Sensitive content

This media may contain sensitive content.

Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Ye Shunguang after training. 💦 Special thanks to VA: Alaitz ♡ VA🌸 ❤️ models by: Hairy Harzoo DaB map by: Lambo🔞COMMS CLOSSED (3/3) #YeShunguang #ZenlessZoneZero #zzzero

vexura3d's profile picture

Vexura

13,418 subscribers

100,616 görüntüleme • 8 ay önce •via X (Twitter)

NSFW#YeShunguang#ZenlessZoneZero#zzzero
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Miyabi's secret training Rigs by : The_Crabnuts BG : Lambo🔞COMMS CLOSSED (3/3) #zzzero #ZZZ #ZenlessZoneZero
0:39

Sensitive content

Miyabi's secret training Rigs by : The_Crabnuts BG : Lambo🔞COMMS CLOSSED (3/3) #zzzero #ZZZ #ZenlessZoneZero

Miiza

1,974,503 görüntüleme • 9 ay önce

Belle Foot Worship Model: DaB Map: Lambo🔞COMMS CLOSSED (3/3) Ability to remove socks aquired ✅ #zzzero #belle
0:53

Sensitive content

Belle Foot Worship Model: DaB Map: Lambo🔞COMMS CLOSSED (3/3) Ability to remove socks aquired ✅ #zzzero #belle

BinaryGloves

755,540 görüntüleme • 4 ay önce

ZZZ Feet Party - Map: Lambo🔞COMMS CLOSSED (3/3) - Belle, Burnice, Lucia, Ellen: DaB - Nangong, Sunna, Aria: Never Yapph - Seed, Alice, Yuzuha: D-CHAN - Remielle, Caesar, Cissia, Jane Doe: 🇲🇽Waffle 🔞 A lot of characters in this one! #zzzero #remielle
2:15

Sensitive content

ZZZ Feet Party - Map: Lambo🔞COMMS CLOSSED (3/3) - Belle, Burnice, Lucia, Ellen: DaB - Nangong, Sunna, Aria: Never Yapph - Seed, Alice, Yuzuha: D-CHAN - Remielle, Caesar, Cissia, Jane Doe: 🇲🇽Waffle 🔞 A lot of characters in this one! #zzzero #remielle

BinaryGloves

795,542 görüntüleme • 2 ay önce

🚨 NOW: Marxist influencer Hasan Piker is facing calls to be VISITED BY THE FBI after he appeared to urge "VIOLENT REVOLUTION" at the Virginia Supreme Court striking down an illegal 10D-1R House map This TRASH wants the DOWNFALL of our society! "Those who make peaceful revolution impossible, make violent revolution inevitable." "The va supreme court denied the results of the redistricting referendum. scotus gutted the voting rights act and tennessee carved up the last dem district destroying black voter power in the state." Flat out urging leftists to violently attack Republicans, lawmakers and judges. This is the kind of rhetoric that has caused 3 assassination attempts on President Trump!
0:49

🚨 NOW: Marxist influencer Hasan Piker is facing calls to be VISITED BY THE FBI after he appeared to urge "VIOLENT REVOLUTION" at the Virginia Supreme Court striking down an illegal 10D-1R House map This TRASH wants the DOWNFALL of our society! "Those who make peaceful revolution impossible, make violent revolution inevitable." "The va supreme court denied the results of the redistricting referendum. scotus gutted the voting rights act and tennessee carved up the last dem district destroying black voter power in the state." Flat out urging leftists to violently attack Republicans, lawmakers and judges. This is the kind of rhetoric that has caused 3 assassination attempts on President Trump!

Eric Daugherty

768,064 görüntüleme • 4 ay önce

HERE IT IS, VeeCon THE MOVIE 2023 QUOTE-TWEET/RE-TWEET with a comment of your favorite part of the movie to enter to win a “GIFT GOAT” VeeFriends Series 2 NFT. Winner gets chosen within 24 hours. Help us spread the word. Sponsored by my alcohol company: Use coupon code VEECON for 35% off. Help support my cousin and myself for growing organic sustainable myrtle berry liquor on our 23 acre farm in Fallbrook, California. Those who buy a bottle of Amati will get a special something in the mail from me :) Must be over 21+ in America to order alcohol. Please drink responsibly. TIMESTAMPS: DAY ONE [0:35] The Barbershop - shoutout to 💈Da Barbershop💈, veeconbarber.eth, and JustTwinn | Afakasi for getting my hair game on point. [1:49] Hangout Hawk Lunch - shoutout to laptoplaura for organizing it. KAWWWWW! [2:23] Leaders On Chain - shoutout to LeadHersOnChain ⁦Valerie | 📍San Diego, CA NattyWife💚VeeCon Alum Winchester VA REALTOR® Holly Roznowski @JapaneseNFTgirl Seema Alexander & Carly aka ComTar for organizing that…oh and meeting KATZ who organized the Veefriends event in Miami during Art Basil. [2:59] PreCon - Ran by @wellbeings_xyz x GentleTornado❤️🌪 [3:20] My first after-party. Shoutout to Jeremy Knows | VeeFriends for taking the time out of his insane schedule to say hi to everyone. DAY TWO [5:50] Spectacular Lunch - shoutout to Big Fish Benny and D8nger.eth for putting this together. It was getting a word in with drock & Dustin Not Justin here. [6:48] VeeCon Welcome Party - shoutout to Gentle Tornado and all the other community speakers. [7:49] Busta Rhymes Concert - Busta Rhymes DAY THREE [10:40] Gift Goat Signing Event - shoutout to danny cole from Creature World for literally signing people’s stuff with proper artwork. Insane dedication. [11:14] VEECON! Shoutout to Jesse Itzler Scott “Scooter” Braun Andrew Schulz 👑HEZI Jordin Sparks and more. [13:16] SHIV’S AFTER PARTY: Is it weird to give a shoutout to me on this one? Haha. Somehow got 800+ people to party together. Idk how we pulled it off. [15:56] SHIV’S AFTER AFTER PARTY: Did I mention I make alcohol called Amati. Buy a bottle. Help a brotha out. Kthx. DAY FOUR [16:31] Unboxing 10 C&C Boxes [18:48] Pulling a spectacular & talking to Gary Vaynerchuk! [19:25] My final farewell after after after party in the JW Lobby. Ends up Adam Ripps dad is super dope. Drink responsibly. Support me by buying a bottle of ‘Amati’: Use coupon code VEECON for 35% off. Help support my cousin and myself for growing organic sustainable myrtle berry liquor on our 23 acre farm in Fallbrook, California. Shout-out to @JassimJust for filming and editing this masterpiece and of course Gary Vaynerchuk, Andy Krainak and the entire Veefriends/Veecon team.
20:58

HERE IT IS, VeeCon THE MOVIE 2023 QUOTE-TWEET/RE-TWEET with a comment of your favorite part of the movie to enter to win a “GIFT GOAT” VeeFriends Series 2 NFT. Winner gets chosen within 24 hours. Help us spread the word. Sponsored by my alcohol company: Use coupon code VEECON for 35% off. Help support my cousin and myself for growing organic sustainable myrtle berry liquor on our 23 acre farm in Fallbrook, California. Those who buy a bottle of Amati will get a special something in the mail from me :) Must be over 21+ in America to order alcohol. Please drink responsibly. TIMESTAMPS: DAY ONE [0:35] The Barbershop - shoutout to 💈Da Barbershop💈, veeconbarber.eth, and JustTwinn | Afakasi for getting my hair game on point. [1:49] Hangout Hawk Lunch - shoutout to laptoplaura for organizing it. KAWWWWW! [2:23] Leaders On Chain - shoutout to LeadHersOnChain ⁦Valerie | 📍San Diego, CA NattyWife💚VeeCon Alum Winchester VA REALTOR® Holly Roznowski @JapaneseNFTgirl Seema Alexander & Carly aka ComTar for organizing that…oh and meeting KATZ who organized the Veefriends event in Miami during Art Basil. [2:59] PreCon - Ran by @wellbeings_xyz x GentleTornado❤️🌪 [3:20] My first after-party. Shoutout to Jeremy Knows | VeeFriends for taking the time out of his insane schedule to say hi to everyone. DAY TWO [5:50] Spectacular Lunch - shoutout to Big Fish Benny and D8nger.eth for putting this together. It was getting a word in with drock & Dustin Not Justin here. [6:48] VeeCon Welcome Party - shoutout to Gentle Tornado and all the other community speakers. [7:49] Busta Rhymes Concert - Busta Rhymes DAY THREE [10:40] Gift Goat Signing Event - shoutout to danny cole from Creature World for literally signing people’s stuff with proper artwork. Insane dedication. [11:14] VEECON! Shoutout to Jesse Itzler Scott “Scooter” Braun Andrew Schulz 👑HEZI Jordin Sparks and more. [13:16] SHIV’S AFTER PARTY: Is it weird to give a shoutout to me on this one? Haha. Somehow got 800+ people to party together. Idk how we pulled it off. [15:56] SHIV’S AFTER AFTER PARTY: Did I mention I make alcohol called Amati. Buy a bottle. Help a brotha out. Kthx. DAY FOUR [16:31] Unboxing 10 C&C Boxes [18:48] Pulling a spectacular & talking to Gary Vaynerchuk! [19:25] My final farewell after after after party in the JW Lobby. Ends up Adam Ripps dad is super dope. Drink responsibly. Support me by buying a bottle of ‘Amati’: Use coupon code VEECON for 35% off. Help support my cousin and myself for growing organic sustainable myrtle berry liquor on our 23 acre farm in Fallbrook, California. Shout-out to @JassimJust for filming and editing this masterpiece and of course Gary Vaynerchuk, Andy Krainak and the entire Veefriends/Veecon team.

SHIVALRY

36,363 görüntüleme • 3 yıl önce

Finally video emerges of an actual Ukrainian attack in Toretsk, two weeks after their supposed ninja counteroffensive kicked off... or is it?⬇️ I'm referring to this video, which was published Wednesday, showing the destruction of two AFU M113s withdrawing troops from a southern suburb of Toretsk (Zabalka) that basically all mappers had placed well inside Russian lines (video 1, see figure 3 for the map). Video emerged later that day of Russian infantry destroying an apparently Ukrainian-held house in the same area with a satchel charge after suppressing the defenders with an RPG and small arms (the first segment of video 2). Interesting, I thought. So I looked at the map again, harder. And then I looked at the geolocated positions of Russian units, which the TG channel Creamy Caprice keeps a map of (figure 3, showing the location of the first video - the red and blue dots mark Russian and Ukrainian positions logged over the entire course of the battle). And that's when it struck me - Russian troops have never been spotted in the northwestern corner of the Zabalka district. They've been seen in the central part and they've been seen around the slag heaps, but not the northwest corner. And there's a relatively short route into it through the Ukrainian-held forest to the northwest, although the last mile is through an open field around the base of (and dominated by) the nearby slag heap. It would be an extremely dangerous journey. Now we come to what the video actually depicts - the withdrawal of troops. The first APC is hit while withdrawing and the escaping dismounts are effectively engaged after going to ground in the forest. The second APC then arrives in Zabalka, loads up, and is heavily hit and knocked out as it withdraws. A couple soldiers are seen heading deeper into the city on foot as it departs, but not a full squad - there may not have been room for them on the transport, or they may have (correctly) thought their chances were better on foot. Then the subsequent video emerged of Russian infantry clearing holdouts in the area, which could have very well been those men. So what happened here? Well, I pointed out earlier that it's standard practice to back-clear an urban area after taking it, to clear bypassed enemy strongholds and booby-traps and render the area safe to support further operations. I suspect there was actually a pocket of bypassed Ukrainian troops holed up in the western Zabalka District, a couple kilometers behind Russian lines, whom the AFU command tried to evacuate via APC earlier this week while they still had the chance to do so. And the Russians may very well have allowed those APCs to pull in and load up so they could - as they did - very coldly kill them while they were packed with infantry to withdraw. Why did the Ukrainian command undertake such a high-risk operation instead of simply writing these men off and telling them that it was every man for himself? Probably because these were Azov fighters and thus entitled to special treatment and consideration - one of their brigades operates in the area. In any event, far from suggesting a Ukrainian counterattack into south Toretsk, this turn of events suggests to me that the Russians are instead mopping up remaining resistance in the city. (As an addendum, the people behind Creamy Caprice think that the third segment of video 2 shows Ukrainian activity somewhat farther into the northwest corner of the Zabalka Dictrict, but that segment also doesn't show live troops - for all we know they were bombing an AFU comms repeater or something on the roof of that building. The second segment of that video is old judging by the snow on the ground.)
1:42

Finally video emerges of an actual Ukrainian attack in Toretsk, two weeks after their supposed ninja counteroffensive kicked off... or is it?⬇️ I'm referring to this video, which was published Wednesday, showing the destruction of two AFU M113s withdrawing troops from a southern suburb of Toretsk (Zabalka) that basically all mappers had placed well inside Russian lines (video 1, see figure 3 for the map). Video emerged later that day of Russian infantry destroying an apparently Ukrainian-held house in the same area with a satchel charge after suppressing the defenders with an RPG and small arms (the first segment of video 2). Interesting, I thought. So I looked at the map again, harder. And then I looked at the geolocated positions of Russian units, which the TG channel Creamy Caprice keeps a map of (figure 3, showing the location of the first video - the red and blue dots mark Russian and Ukrainian positions logged over the entire course of the battle). And that's when it struck me - Russian troops have never been spotted in the northwestern corner of the Zabalka district. They've been seen in the central part and they've been seen around the slag heaps, but not the northwest corner. And there's a relatively short route into it through the Ukrainian-held forest to the northwest, although the last mile is through an open field around the base of (and dominated by) the nearby slag heap. It would be an extremely dangerous journey. Now we come to what the video actually depicts - the withdrawal of troops. The first APC is hit while withdrawing and the escaping dismounts are effectively engaged after going to ground in the forest. The second APC then arrives in Zabalka, loads up, and is heavily hit and knocked out as it withdraws. A couple soldiers are seen heading deeper into the city on foot as it departs, but not a full squad - there may not have been room for them on the transport, or they may have (correctly) thought their chances were better on foot. Then the subsequent video emerged of Russian infantry clearing holdouts in the area, which could have very well been those men. So what happened here? Well, I pointed out earlier that it's standard practice to back-clear an urban area after taking it, to clear bypassed enemy strongholds and booby-traps and render the area safe to support further operations. I suspect there was actually a pocket of bypassed Ukrainian troops holed up in the western Zabalka District, a couple kilometers behind Russian lines, whom the AFU command tried to evacuate via APC earlier this week while they still had the chance to do so. And the Russians may very well have allowed those APCs to pull in and load up so they could - as they did - very coldly kill them while they were packed with infantry to withdraw. Why did the Ukrainian command undertake such a high-risk operation instead of simply writing these men off and telling them that it was every man for himself? Probably because these were Azov fighters and thus entitled to special treatment and consideration - one of their brigades operates in the area. In any event, far from suggesting a Ukrainian counterattack into south Toretsk, this turn of events suggests to me that the Russians are instead mopping up remaining resistance in the city. (As an addendum, the people behind Creamy Caprice think that the third segment of video 2 shows Ukrainian activity somewhat farther into the northwest corner of the Zabalka Dictrict, but that segment also doesn't show live troops - for all we know they were bombing an AFU comms repeater or something on the roof of that building. The second segment of that video is old judging by the snow on the ground.)

Armchair Warlord

15,294 görüntüleme • 1 yıl önce

I know your timeline is flooded now with word salads of "insane, HER, 10 features you missed, we're so back". Sit down. Chill. Take a deep breath like Mark does in the demo . Let's think step by step: - Technique-wise, OpenAI has figured out a way to map audio to audio directly as first-class modality, and stream videos to a transformer in real-time. These require some new research on tokenization and architecture, but overall it's a data and system optimization problem (as most things are). High-quality data can come from at least 2 sources: 1) Naturally occurring dialogues on YouTube, podcasts, TV series, movies, etc. Whisper can be trained to identify speaker turns in a dialogue or separate overlapping speeches for automated annotation. 2) Synthetic data. Run the slow 3-stage pipeline using the most powerful models: speech1->text1 (ASR), text1->text2 (LLM), text2->speech2 (TTS). The middle LLM can decide when to stop and also simulate how to resume from interruption. It could output additional "thought traces" that are not verbalized to help generate better reply. Then GPT-4o distills directly from speech1->speech2, with optional auxiliary loss functions based on the 3-stage data. After distillation, these behaviors are now baked into the model without emitting intermediate texts. On the system side: the latency would not meet real-time threshold if every video frame is decompressed into an RGB image. OpenAI has likely developed their own neural-first, streaming video codec to transmit the motion deltas as tokens. The communication protocol and NN inference must be co-optimized. For example, there could be a small and energy-efficient NN running on the edge device that decides to transmit more tokens if the video is interesting, and fewer otherwise. - I didn't expect GPT-4o to be closer to GPT-5, the rumored "Arrakis" model that takes multimodal in and out. In fact, it's likely an early checkpoint of GPT-5 that hasn't finished training yet. The branding betrays a certain insecurity. Ahead of Google I/O, OpenAI would rather beat our mental projection of GPT-4.5 than disappoint by missing the sky-high expectation for GPT-5. A smart move to buy more time. - Notably, the assistant is much more lively and even a bit flirty. GPT-4o is trying (perhaps a bit too hard) to sound like HER. OpenAI is eating Character AI's lunch, with almost 100% overlap in form factor and huge distribution channels. It's a pivot towards more emotional AI with strong personality, which OpenAI seemed to actively suppress in the past. - Whoever wins Apple first wins big time. I see 3 levels of integration with iOS: 1) Ditch Siri. OpenAI distills a smaller-tier, purely on-device GPT-4o for iOS, with optional paid upgrade to use the cloud. 2) Native features to stream the camera or screen into the model. Chip-level support for neural audio/video codec. 3) Integrate with iOS system-level action API and smart home APIs. No one uses Siri Shortcuts, but it's time to resurrect. This could become the AI agent product with a billion users from the get-go. The FSD for smartphones with a Tesla-scale data flywheel.
14:22

I know your timeline is flooded now with word salads of "insane, HER, 10 features you missed, we're so back". Sit down. Chill. Take a deep breath like Mark does in the demo . Let's think step by step: - Technique-wise, OpenAI has figured out a way to map audio to audio directly as first-class modality, and stream videos to a transformer in real-time. These require some new research on tokenization and architecture, but overall it's a data and system optimization problem (as most things are). High-quality data can come from at least 2 sources: 1) Naturally occurring dialogues on YouTube, podcasts, TV series, movies, etc. Whisper can be trained to identify speaker turns in a dialogue or separate overlapping speeches for automated annotation. 2) Synthetic data. Run the slow 3-stage pipeline using the most powerful models: speech1->text1 (ASR), text1->text2 (LLM), text2->speech2 (TTS). The middle LLM can decide when to stop and also simulate how to resume from interruption. It could output additional "thought traces" that are not verbalized to help generate better reply. Then GPT-4o distills directly from speech1->speech2, with optional auxiliary loss functions based on the 3-stage data. After distillation, these behaviors are now baked into the model without emitting intermediate texts. On the system side: the latency would not meet real-time threshold if every video frame is decompressed into an RGB image. OpenAI has likely developed their own neural-first, streaming video codec to transmit the motion deltas as tokens. The communication protocol and NN inference must be co-optimized. For example, there could be a small and energy-efficient NN running on the edge device that decides to transmit more tokens if the video is interesting, and fewer otherwise. - I didn't expect GPT-4o to be closer to GPT-5, the rumored "Arrakis" model that takes multimodal in and out. In fact, it's likely an early checkpoint of GPT-5 that hasn't finished training yet. The branding betrays a certain insecurity. Ahead of Google I/O, OpenAI would rather beat our mental projection of GPT-4.5 than disappoint by missing the sky-high expectation for GPT-5. A smart move to buy more time. - Notably, the assistant is much more lively and even a bit flirty. GPT-4o is trying (perhaps a bit too hard) to sound like HER. OpenAI is eating Character AI's lunch, with almost 100% overlap in form factor and huge distribution channels. It's a pivot towards more emotional AI with strong personality, which OpenAI seemed to actively suppress in the past. - Whoever wins Apple first wins big time. I see 3 levels of integration with iOS: 1) Ditch Siri. OpenAI distills a smaller-tier, purely on-device GPT-4o for iOS, with optional paid upgrade to use the cloud. 2) Native features to stream the camera or screen into the model. Chip-level support for neural audio/video codec. 3) Integrate with iOS system-level action API and smart home APIs. No one uses Siri Shortcuts, but it's time to resurrect. This could become the AI agent product with a billion users from the get-go. The FSD for smartphones with a Tesla-scale data flywheel.

Jim Fan

992,016 görüntüleme • 2 yıl önce

I had to test it myself to believe this unreal inference speed. 3,000 tokens/s for 1 user on standard datacenter GPUs. They leveraged a hidden efficiency gap in how GPUs generate tokens. Kog just achieved 3,000 tokens/s on 8× AMD MI300X GPUs and 2,100 on 8× NVIDIA H200 (FP16, no speculative decoding). Their tech preview is on a 2B model, and they show how their techniques will scale to large frontier MoE models at similar speeds. That's a huge number because normal low-batch GPU decoding for 2B to 8B models is usually closer to 100 to 300 tokens/s per request, so Kog is claiming something like a 10X to 30X jump in the speed one user actually feels. Their trick: they are getting the speed by treating LLM decoding as a memory streaming problem, not mainly a math problem. For 1 user at batch size 1, the GPU is not doing big, efficient matrix-matrix work like in training or large-batch serving; it is repeatedly pulling the model’s active weights from high-bandwidth memory for each new token, so speed depends on how smoothly those weights keep flowing. Normal inference stacks keep breaking that flow. They run many separate GPU programs for different parts of the model, move intermediate results through memory, wait at synchronization points, talk back to the CPU for scheduling or sampling, and then repeat this token after token. Kog’s answer is to co-design 3 things that are usually tuned separately: the runtime, the low-level GPU code, and the model architecture. The biggest engineering move is the monokernel, where the whole decode pass runs as 1 persistent GPU-resident program, including sampling, so the system does not keep stopping for kernel launches, CPU scheduling, and intermediate memory round trips. They also rebuilt synchronization, because their own measurements say grid sync was eating around 35% of token-generation time; instead of making every compute unit wait at a broad barrier, each unit waits only for the exact data it needs. On AMD MI300X, they also map memory access around the chiplet layout, because memory latency changes depending on which die makes the request. Then their Laneformer model uses Delayed Tensor Parallelism, which lets cross-GPU communication happen in the background instead of blocking every layer.
1:08

I had to test it myself to believe this unreal inference speed. 3,000 tokens/s for 1 user on standard datacenter GPUs. They leveraged a hidden efficiency gap in how GPUs generate tokens. Kog just achieved 3,000 tokens/s on 8× AMD MI300X GPUs and 2,100 on 8× NVIDIA H200 (FP16, no speculative decoding). Their tech preview is on a 2B model, and they show how their techniques will scale to large frontier MoE models at similar speeds. That's a huge number because normal low-batch GPU decoding for 2B to 8B models is usually closer to 100 to 300 tokens/s per request, so Kog is claiming something like a 10X to 30X jump in the speed one user actually feels. Their trick: they are getting the speed by treating LLM decoding as a memory streaming problem, not mainly a math problem. For 1 user at batch size 1, the GPU is not doing big, efficient matrix-matrix work like in training or large-batch serving; it is repeatedly pulling the model’s active weights from high-bandwidth memory for each new token, so speed depends on how smoothly those weights keep flowing. Normal inference stacks keep breaking that flow. They run many separate GPU programs for different parts of the model, move intermediate results through memory, wait at synchronization points, talk back to the CPU for scheduling or sampling, and then repeat this token after token. Kog’s answer is to co-design 3 things that are usually tuned separately: the runtime, the low-level GPU code, and the model architecture. The biggest engineering move is the monokernel, where the whole decode pass runs as 1 persistent GPU-resident program, including sampling, so the system does not keep stopping for kernel launches, CPU scheduling, and intermediate memory round trips. They also rebuilt synchronization, because their own measurements say grid sync was eating around 35% of token-generation time; instead of making every compute unit wait at a broad barrier, each unit waits only for the exact data it needs. On AMD MI300X, they also map memory access around the chiplet layout, because memory latency changes depending on which die makes the request. Then their Laneformer model uses Delayed Tensor Parallelism, which lets cross-GPU communication happen in the background instead of blocking every layer.

Rohan Paul

13,282 görüntüleme • 3 ay önce

My fox shooting garden defending AI robot is finally done and WORKING! 🤩 (Don’t worry it only shoots 💦 water) After months of slowly moving forward with each part I finished the last step to train a TensorFlow model on the footage of the 🦊 fox I collected hours of footage 📹 with the fox roaming around my garden, from this I labeled around 2000 images with the fox by hand ✋ Honestly, I was quite skeptical training the model was actually gonna work, maybe this was partly the reason I avoided working on this until the very end. If I couldn’t train a model to detect the fox, this whole robot would never be able to function properly. On the flipside though, with no previous experience in hardware or electronics there was a bit of a learning curve and I didn’t want to end up labeling thousands of images, training a TensorFlow model, only to fail on building the hardware. As I started building, I realized that mixing hardware and software adds quite another dimension to debugging things. At times I wasted hours debugging code in my IDE, only to realize the issue was somewhere in the electronics. Furthermore, combining this side project with a full time job and a young family, is not always easy. It can be quite frustrating, to know you only need 4 hours of concentrated effort for a small task, having to spread it out across a week of 20min increments. Then, a few months into the build I noticed the fox had stopped coming to my garden, in fact one day, I recorded her walking with 3 cute little 🐶 pups, and the next day I saw her moving out of my garden completely. Did she know I was building a robot? I had this strange mix of feelings, happy my garden was safe from poop and digging, happy she was safe with her pups, but how was I gonna finish this project if my robot had no fox to detect? For sure they would be back next year, I figured I could postpone the whole thing until next winter, but I also knew it was gonna be much harder to pick up momentum if I did let it sit there for six months. So I decided to keep working, hoping the fox would reappear,.. but she never did. As I finished labeling the footage and started training my model, I could finally see the mAP results, quantifying the precision of my object detection model. It was measuring at 78% across different metrics on detecting my fox. I quickly ran the model on some of the video footage I got from my fox. Inference speed took a hit, but it did a near perfect job detecting the fox, even when she was deep down in the grass or wizzing past in a motion blur. It took me by surprise how well it worked. With the default model I had to drop my confidence threshold way down to 15%, to recognize the fox as 🦜“bird” in one or two frames, with my custom model it followed the fox all the way down to the back of the garden! Still this didn’t solve the issue of there being no actual fox in my garden and how was I gonna wrap this project in a short timeframe. I played with the idea of putting a fox toy 🧸 on an RC 🚗 car, or borrowing a dog to run around the garden to test. Friends suggested I run around the garden in a fox costume.. what a ridiculous idea. I wasn’t really feeling the idea of running around the garden in a floppy cloth fox 🎭 costume, but had a look anyway. I came across these self inflating costumes. This actually could be perfect. Since it’s inflated, it would hold its shape super well, making it much easier to label, train and be recognized by my robot. So I got the costume and shot a time lapse of myself as a fox walking around the garden. I labeled it to around 600 images. Ran the model training again and got a mAP result of 82%. This was even better than my real fox! At this point I knew this was gonna work. So here’s the final 🎥 video, just having some fun with it. I’ll update here whenever the real fox does come back. On a final note, I’m looking for (remote) jobs in these fields of AI now: - object detection - visual generative AI - 3D (nerfs + gaussian splats) So if you know anything let me know! My DMs are open 😊
2:05

My fox shooting garden defending AI robot is finally done and WORKING! 🤩 (Don’t worry it only shoots 💦 water) After months of slowly moving forward with each part I finished the last step to train a TensorFlow model on the footage of the 🦊 fox I collected hours of footage 📹 with the fox roaming around my garden, from this I labeled around 2000 images with the fox by hand ✋ Honestly, I was quite skeptical training the model was actually gonna work, maybe this was partly the reason I avoided working on this until the very end. If I couldn’t train a model to detect the fox, this whole robot would never be able to function properly. On the flipside though, with no previous experience in hardware or electronics there was a bit of a learning curve and I didn’t want to end up labeling thousands of images, training a TensorFlow model, only to fail on building the hardware. As I started building, I realized that mixing hardware and software adds quite another dimension to debugging things. At times I wasted hours debugging code in my IDE, only to realize the issue was somewhere in the electronics. Furthermore, combining this side project with a full time job and a young family, is not always easy. It can be quite frustrating, to know you only need 4 hours of concentrated effort for a small task, having to spread it out across a week of 20min increments. Then, a few months into the build I noticed the fox had stopped coming to my garden, in fact one day, I recorded her walking with 3 cute little 🐶 pups, and the next day I saw her moving out of my garden completely. Did she know I was building a robot? I had this strange mix of feelings, happy my garden was safe from poop and digging, happy she was safe with her pups, but how was I gonna finish this project if my robot had no fox to detect? For sure they would be back next year, I figured I could postpone the whole thing until next winter, but I also knew it was gonna be much harder to pick up momentum if I did let it sit there for six months. So I decided to keep working, hoping the fox would reappear,.. but she never did. As I finished labeling the footage and started training my model, I could finally see the mAP results, quantifying the precision of my object detection model. It was measuring at 78% across different metrics on detecting my fox. I quickly ran the model on some of the video footage I got from my fox. Inference speed took a hit, but it did a near perfect job detecting the fox, even when she was deep down in the grass or wizzing past in a motion blur. It took me by surprise how well it worked. With the default model I had to drop my confidence threshold way down to 15%, to recognize the fox as 🦜“bird” in one or two frames, with my custom model it followed the fox all the way down to the back of the garden! Still this didn’t solve the issue of there being no actual fox in my garden and how was I gonna wrap this project in a short timeframe. I played with the idea of putting a fox toy 🧸 on an RC 🚗 car, or borrowing a dog to run around the garden to test. Friends suggested I run around the garden in a fox costume.. what a ridiculous idea. I wasn’t really feeling the idea of running around the garden in a floppy cloth fox 🎭 costume, but had a look anyway. I came across these self inflating costumes. This actually could be perfect. Since it’s inflated, it would hold its shape super well, making it much easier to label, train and be recognized by my robot. So I got the costume and shot a time lapse of myself as a fox walking around the garden. I labeled it to around 600 images. Ran the model training again and got a mAP result of 82%. This was even better than my real fox! At this point I knew this was gonna work. So here’s the final 🎥 video, just having some fun with it. I’ll update here whenever the real fox does come back. On a final note, I’m looking for (remote) jobs in these fields of AI now: - object detection - visual generative AI - 3D (nerfs + gaussian splats) So if you know anything let me know! My DMs are open 😊

Jeroen Pixel

55,797 görüntüleme • 2 yıl önce

Call on all Global Pioneers to join the Vow Praying Ceremony on April 13th GMT 5: 30-6:00 am, USA EST 1:30 am- 2:00 am online at Voov Meeting, using the above room number. If you're unable to participate, you're welcome to take some time to meditate and pray for the success of the Pi Network on Pi2Day, 2024 - with a GCV of $314,159.🙏🙏🙏 There is a special type of energy that is created when people come together to pray at the same time. Studies have shown that good intentions have a massive impact on the world around us. For example, Lake Biwa in Japan had been severely polluted for over 20 years, but after 350 people prayed for it one morning, the lake became clear and clean after just a week. This is why I'm calling on all Pi Network Pioneers to join the vow praying ceremony. If you're unable to attend, I suggest you pray and meditate at home. During the ceremony, I will read the vow and pray for the success of the Pi Network in both English and Chinese.🙏🙏🙏❤️❤️❤️ Yesterday, I listened to two versions of Pi Network songs from Burundi, and they were incredibly touching. I could feel their deep appreciation for the Pi Network. I believe that most of our Pioneers hold a grateful heart towards the Pi Network. I'm proud that our Pioneers are full of love and compassion, and this is why I believe that the Pi Network will succeed. To show our sincerity, I will fast and bathe for three days. In ancient China, we would keep our spirit and body clean before important ceremonies so that we can gain higher universal power and energy. You're welcome to do the same. I express my gratitude to the CEO of China Ni Diao Group, Ms. Jia Yu Chen, and all the top management of Ni Diao Group, as well as Pi Network GCV Community China Ambassador, Mr. Daniel, and the entire Conference Organization Committee, for their support and efforts. The forthcoming Ni Diao Group Global GCV International Commodity Trade Demonstration and Entertainment Exchange Seminar, to be held in Anshun City, Guizhou Province, the hometown of Pi Network Founder Dr. Fan. Ni Diao is expected to become Pi Network's strongest ecosystem after OM, and possibly even before it. This seminar will serve as the cradle of international trade, where Pi will be used as payment. All Ni Diao Group members will participate in the GCV Pi2Day OM petition, representing the global ecosystem. I deeply appreciate Ms. Chen's support of GCV $314,159 as an ecosystem, and for her remarkable efforts and dedication to cross-border barter. 🙏🙏🙏 The upcoming seminar is poised to be a significant event, with an estimated attendance of over 1000, including Ni Diao Group along with other notable Chinese GCV merchants and pioneers. The seminar will feature speeches by global importers and exporters, providing an excellent opportunity to explore global barter prospects. As this is an offline seminar, attendees can look forward to additional performances and tours. It is recommended to review the conference schedule to select the most suitable time to attend the speech time for online pioneers. I hope that the conference achieves a successful outcome. Best wishes for the event. I'm looking forward to seeing you all soon on the Voov Meeting.🌹🌹🌹 P.S. Sorry for not replying to many Pioneers DM messages. Due to my busy schedule and high volume of DMs, I cannot respond to all of them. Please follow my updates on Twitter or the group for more information. Thanks for understanding. Global GCV Ambassador: Doris Yin 🪷🪷🪷🪷3
0:29

Call on all Global Pioneers to join the Vow Praying Ceremony on April 13th GMT 5: 30-6:00 am, USA EST 1:30 am- 2:00 am online at Voov Meeting, using the above room number. If you're unable to participate, you're welcome to take some time to meditate and pray for the success of the Pi Network on Pi2Day, 2024 - with a GCV of $314,159.🙏🙏🙏 There is a special type of energy that is created when people come together to pray at the same time. Studies have shown that good intentions have a massive impact on the world around us. For example, Lake Biwa in Japan had been severely polluted for over 20 years, but after 350 people prayed for it one morning, the lake became clear and clean after just a week. This is why I'm calling on all Pi Network Pioneers to join the vow praying ceremony. If you're unable to attend, I suggest you pray and meditate at home. During the ceremony, I will read the vow and pray for the success of the Pi Network in both English and Chinese.🙏🙏🙏❤️❤️❤️ Yesterday, I listened to two versions of Pi Network songs from Burundi, and they were incredibly touching. I could feel their deep appreciation for the Pi Network. I believe that most of our Pioneers hold a grateful heart towards the Pi Network. I'm proud that our Pioneers are full of love and compassion, and this is why I believe that the Pi Network will succeed. To show our sincerity, I will fast and bathe for three days. In ancient China, we would keep our spirit and body clean before important ceremonies so that we can gain higher universal power and energy. You're welcome to do the same. I express my gratitude to the CEO of China Ni Diao Group, Ms. Jia Yu Chen, and all the top management of Ni Diao Group, as well as Pi Network GCV Community China Ambassador, Mr. Daniel, and the entire Conference Organization Committee, for their support and efforts. The forthcoming Ni Diao Group Global GCV International Commodity Trade Demonstration and Entertainment Exchange Seminar, to be held in Anshun City, Guizhou Province, the hometown of Pi Network Founder Dr. Fan. Ni Diao is expected to become Pi Network's strongest ecosystem after OM, and possibly even before it. This seminar will serve as the cradle of international trade, where Pi will be used as payment. All Ni Diao Group members will participate in the GCV Pi2Day OM petition, representing the global ecosystem. I deeply appreciate Ms. Chen's support of GCV $314,159 as an ecosystem, and for her remarkable efforts and dedication to cross-border barter. 🙏🙏🙏 The upcoming seminar is poised to be a significant event, with an estimated attendance of over 1000, including Ni Diao Group along with other notable Chinese GCV merchants and pioneers. The seminar will feature speeches by global importers and exporters, providing an excellent opportunity to explore global barter prospects. As this is an offline seminar, attendees can look forward to additional performances and tours. It is recommended to review the conference schedule to select the most suitable time to attend the speech time for online pioneers. I hope that the conference achieves a successful outcome. Best wishes for the event. I'm looking forward to seeing you all soon on the Voov Meeting.🌹🌹🌹 P.S. Sorry for not replying to many Pioneers DM messages. Due to my busy schedule and high volume of DMs, I cannot respond to all of them. Please follow my updates on Twitter or the group for more information. Thanks for understanding. Global GCV Ambassador: Doris Yin 🪷🪷🪷🪷3

Doris Yin 东方紫莲🪷

33,845 görüntüleme • 2 yıl önce

Dear ICP community, the Internet Computer has now been running strong for 5 years 👏👏👏 Here is a celebratory preview of ICP "cloud engines," the sovereign frontier cloud technology the network shall soon provide from Main points: — Cloud engines enable anyone to spin up their own sovereign frontier cloud. The technology involves an extraordinary inventive step, in which cloud is created from a mathematically secure network of nodes. The nodes run as part of the Internet Computer network ( but are selected and configured by the cloud engine's owner. — The frontier cloud provided by engines is strongly focused on enabling AI agents to build and update online applications and services for us. The world is changing fast, and nearly all new online apps and services are already being built with the help of AI, and thus cloud engines target the future of cloud. — Software hosted on cloud engines is tamperproof, which means that it is immune to infrastructure hacks, because it runs inside a mathematically secure network protocol, rather than on computers directly. This means that AI agents, and those building with them, don't need to have a security team in the loop, or to trust someone else's security team. This is crucial, because in the future, non technical people will demand the freedom to build with full automation — where they just need to issue instructions to AI about what to build, and don't need to worry about anything or anyone else. Of course, apps and services running on engines are also vastly safer from the new breed of hacker being enabled by frontier AI. (The cloud engines themselves are also "tamperproof." Even if a hacker gains physical access to some portion of a cloud engine's nodes, and can make arbitrary changes, the computations and data of the hosted apps and services cannot be corrupted or interrupted so long as the network's fault bounds aren't exceeded. The recent hack of Vercel, a major cloud platform, which gave hackers access to the apps it hosted, provides additional perspective on the importance of this advantage.) — Software hosted on cloud engines is guaranteed to run, so long as a sufficient number of the engine's nodes are running. This means that AI can build applications and services without the need to have a human systems admin team constantly tinkering with the underlying platform to keep it running, which is again crucial, because in the future, non technical people will expect the freedom to use AI to build without the support of others. — New frontier programming language technology, in the form of the Motoko language developed by Caffeine Labs, leverages seminal "orthogonal persistence" technology that unifies program logic and data to deliver further unlocks for AI (Motoko is the first computer language being developed that targets agents that are writing software rather than humans engineers per se). Nowadays, AI can build and update production apps at a prodigious rate, even at the speed of conversation. But it can also make mistakes, and there's a risk that an update it creates might be "lossy" in the sense it causes some transformed data to be lost. Again, in this new world, it's both undesirable and impractical for everyone to have to have a systems admin team on-hand to detect lossy updates and roll them back, but Motoko provides a solution: it can detect new software updates are lossy before they are applied, reducing potentially catastrophic errors by AI to harmless coding retries. — Software hosted on cloud engines is "serverless" but unlike traditional serverless software, directly it directly incorporates data through "orthogonal persistence." Another key purpose is simplify backend software logic and fuel the modeling power of AI by increasing abstraction (sorry for the technical language!!!). Put simply, this enables AI to produce more sophisticated backends, faster, and at dramatically lower costs, as measured by the number AI API tokens consumed during coding. (Tip for the technical: orthogonal persistence is a new paradigm where "the program is the database," and data lives inside program variables, which is possible because it's as if hosted software runs forever in persistent memory). — An expanding database of skills at shall make it possible to develop and directly deploy apps and services to your cloud engines directly from Claude Code, Perplexity, Codex and other AI platforms. Further, your account on can be connected, so that new apps and updates created through conversation automatically appear hosted from your cloud engine. In the future, R&D is going to be very seamless. You converse with AI, and your secure and unstoppable apps or services are created or updated. Cloud engines are designed to directly support this "self-writing cloud" future where we can work hands-free. — Tech sovereignty is becoming a huge issue worldwide, with governments and corporations seeking to create sovereign tech stacks owing to geopolitical tensions. Increasingly, people are realizing that tech provided by foreign nations can come with hidden backdoors and kills switches, from the base platform, right up through hosted apps and services. ICP technology is open source, and those building on ICP using AI own their own source code. When you have the source code, you can verify that there are no backdoors, and when you own the source code thanks to AI, you can update it at will, freeing you from vendor lock-in. But cloud engines take sovereignty much further... — You create a cloud engine by selecting the nodes that will be combined. You can choose the class of nodes used, and their number, but more importantly, you can choose who operates the nodes, and where they are located. Almost any configuration is possible, because the Internet Computer scales the security privileges afforded to hosted software within the network according to configuration (software hosted on cloud engines can directly interoperate with software on other engines and traditional subnets, but base restrictions are applied according to security rules). A cloud engine can be created within a region such as Europe, to comply with regs such as GDPR, or completely within a sovereign state like Switzerland or Pakistan. But cloud engines go further still... — Sovereignty is also about freedom from vendor lock-in. Cloud engines are essentially ICP (Internet Computer Protocol) network configurations, and this means the underlying compute nodes they combine can be swapped out without interrupting their hosted apps and services. This is a big deal. In addition, cloud engines now support nodes that are instances running on Big Tech's clouds, in addition to nodes that are dedicated specialized hardware, as per the Gen I and Gen II nodes that dominate the Internet Computer today. For example, it is possible to have an engine running across different AWS data centers, say, and then reconfigure the engine to run across a mixture of AWS, Google, Azure and Hetzner for even more resilience, without the users of hosted apps and services noticing a thing. That's true freedom. — Sovereign AI is becoming increasingly important too, and cloud engines allow special "AI nodes" to be added to them, so that hosted software can perform inference on hardware provisioned by the owner from a location the owner has selected. Even though the AI nodes are only accessible within the cloud engine, they can still benefit from the forthcoming Internet Intelligence Gateway (IG), which will make it possible to validate inference performed on key frontier open weights LLMs, even when the inference is performed on completely independent AI clouds. When the results of inference are received, this technology can verify that neither the prompt+context (input) nor the inference result (output) have been modified, and that the results were produced by the precise LLM expected. This ensures that AI clouds don't cheat by running inference on cheaper models than are being paid for, and bad actors aren't modifying the inputs or outputs to surreptitiously insert advertising into results, say, or change facts, or insert malware when code is being generated. What's super cool about this technology is the cost of the verification is scalable. A very valuable additional security can be achieved with only 1-2% of extra cost. — Scaling apps and services when they hit capacity limits is another thorny problem that cloud engines help the world address. Engines make scaling possible without rewriting or reconfiguring software. The query workload capacity of hosted software can be horizontally scaled simply by adding new nodes to an engine, and nodes can also be added in geographical proximity to demand. Meanwhile, update workload capacity can first be scaled-up by swapping an engine's nodes out for the next class up, and then when no larger class of node is available, horizontally scaled-out by "splitting" the engine into two, which doubles available capacity. (Technical tip: horizontally scaling update capacity by splitting engines requires multi-canister architectures). — For those who have been following how Caffeine builds apps that can efficiently store large numbers of files, I should mention that apps built on cloud engines will also support the new ICP Blob Storage cloud network (since cloud engines currently have up to about 3 TB of memory, which apps storing large amounts of files can easily exceed). We are also working on allowing blob storage nodes to be added to cloud engines, to enable sovereign mass blob storage within an engine, similarly to how AI nodes can be added currently. — Lastly, but certainly not least, I should mention that cloud engines are multi-blockchain capable, and ready for digital assets, thanks to the clever math at their core. For example, an e-commerce service built on a cloud engine can securely accept and custody stablecoin payments, or a multi-chain DEX could be hosted. Further, engines can support software autonomy (software orchestrated and controlled by other autonomous software, in a decentralized way) and can themselves be orchestrated by SNS technology, and thus run autonomously too. Today, though, the focus is on *mainstream* cloud. This year, the cloud industry will generate approximately one trillion dollars in revenue. That number is already huge, but is expected to grow to two trillion dollars by 2030. After years of continuous development, which have seen more than $500m spent on R&D, the Internet Computer network is now tacking directly toward this mainstream cloud market with cloud engine technology. In their first version, cloud engines are not meant to be a cloud panacea. For example, currently they are not ideal for working with big data. You should use something like DataBricks for that. Cloud engines are carefully targeted at enabling AI to produce traditional online applications and services, including SaaS, in a safer and more productive way, which represents a new market segment with tremendous potential. Of course, DFINITY will continue to work relentlessly to push forward ICP's capabilities, so expect further developments. It's worth mentioning that this cloud segment isn't just about creating new apps and services using AI, it's also about replacing legacy systems and apps built on super expensive SaaS services. Caffeine Labs is working to produce technology (Caffeine Snorkel) that can study an enterprise's legacy systems and app built on SaaS, create replacement systems and apps, and migrate the data, while supporting key stakeholders through the process over email and chat, with full automation. Thus the legacy systems and SaaS markets shall also be addressed by cloud engines. Zooming out, and reasoning in a more metaphysical way, we believe, as we always have, that there is room for a new kind of cloud created by mathematical networks, that provides seminal advances in the fields of security and resilience, as well as true sovereignty and freedom from lock-in. That this same technology, with the help of additional technologies like orthogonal persistence and Motoko, enables AI to build for us without the need for so much oversight, and to create more backend sophistication while consuming fewer AI API tokens, enables ICP to bring game-changing advances to the world. Cloud engines will work synergistically with the Intelligence Gateway, which will enable apps and services running on engines to seamlessly leverage AI, wherever that AI is running, while providing verifiability at extremely low cost for open weights frontier models. We believe that cloud engines represent an inflection point in the storied history of the Internet Computer project, and I'm very proud to be sharing the details with you on the network's fifth birthday 💪 I'll be back with more news soon!!
19:06

Dear ICP community, the Internet Computer has now been running strong for 5 years 👏👏👏 Here is a celebratory preview of ICP "cloud engines," the sovereign frontier cloud technology the network shall soon provide from Main points: — Cloud engines enable anyone to spin up their own sovereign frontier cloud. The technology involves an extraordinary inventive step, in which cloud is created from a mathematically secure network of nodes. The nodes run as part of the Internet Computer network ( but are selected and configured by the cloud engine's owner. — The frontier cloud provided by engines is strongly focused on enabling AI agents to build and update online applications and services for us. The world is changing fast, and nearly all new online apps and services are already being built with the help of AI, and thus cloud engines target the future of cloud. — Software hosted on cloud engines is tamperproof, which means that it is immune to infrastructure hacks, because it runs inside a mathematically secure network protocol, rather than on computers directly. This means that AI agents, and those building with them, don't need to have a security team in the loop, or to trust someone else's security team. This is crucial, because in the future, non technical people will demand the freedom to build with full automation — where they just need to issue instructions to AI about what to build, and don't need to worry about anything or anyone else. Of course, apps and services running on engines are also vastly safer from the new breed of hacker being enabled by frontier AI. (The cloud engines themselves are also "tamperproof." Even if a hacker gains physical access to some portion of a cloud engine's nodes, and can make arbitrary changes, the computations and data of the hosted apps and services cannot be corrupted or interrupted so long as the network's fault bounds aren't exceeded. The recent hack of Vercel, a major cloud platform, which gave hackers access to the apps it hosted, provides additional perspective on the importance of this advantage.) — Software hosted on cloud engines is guaranteed to run, so long as a sufficient number of the engine's nodes are running. This means that AI can build applications and services without the need to have a human systems admin team constantly tinkering with the underlying platform to keep it running, which is again crucial, because in the future, non technical people will expect the freedom to use AI to build without the support of others. — New frontier programming language technology, in the form of the Motoko language developed by Caffeine Labs, leverages seminal "orthogonal persistence" technology that unifies program logic and data to deliver further unlocks for AI (Motoko is the first computer language being developed that targets agents that are writing software rather than humans engineers per se). Nowadays, AI can build and update production apps at a prodigious rate, even at the speed of conversation. But it can also make mistakes, and there's a risk that an update it creates might be "lossy" in the sense it causes some transformed data to be lost. Again, in this new world, it's both undesirable and impractical for everyone to have to have a systems admin team on-hand to detect lossy updates and roll them back, but Motoko provides a solution: it can detect new software updates are lossy before they are applied, reducing potentially catastrophic errors by AI to harmless coding retries. — Software hosted on cloud engines is "serverless" but unlike traditional serverless software, directly it directly incorporates data through "orthogonal persistence." Another key purpose is simplify backend software logic and fuel the modeling power of AI by increasing abstraction (sorry for the technical language!!!). Put simply, this enables AI to produce more sophisticated backends, faster, and at dramatically lower costs, as measured by the number AI API tokens consumed during coding. (Tip for the technical: orthogonal persistence is a new paradigm where "the program is the database," and data lives inside program variables, which is possible because it's as if hosted software runs forever in persistent memory). — An expanding database of skills at shall make it possible to develop and directly deploy apps and services to your cloud engines directly from Claude Code, Perplexity, Codex and other AI platforms. Further, your account on can be connected, so that new apps and updates created through conversation automatically appear hosted from your cloud engine. In the future, R&D is going to be very seamless. You converse with AI, and your secure and unstoppable apps or services are created or updated. Cloud engines are designed to directly support this "self-writing cloud" future where we can work hands-free. — Tech sovereignty is becoming a huge issue worldwide, with governments and corporations seeking to create sovereign tech stacks owing to geopolitical tensions. Increasingly, people are realizing that tech provided by foreign nations can come with hidden backdoors and kills switches, from the base platform, right up through hosted apps and services. ICP technology is open source, and those building on ICP using AI own their own source code. When you have the source code, you can verify that there are no backdoors, and when you own the source code thanks to AI, you can update it at will, freeing you from vendor lock-in. But cloud engines take sovereignty much further... — You create a cloud engine by selecting the nodes that will be combined. You can choose the class of nodes used, and their number, but more importantly, you can choose who operates the nodes, and where they are located. Almost any configuration is possible, because the Internet Computer scales the security privileges afforded to hosted software within the network according to configuration (software hosted on cloud engines can directly interoperate with software on other engines and traditional subnets, but base restrictions are applied according to security rules). A cloud engine can be created within a region such as Europe, to comply with regs such as GDPR, or completely within a sovereign state like Switzerland or Pakistan. But cloud engines go further still... — Sovereignty is also about freedom from vendor lock-in. Cloud engines are essentially ICP (Internet Computer Protocol) network configurations, and this means the underlying compute nodes they combine can be swapped out without interrupting their hosted apps and services. This is a big deal. In addition, cloud engines now support nodes that are instances running on Big Tech's clouds, in addition to nodes that are dedicated specialized hardware, as per the Gen I and Gen II nodes that dominate the Internet Computer today. For example, it is possible to have an engine running across different AWS data centers, say, and then reconfigure the engine to run across a mixture of AWS, Google, Azure and Hetzner for even more resilience, without the users of hosted apps and services noticing a thing. That's true freedom. — Sovereign AI is becoming increasingly important too, and cloud engines allow special "AI nodes" to be added to them, so that hosted software can perform inference on hardware provisioned by the owner from a location the owner has selected. Even though the AI nodes are only accessible within the cloud engine, they can still benefit from the forthcoming Internet Intelligence Gateway (IG), which will make it possible to validate inference performed on key frontier open weights LLMs, even when the inference is performed on completely independent AI clouds. When the results of inference are received, this technology can verify that neither the prompt+context (input) nor the inference result (output) have been modified, and that the results were produced by the precise LLM expected. This ensures that AI clouds don't cheat by running inference on cheaper models than are being paid for, and bad actors aren't modifying the inputs or outputs to surreptitiously insert advertising into results, say, or change facts, or insert malware when code is being generated. What's super cool about this technology is the cost of the verification is scalable. A very valuable additional security can be achieved with only 1-2% of extra cost. — Scaling apps and services when they hit capacity limits is another thorny problem that cloud engines help the world address. Engines make scaling possible without rewriting or reconfiguring software. The query workload capacity of hosted software can be horizontally scaled simply by adding new nodes to an engine, and nodes can also be added in geographical proximity to demand. Meanwhile, update workload capacity can first be scaled-up by swapping an engine's nodes out for the next class up, and then when no larger class of node is available, horizontally scaled-out by "splitting" the engine into two, which doubles available capacity. (Technical tip: horizontally scaling update capacity by splitting engines requires multi-canister architectures). — For those who have been following how Caffeine builds apps that can efficiently store large numbers of files, I should mention that apps built on cloud engines will also support the new ICP Blob Storage cloud network (since cloud engines currently have up to about 3 TB of memory, which apps storing large amounts of files can easily exceed). We are also working on allowing blob storage nodes to be added to cloud engines, to enable sovereign mass blob storage within an engine, similarly to how AI nodes can be added currently. — Lastly, but certainly not least, I should mention that cloud engines are multi-blockchain capable, and ready for digital assets, thanks to the clever math at their core. For example, an e-commerce service built on a cloud engine can securely accept and custody stablecoin payments, or a multi-chain DEX could be hosted. Further, engines can support software autonomy (software orchestrated and controlled by other autonomous software, in a decentralized way) and can themselves be orchestrated by SNS technology, and thus run autonomously too. Today, though, the focus is on *mainstream* cloud. This year, the cloud industry will generate approximately one trillion dollars in revenue. That number is already huge, but is expected to grow to two trillion dollars by 2030. After years of continuous development, which have seen more than $500m spent on R&D, the Internet Computer network is now tacking directly toward this mainstream cloud market with cloud engine technology. In their first version, cloud engines are not meant to be a cloud panacea. For example, currently they are not ideal for working with big data. You should use something like DataBricks for that. Cloud engines are carefully targeted at enabling AI to produce traditional online applications and services, including SaaS, in a safer and more productive way, which represents a new market segment with tremendous potential. Of course, DFINITY will continue to work relentlessly to push forward ICP's capabilities, so expect further developments. It's worth mentioning that this cloud segment isn't just about creating new apps and services using AI, it's also about replacing legacy systems and apps built on super expensive SaaS services. Caffeine Labs is working to produce technology (Caffeine Snorkel) that can study an enterprise's legacy systems and app built on SaaS, create replacement systems and apps, and migrate the data, while supporting key stakeholders through the process over email and chat, with full automation. Thus the legacy systems and SaaS markets shall also be addressed by cloud engines. Zooming out, and reasoning in a more metaphysical way, we believe, as we always have, that there is room for a new kind of cloud created by mathematical networks, that provides seminal advances in the fields of security and resilience, as well as true sovereignty and freedom from lock-in. That this same technology, with the help of additional technologies like orthogonal persistence and Motoko, enables AI to build for us without the need for so much oversight, and to create more backend sophistication while consuming fewer AI API tokens, enables ICP to bring game-changing advances to the world. Cloud engines will work synergistically with the Intelligence Gateway, which will enable apps and services running on engines to seamlessly leverage AI, wherever that AI is running, while providing verifiability at extremely low cost for open weights frontier models. We believe that cloud engines represent an inflection point in the storied history of the Internet Computer project, and I'm very proud to be sharing the details with you on the network's fifth birthday 💪 I'll be back with more news soon!!

dom | icp

320,618 görüntüleme • 4 ay önce

Rolled credits on Hell is Us after 25 hours! REVIEW: Hell is Us is a rare risk taker in the modern gaming landscape. It foregoes things like quest logs and map markers, instead challenging you to figure things out on your own, without holding your hand. This is a game full of "lightbulb moments." The large, open maps are filled with puzzles, and figuring out these solutions on your own can be very satisfying. Items you collected hours ago could be the key to a solution in a completely different area. It forces you to THINK, and when it finally clicks, it's an extremely rewarding feeling. One of the game's most impressive qualities is the sheer immersion and worldbuilding. Hadea is a war-torn country that is deeply layered with lore. It can at times be a dark story - touching on subjects like prejudice, abuse, and the horrors of war. The country is in an active civil war and Rogue Factor isn't shy about it putting it directly in your face. The two warring factions have commited atrocities against one another, and it's often put front and center. You'll learn the history of the war, as well as the different cultures, religous beliefs, cults, uprisings, and other major events in each area. The visuals, while not the most cutting edge, are still a highlight with strong art design. From thick forests and foggy mists, to dark underground caverns with supernatural phenomena, to wide open lakes and secret underground facilities - none of them felt out of place. Performance was great for me and was much improved from the demo in my experience. Sadly though, combat was a mixed bag. The core mechanics are solid, but the overall execution still needs some work. The parry/block felt inconsistent and visual clarity can be a huge issue. The effects look great but can make it hard to read enemy telegraphs. There's a good amount of build potential however, with dozens of different abilities to choose from for both your weapons and your drone. These cost 'Lymbic Energy' which is recharged by dealing damage. Health and stamina can also be recovered by timing a Ki Pulse ability after every attack. It's a good system that allows you to keep up nonstop pressure once you master it. Difficulty ramps up in Act 2 and beyond, but there were surprisingly few actual boss fights. Enemy variety was also very low with 5 core enemies, and 5 "Hazes." Each can be slightly altered depending on what grade they are (1-3), but in the end it can get quite repetitive. Hell is Us isn't perfect - but it makes some bold design choices that I can appreciate. It's a strong effort from a relatively unknown studio, and there's a LOT of potential for a sequel. The world is rich with lore, and it's already a very fleshed out universe. I completed my playthrough in around 25 hours, but there's easily 30-40 hours worth of content to discover. I've seen some criticism of the story's ending, but I thought it was fine, although somewhat abrupt. It's very clearly teasing a sequel or DLC of some kind. It's well-priced, and ultimately a worthwhile experience IMO! Final Score: 8/10 Playtime: 25 hours Platform: Steam Thanks to Nacon and Dead Good PR for providing a code for review. ❤️
6:39

Rolled credits on Hell is Us after 25 hours! REVIEW: Hell is Us is a rare risk taker in the modern gaming landscape. It foregoes things like quest logs and map markers, instead challenging you to figure things out on your own, without holding your hand. This is a game full of "lightbulb moments." The large, open maps are filled with puzzles, and figuring out these solutions on your own can be very satisfying. Items you collected hours ago could be the key to a solution in a completely different area. It forces you to THINK, and when it finally clicks, it's an extremely rewarding feeling. One of the game's most impressive qualities is the sheer immersion and worldbuilding. Hadea is a war-torn country that is deeply layered with lore. It can at times be a dark story - touching on subjects like prejudice, abuse, and the horrors of war. The country is in an active civil war and Rogue Factor isn't shy about it putting it directly in your face. The two warring factions have commited atrocities against one another, and it's often put front and center. You'll learn the history of the war, as well as the different cultures, religous beliefs, cults, uprisings, and other major events in each area. The visuals, while not the most cutting edge, are still a highlight with strong art design. From thick forests and foggy mists, to dark underground caverns with supernatural phenomena, to wide open lakes and secret underground facilities - none of them felt out of place. Performance was great for me and was much improved from the demo in my experience. Sadly though, combat was a mixed bag. The core mechanics are solid, but the overall execution still needs some work. The parry/block felt inconsistent and visual clarity can be a huge issue. The effects look great but can make it hard to read enemy telegraphs. There's a good amount of build potential however, with dozens of different abilities to choose from for both your weapons and your drone. These cost 'Lymbic Energy' which is recharged by dealing damage. Health and stamina can also be recovered by timing a Ki Pulse ability after every attack. It's a good system that allows you to keep up nonstop pressure once you master it. Difficulty ramps up in Act 2 and beyond, but there were surprisingly few actual boss fights. Enemy variety was also very low with 5 core enemies, and 5 "Hazes." Each can be slightly altered depending on what grade they are (1-3), but in the end it can get quite repetitive. Hell is Us isn't perfect - but it makes some bold design choices that I can appreciate. It's a strong effort from a relatively unknown studio, and there's a LOT of potential for a sequel. The world is rich with lore, and it's already a very fleshed out universe. I completed my playthrough in around 25 hours, but there's easily 30-40 hours worth of content to discover. I've seen some criticism of the story's ending, but I thought it was fine, although somewhat abrupt. It's very clearly teasing a sequel or DLC of some kind. It's well-priced, and ultimately a worthwhile experience IMO! Final Score: 8/10 Playtime: 25 hours Platform: Steam Thanks to Nacon and Dead Good PR for providing a code for review. ❤️

KAMI

445,459 görüntüleme • 1 yıl önce

The circus.. This is how Richard Palmer reported on Harry being DENIED access to the Cenotaph for remembrance: "In a sign of a more conciliatory approach towards his family, Prince Harry has reluctantly accepted that there is no point lobbying to attend Remembrance Sunday at the Cenotaph and other state events - despite his Army record." Hah!😏 The reality is that YES Harry lobbied his way to the Cenotaph, he was simply DENIED, at many levels beyond the palace, so he has to go pull a PR stunt somewhere else. Yet, Palmer presents it as a slight. What could Harry possibly be doing at the Cenotaph?🥴 He is no longer a working royal, he is no longer in the military. In what capacity should he appear in Uniform alongside other senior Royals? "Oh he deployed to Afghanistan" And?🤔 Let's talk about Harry's military record. A military man wears his military record ON his uniform🔥 ALL harry can show for medals on his chest after a 10-year military career, are the 2 basic completion medals given to •ANY serviceman who competed the initial entrance training in the military •and the one presented to ANY serviceman who deployed to Afghanistan, despite Harry NEVER completing One full deployment. That is all Harry has to show for it. 2 basic completion military medals. Yet, Harry claimed to have flown an apache in war and killed 25 talibans, which if true, would have earned him at least 1 or 3 bravery medals in the military. AND YET....😏 Harry does not have ANY special bravery medal, ANY commendation medal for doing anything worth noting in the military. Yes my dears, it is by Harry's medals on his chest, that we know he NEVER killed 25 talibans as he lied about. How could he fly an apache at war when everytime there was an attack on his base, Harry was shielded in a bunker as a prince, while soldiers from "commoner parents" went to fight and die? How could Harry shoot 25 Talibans in an apache from his bunker, except when he was playing "Calls of Duty" in his bunker while the others were fighting for their lives and his? 🤡🎪🤹‍♂️ Harry's "I shot 25 talibans" story, is the same as Meghan's "when I was 11" dish soap story: they NEVER happened in real life🤷🏽‍♀️ The reality of the Cenotaph royal appearances is that, outside of Prince Edward and Princess Anne who never served in the Military and hold ceremonial roles as senior royals, King Charles served in the military and is the commander in chief; of course he deserves to be there. Prince William has 7.5 years of active military service and is the Heir. Of course he deserves to be there👌🏽 Furthermore, unlike Harry, William actually saw first-hand danger in his military career as a RAF SAR in RAF valley, flying on the coast and in the mountains of Wales to rescue people, in storms and very poor weather condition🔥 Unlike Harry, when it was time to face danger and save the lives of British citizens, William did not hide in a bunker; he was sitting first row, piloting his helicopter with his team to rescue everyday people who have spoken on record and thanked him for his bravery🔥 Unlike Harry who is LYING about killing 25 talibans to puff his noneventful military record, William is CREDITED on record for saving 149 lives while taking part in a 156 rescue missions.Yes unlike Harry, William's military record, is NOT fiction that cannot be substantiated by facts. William's military reecord is factual reality🔥 In 2026, during a visit to RAF valley for its 85th anniversary, Prince William reflected on his part service as rescue pilot and when asked by trainees if he would return to the role if he could, he simply replied "Back in a heartbeat"❤️ Therefore, if someone deserves to be standing at that Cenotaph for his military record, as Palmer wants to imply, then Prince William definitely deserve to be there; not only for his position as Heir to the throne and future commander in chief but ON MERIT for his military record as well. He did his duty and he did it well and saved lives❤️ The real problem the Royal rota will always have with their attempt at diminishing William to to make Harry and his PR manufactured life shine, are that facts and history will always get in the way☕️ #PrinceofWales
0:24

The circus.. This is how Richard Palmer reported on Harry being DENIED access to the Cenotaph for remembrance: "In a sign of a more conciliatory approach towards his family, Prince Harry has reluctantly accepted that there is no point lobbying to attend Remembrance Sunday at the Cenotaph and other state events - despite his Army record." Hah!😏 The reality is that YES Harry lobbied his way to the Cenotaph, he was simply DENIED, at many levels beyond the palace, so he has to go pull a PR stunt somewhere else. Yet, Palmer presents it as a slight. What could Harry possibly be doing at the Cenotaph?🥴 He is no longer a working royal, he is no longer in the military. In what capacity should he appear in Uniform alongside other senior Royals? "Oh he deployed to Afghanistan" And?🤔 Let's talk about Harry's military record. A military man wears his military record ON his uniform🔥 ALL harry can show for medals on his chest after a 10-year military career, are the 2 basic completion medals given to •ANY serviceman who competed the initial entrance training in the military •and the one presented to ANY serviceman who deployed to Afghanistan, despite Harry NEVER completing One full deployment. That is all Harry has to show for it. 2 basic completion military medals. Yet, Harry claimed to have flown an apache in war and killed 25 talibans, which if true, would have earned him at least 1 or 3 bravery medals in the military. AND YET....😏 Harry does not have ANY special bravery medal, ANY commendation medal for doing anything worth noting in the military. Yes my dears, it is by Harry's medals on his chest, that we know he NEVER killed 25 talibans as he lied about. How could he fly an apache at war when everytime there was an attack on his base, Harry was shielded in a bunker as a prince, while soldiers from "commoner parents" went to fight and die? How could Harry shoot 25 Talibans in an apache from his bunker, except when he was playing "Calls of Duty" in his bunker while the others were fighting for their lives and his? 🤡🎪🤹‍♂️ Harry's "I shot 25 talibans" story, is the same as Meghan's "when I was 11" dish soap story: they NEVER happened in real life🤷🏽‍♀️ The reality of the Cenotaph royal appearances is that, outside of Prince Edward and Princess Anne who never served in the Military and hold ceremonial roles as senior royals, King Charles served in the military and is the commander in chief; of course he deserves to be there. Prince William has 7.5 years of active military service and is the Heir. Of course he deserves to be there👌🏽 Furthermore, unlike Harry, William actually saw first-hand danger in his military career as a RAF SAR in RAF valley, flying on the coast and in the mountains of Wales to rescue people, in storms and very poor weather condition🔥 Unlike Harry, when it was time to face danger and save the lives of British citizens, William did not hide in a bunker; he was sitting first row, piloting his helicopter with his team to rescue everyday people who have spoken on record and thanked him for his bravery🔥 Unlike Harry who is LYING about killing 25 talibans to puff his noneventful military record, William is CREDITED on record for saving 149 lives while taking part in a 156 rescue missions.Yes unlike Harry, William's military record, is NOT fiction that cannot be substantiated by facts. William's military reecord is factual reality🔥 In 2026, during a visit to RAF valley for its 85th anniversary, Prince William reflected on his part service as rescue pilot and when asked by trainees if he would return to the role if he could, he simply replied "Back in a heartbeat"❤️ Therefore, if someone deserves to be standing at that Cenotaph for his military record, as Palmer wants to imply, then Prince William definitely deserve to be there; not only for his position as Heir to the throne and future commander in chief but ON MERIT for his military record as well. He did his duty and he did it well and saved lives❤️ The real problem the Royal rota will always have with their attempt at diminishing William to to make Harry and his PR manufactured life shine, are that facts and history will always get in the way☕️ #PrinceofWales

Canellecitadelle

119,119 görüntüleme • 26 gün önce

🚨The DuPont Dynasty's Hidden Hand: Pierre du Pont V, Netanyahu's Rejected Deal, and the Assassination of Charlie Kirk – A Clear Chain of Elite Influence🇫🇷🇮🇱 In an era where the lines between corporate power, foreign policy, and domestic conservatism have blurred into a web of unaccountable influence, few stories cut as deep as the one unfolding around Turning Point USA (TPUSA) and the September 10, 2025, assassination of its founder, Charlie Kirk. On her December 3, 2025, podcast, Candace Owens pulled back the curtain on a revelation that demands scrutiny: a direct financial and geopolitical pipeline from the storied DuPont family – heirs to a French-American industrial empire – to TPUSA, brokered through Israeli Prime Minister Benjamin Netanyahu. Kirk's rejection of this "life-changing" deal, insiders claim, may have marked him as a liability to the very elites he once courted. This is not idle conspiracy; it's a documented trail of donor ties, leaked communications, and post-assassination maneuvers that the FBI has conspicuously ignored. As citizen investigators, we must follow the money – and the motives – wherever they lead. The DuPont Legacy: From French Revolution Refugees to American Power Brokers To understand the stakes, we must first trace the DuPont roots, a saga of transatlantic ambition that began amid the chaos of revolutionary France. In 1802, Pierre Samuel du Pont de Nemours – a physiocrat economist, minor aristocrat, and advisor to figures like Thomas Jefferson – fled Paris after the Reign of Terror. With seed capital from French networks (including ties to the Lavoisier family, infamous for their role in gunpowder production), he established E.I. du Pont de Nemours and Company in Wilmington, Delaware. What started as a modest gunpowder mill exploded into an industrial colossus, supplying the U.S. military during the War of 1812, the Civil War, and both World Wars. By the 20th century, under leaders like Pierre S. du Pont (1870–1954), the family diversified into chemicals, textiles, and composites. DuPont pioneered nylon, Teflon, and Kevlar – materials integral to modern warfare, from fighter jet components to body armor. Yet this innovation masked darker chapters: During World War II, DuPont subsidiaries supplied chemicals and explosives to Nazi Germany through neutral intermediaries, a fact buried in postwar restructurings. The family's mastery of corporate alchemy – spinning off liabilities like Chemours in 2015 to offload $671 million in PFAS "forever chemicals" lawsuits – exemplifies elite impunity. Today, the dynasty's fortune exceeds $14 billion, spread across 3,500 heirs, with Wilmington as its fortified hub: Delaware's lax corporate laws shelter 60% of Fortune 500 entities, making it a haven for anonymous LLCs and offshore flows. Enter Pierre Samuel du Pont V (born circa 1959), the reclusive heir often described as the "low-profile operator" of this empire. A French-American by blood – fluent in the language, raised partly in Europe, and tied to the family's Parisian estates – Pierre V embodies the dynasty's dual loyalties. Unlike flashier relatives like Pierre S. du Pont IV (Delaware governor, 1977–1985), Pierre V shuns the spotlight, channeling influence through think tanks, foundations, and discreet philanthropy. His role? Stewarding the family's post-2019 merger with Dow Chemical, where DuPont de Nemours emerged leaner, greener (on paper), and more entangled in global supply chains – including military tech exported to Israel. Wilmington isn't just home; it's a nexus of power, blocks from FBI field offices and federal courts, where elite disputes vanish into sealed filings. The Netanyahu Brokerage: A "Life-Changing" Deal with Strings Attached The DuPont thread weaves into TPUSA through a 2025 overture from Benjamin Netanyahu, whose administration has long mastered the art of "influencer diplomacy." Reports confirm Bibi's office allocated up to $7,000 per post for U.S. conservative voices to amplify pro-Israel narratives, a $45 million Google contract in one case. In early 2025, Netanyahu extended a "huge" funding offer to TPUSA – whispers of $150 million or more – aimed at scaling its campus operations amid rising anti-Zionist sentiment among Gen Z. Kirk's September 2025 letter to Bibi urged countering "pro-Palestinian narratives" via tours and social campaigns, but privately, Kirk chafed under donor pressure from figures like Robert Shillman (Jewish backers who pulled millions over Tucker Carlson invites). Owens' insiders – six TPUSA staffers – paint Pierre du Pont V as the "whale": a "life-changing" tech IPO infusion, potentially hundreds of millions, funneled through DuPont channels for AI surveillance, donor analytics, and event drones. Not mere philanthropy: strings included locking in pro-Israel loyalty – more "security" tech on campuses, fewer platforms for critics like Owens or Carlson. DuPont's military-grade composites (e.g., Kevlar for IDF gear) made Pierre V a natural fit for Netanyahu's orbit, where family foundations overlap with Chabad and AIPAC networks. Kirk demurred. Leaked texts (verified by TPUSA's Andrew Kolvet) reveal his fury: losing $2 million annually for resisting donor "stereotypes," rejecting Bibi's Israel trip, and probing IRS 990s for ghost employees and missing funds. A Hamptons "intervention" with Bill Ackman – elite pressure tactics – followed. Kirk eyed a DOGE-style audit, pivoting TPUSA toward transparency over transatlantic cash. To the DuPont-Netanyahu axis, he was no longer an asset – he was a threat. The Assassination and the Post-Hit Pivot: Damage Control or Consolidation? September 10, 2025: Kirk is assassinated at a Utah Valley University event. Hours later, TPUSA COO Justin Streiff allegedly dials Pierre du Pont V – "immediate," per Owens' sources. Damage control? Or securing the billions now that the "roadblock" is gone? TPUSA has since accepted DuPont-linked funds; 2026 filings will likely confirm it. Kirk's pivot threatened billions: rejecting Netanyahu meant cratering elite support from Chabad donors (Leviev, Adelson) to AIPAC influencers. Owens catalogs 10 post-hit "lies" – scrubbed finances, silenced insiders – echoing historical hits: JFK's donor clashes, Rabin's Oslo backlash, Trump's Epstein pivot. The French Connection: Macron, Wilmington, and Transatlantic Shadows Pierre V's dual heritage – French by blood, American by empire – bridges to Emmanuel Macron's circle. The du Ponts never severed Parisian ties: French properties, banks, and aristocratic networks persist. Brigitte Macron's July 2025 defamation suit against Owens? Filed in Wilmington Superior Court – a 218-page behemoth seeking punitive damages over "verifiably false" transition claims. Why Delaware? Secrecy laws, proximity to DuPont HQ, FBI offices, and federal courts – all within blocks. Macron's lawyers, per Owens, operate from the same firms shielding du Pont trusts. Egyptian-registered planes tailed Erica Kirk for years (confirmed by Tucker Carlson). Routed through Delaware pre-hit, they align with U.S.-Egypt-Israel drills. DuPont's WWII Nazi supplies echo Bolshevik-era alliances; today, it's Macron's Gendarmerie (elite ops) training with U.S.-Israel units. Wilmington is the nexus: DuPont nerve center, FBI hub, Macron's legal beachhead – a protection racket since 1802. Motive, Means, and the Wilmington Web: Why Kirk Had to Go Greed and control: The DuPont-IPO was billions for TPUSA's ascent, but Kirk's "no" exposed rot – ghost payrolls, donor hijacks. Post-hit call to Pierre V? Locking in the windfall. TPUSA's $140M post-assassination haul? Elite cash, funneled quietly. FBI opacity? Their Wilmington office sits amid it all. Timeline: Jan 2025 – Netanyahu pitches. Summer – Kirk's donor fury texts. Aug – Finance probes. Sep 10 – Hit. Hours later – Pierre V call. Oct – Owens exposes texts. Dec 3 – Insider bombshell. Sequential elimination. Call to Action: Demand the Audits, FOIAs, and Truth Kirk's death wasn't "lone gunman" chaos; it was to safeguard billions and agendas. TPUSA's silence? Complicity. Owens invokes spiritual warfare: Kirk as martyr. Demand IRS 990 audits, FOIAs on DuPont donations, Streiff's call logs. Cross-reference Chabad overlaps. Protect Owens' insiders. Share, scrutinize, amplify. This exposes how foreign influence colonizes U.S. conservatism, assassinating dissent. The narrative crumbles; the hidden hand trembles. The people will prevail. Special Thanks to Lance Henderson who compiled all this information for me. Go FOLLOW Them. To Sources are in the comments below...
1:14

🚨The DuPont Dynasty's Hidden Hand: Pierre du Pont V, Netanyahu's Rejected Deal, and the Assassination of Charlie Kirk – A Clear Chain of Elite Influence🇫🇷🇮🇱 In an era where the lines between corporate power, foreign policy, and domestic conservatism have blurred into a web of unaccountable influence, few stories cut as deep as the one unfolding around Turning Point USA (TPUSA) and the September 10, 2025, assassination of its founder, Charlie Kirk. On her December 3, 2025, podcast, Candace Owens pulled back the curtain on a revelation that demands scrutiny: a direct financial and geopolitical pipeline from the storied DuPont family – heirs to a French-American industrial empire – to TPUSA, brokered through Israeli Prime Minister Benjamin Netanyahu. Kirk's rejection of this "life-changing" deal, insiders claim, may have marked him as a liability to the very elites he once courted. This is not idle conspiracy; it's a documented trail of donor ties, leaked communications, and post-assassination maneuvers that the FBI has conspicuously ignored. As citizen investigators, we must follow the money – and the motives – wherever they lead. The DuPont Legacy: From French Revolution Refugees to American Power Brokers To understand the stakes, we must first trace the DuPont roots, a saga of transatlantic ambition that began amid the chaos of revolutionary France. In 1802, Pierre Samuel du Pont de Nemours – a physiocrat economist, minor aristocrat, and advisor to figures like Thomas Jefferson – fled Paris after the Reign of Terror. With seed capital from French networks (including ties to the Lavoisier family, infamous for their role in gunpowder production), he established E.I. du Pont de Nemours and Company in Wilmington, Delaware. What started as a modest gunpowder mill exploded into an industrial colossus, supplying the U.S. military during the War of 1812, the Civil War, and both World Wars. By the 20th century, under leaders like Pierre S. du Pont (1870–1954), the family diversified into chemicals, textiles, and composites. DuPont pioneered nylon, Teflon, and Kevlar – materials integral to modern warfare, from fighter jet components to body armor. Yet this innovation masked darker chapters: During World War II, DuPont subsidiaries supplied chemicals and explosives to Nazi Germany through neutral intermediaries, a fact buried in postwar restructurings. The family's mastery of corporate alchemy – spinning off liabilities like Chemours in 2015 to offload $671 million in PFAS "forever chemicals" lawsuits – exemplifies elite impunity. Today, the dynasty's fortune exceeds $14 billion, spread across 3,500 heirs, with Wilmington as its fortified hub: Delaware's lax corporate laws shelter 60% of Fortune 500 entities, making it a haven for anonymous LLCs and offshore flows. Enter Pierre Samuel du Pont V (born circa 1959), the reclusive heir often described as the "low-profile operator" of this empire. A French-American by blood – fluent in the language, raised partly in Europe, and tied to the family's Parisian estates – Pierre V embodies the dynasty's dual loyalties. Unlike flashier relatives like Pierre S. du Pont IV (Delaware governor, 1977–1985), Pierre V shuns the spotlight, channeling influence through think tanks, foundations, and discreet philanthropy. His role? Stewarding the family's post-2019 merger with Dow Chemical, where DuPont de Nemours emerged leaner, greener (on paper), and more entangled in global supply chains – including military tech exported to Israel. Wilmington isn't just home; it's a nexus of power, blocks from FBI field offices and federal courts, where elite disputes vanish into sealed filings. The Netanyahu Brokerage: A "Life-Changing" Deal with Strings Attached The DuPont thread weaves into TPUSA through a 2025 overture from Benjamin Netanyahu, whose administration has long mastered the art of "influencer diplomacy." Reports confirm Bibi's office allocated up to $7,000 per post for U.S. conservative voices to amplify pro-Israel narratives, a $45 million Google contract in one case. In early 2025, Netanyahu extended a "huge" funding offer to TPUSA – whispers of $150 million or more – aimed at scaling its campus operations amid rising anti-Zionist sentiment among Gen Z. Kirk's September 2025 letter to Bibi urged countering "pro-Palestinian narratives" via tours and social campaigns, but privately, Kirk chafed under donor pressure from figures like Robert Shillman (Jewish backers who pulled millions over Tucker Carlson invites). Owens' insiders – six TPUSA staffers – paint Pierre du Pont V as the "whale": a "life-changing" tech IPO infusion, potentially hundreds of millions, funneled through DuPont channels for AI surveillance, donor analytics, and event drones. Not mere philanthropy: strings included locking in pro-Israel loyalty – more "security" tech on campuses, fewer platforms for critics like Owens or Carlson. DuPont's military-grade composites (e.g., Kevlar for IDF gear) made Pierre V a natural fit for Netanyahu's orbit, where family foundations overlap with Chabad and AIPAC networks. Kirk demurred. Leaked texts (verified by TPUSA's Andrew Kolvet) reveal his fury: losing $2 million annually for resisting donor "stereotypes," rejecting Bibi's Israel trip, and probing IRS 990s for ghost employees and missing funds. A Hamptons "intervention" with Bill Ackman – elite pressure tactics – followed. Kirk eyed a DOGE-style audit, pivoting TPUSA toward transparency over transatlantic cash. To the DuPont-Netanyahu axis, he was no longer an asset – he was a threat. The Assassination and the Post-Hit Pivot: Damage Control or Consolidation? September 10, 2025: Kirk is assassinated at a Utah Valley University event. Hours later, TPUSA COO Justin Streiff allegedly dials Pierre du Pont V – "immediate," per Owens' sources. Damage control? Or securing the billions now that the "roadblock" is gone? TPUSA has since accepted DuPont-linked funds; 2026 filings will likely confirm it. Kirk's pivot threatened billions: rejecting Netanyahu meant cratering elite support from Chabad donors (Leviev, Adelson) to AIPAC influencers. Owens catalogs 10 post-hit "lies" – scrubbed finances, silenced insiders – echoing historical hits: JFK's donor clashes, Rabin's Oslo backlash, Trump's Epstein pivot. The French Connection: Macron, Wilmington, and Transatlantic Shadows Pierre V's dual heritage – French by blood, American by empire – bridges to Emmanuel Macron's circle. The du Ponts never severed Parisian ties: French properties, banks, and aristocratic networks persist. Brigitte Macron's July 2025 defamation suit against Owens? Filed in Wilmington Superior Court – a 218-page behemoth seeking punitive damages over "verifiably false" transition claims. Why Delaware? Secrecy laws, proximity to DuPont HQ, FBI offices, and federal courts – all within blocks. Macron's lawyers, per Owens, operate from the same firms shielding du Pont trusts. Egyptian-registered planes tailed Erica Kirk for years (confirmed by Tucker Carlson). Routed through Delaware pre-hit, they align with U.S.-Egypt-Israel drills. DuPont's WWII Nazi supplies echo Bolshevik-era alliances; today, it's Macron's Gendarmerie (elite ops) training with U.S.-Israel units. Wilmington is the nexus: DuPont nerve center, FBI hub, Macron's legal beachhead – a protection racket since 1802. Motive, Means, and the Wilmington Web: Why Kirk Had to Go Greed and control: The DuPont-IPO was billions for TPUSA's ascent, but Kirk's "no" exposed rot – ghost payrolls, donor hijacks. Post-hit call to Pierre V? Locking in the windfall. TPUSA's $140M post-assassination haul? Elite cash, funneled quietly. FBI opacity? Their Wilmington office sits amid it all. Timeline: Jan 2025 – Netanyahu pitches. Summer – Kirk's donor fury texts. Aug – Finance probes. Sep 10 – Hit. Hours later – Pierre V call. Oct – Owens exposes texts. Dec 3 – Insider bombshell. Sequential elimination. Call to Action: Demand the Audits, FOIAs, and Truth Kirk's death wasn't "lone gunman" chaos; it was to safeguard billions and agendas. TPUSA's silence? Complicity. Owens invokes spiritual warfare: Kirk as martyr. Demand IRS 990 audits, FOIAs on DuPont donations, Streiff's call logs. Cross-reference Chabad overlaps. Protect Owens' insiders. Share, scrutinize, amplify. This exposes how foreign influence colonizes U.S. conservatism, assassinating dissent. The narrative crumbles; the hidden hand trembles. The people will prevail. Special Thanks to Lance Henderson who compiled all this information for me. Go FOLLOW Them. To Sources are in the comments below...

Project Constitution

86,069 görüntüleme • 9 ay önce

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.
10:20

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.

TheValueist

102,145 görüntüleme • 9 ay önce

Today my power wheelchair hit 15,000 km. Almost every one of those kilometres had a Rottweiler beside it. Losing my legs meant relearning everything. And it was and still is, hard. Opening a door and getting through it. Getting out of bed at 3 a.m. to pee. Even getting into bed. I am eternally grateful a Rottweiler was there for all of it. Kuno, a huge dog with a massive heart, taught me how to live this life. Without him I am fairly certain I would have stayed in bed, stared at the TV, and hated the world. Instead he got me outside, and we learned how to navigate a community built when people like me were expected to live in institutions. Those kilometres were full of firsts. First grocery shop with no help. First time stuck in a couple inches of snow. First time I could not get into a business because they chose an inaccessible location. First time I trained a puppy from a wheelchair. I was painfully self-conscious those first years. The pity smiles. The awkward questions. The stranger in the grocery store who asked to lay hands on me and pray. I coped by getting better at dog training. Kuno and I put a ridiculous number of miles on together. We took classes and found mentors. I learned to drive with hand controls, then bought a used accessible van so we could travel. This wheelchair has rolled all over the prairies when we've completed in Rally Obedience trials. We played in parks in the sun. Trained in the underground parking when it snowed. We gave presentations on service dogs, accessibility, and disability at community groups, schools, and post-secondary institutions. We challenged the city on multiple issues and became the unofficial accessibility watchdog team. At Christmas we raised funds for charities. More importantly, except for a couple of brutal weather days when I had to call emergency dog-walking help, we were out in the community being active. A lot of those kilometres were spent training very specific work: backing in a perfect heel, spins, pivoting beside the chair, even pivots with the dog in front facing me. We got so good it almost became a dance. He walked beside me the first time I rolled down the sidewalk in this chair. I rolled beside him on the last steps he took before he passed away. A few years ago Chesnyy joined the household and the kilometres started adding up faster. We started working with a government-qualified service dog organization and now we teach some of their classes. Last year Chesnyy and I moved into an accessible house with my best friend, and another Rottweiler puppy arrived. There are still a lot of walks left. Winter is still highly inaccessible. After 15,000 km the challenges have shifted. Many barriers are still there and we will keep tackling them. I am more confident now. It took a lot of work, but I have proven myself in the dog training community. I still miss things I can no longer do, but I have accepted this disability and I am trying to live my best life. The bigger fight now is pain. Chronic pain is not new. It was an issue before amputation. Phantom pain no longer surprises me. My shoulders hurts tremendously from transferring and lifting from a seated position. My back aches from sitting. And yes, my butt hurts. Pain makes motivation really hard. The dogs help with that. On the days when I just want to give up, wet noses nudge me on. 15,000 km is a lot for a power chair. I got this one in the fall of 2020. Techs are always shocked at the mileage. They typically see about 5,000 km over the life of a chair. This one will probably need replacing in the next year or so. Thankfully there is some government funding but they only fund base models. Good chairs that are easier on the body and actually work well cost a lot more. A full-time power wheelchair in Canada often starts around $15,000. Add tilt to help prevent pressure sores and you can easily add another $10,000. When you live in the chair it has to be customized. One of the hardest parts of disability is the cost. I am fortunate to have a consistent disability income that covers monthly bills but I don't have extended medical insurance to get upgraded equipment. Prescriptions, physical therapy, and a lot of disability related equipment is all out of pocket. I am not asking for money. I just want people to understand that power mobility is ridiculously expensive, and that a couple of giant-hearted Rottweilers made 15,000 km possible. Today I took the chair off the pavement and did a little 4x4ing so the dogs could run off-leash on a beautiful summer day. Watching them play, I thought about every hard kilometre that got us here. That still makes my heart smile. ❤️🐾❤️ Admin
0:20

Today my power wheelchair hit 15,000 km. Almost every one of those kilometres had a Rottweiler beside it. Losing my legs meant relearning everything. And it was and still is, hard. Opening a door and getting through it. Getting out of bed at 3 a.m. to pee. Even getting into bed. I am eternally grateful a Rottweiler was there for all of it. Kuno, a huge dog with a massive heart, taught me how to live this life. Without him I am fairly certain I would have stayed in bed, stared at the TV, and hated the world. Instead he got me outside, and we learned how to navigate a community built when people like me were expected to live in institutions. Those kilometres were full of firsts. First grocery shop with no help. First time stuck in a couple inches of snow. First time I could not get into a business because they chose an inaccessible location. First time I trained a puppy from a wheelchair. I was painfully self-conscious those first years. The pity smiles. The awkward questions. The stranger in the grocery store who asked to lay hands on me and pray. I coped by getting better at dog training. Kuno and I put a ridiculous number of miles on together. We took classes and found mentors. I learned to drive with hand controls, then bought a used accessible van so we could travel. This wheelchair has rolled all over the prairies when we've completed in Rally Obedience trials. We played in parks in the sun. Trained in the underground parking when it snowed. We gave presentations on service dogs, accessibility, and disability at community groups, schools, and post-secondary institutions. We challenged the city on multiple issues and became the unofficial accessibility watchdog team. At Christmas we raised funds for charities. More importantly, except for a couple of brutal weather days when I had to call emergency dog-walking help, we were out in the community being active. A lot of those kilometres were spent training very specific work: backing in a perfect heel, spins, pivoting beside the chair, even pivots with the dog in front facing me. We got so good it almost became a dance. He walked beside me the first time I rolled down the sidewalk in this chair. I rolled beside him on the last steps he took before he passed away. A few years ago Chesnyy joined the household and the kilometres started adding up faster. We started working with a government-qualified service dog organization and now we teach some of their classes. Last year Chesnyy and I moved into an accessible house with my best friend, and another Rottweiler puppy arrived. There are still a lot of walks left. Winter is still highly inaccessible. After 15,000 km the challenges have shifted. Many barriers are still there and we will keep tackling them. I am more confident now. It took a lot of work, but I have proven myself in the dog training community. I still miss things I can no longer do, but I have accepted this disability and I am trying to live my best life. The bigger fight now is pain. Chronic pain is not new. It was an issue before amputation. Phantom pain no longer surprises me. My shoulders hurts tremendously from transferring and lifting from a seated position. My back aches from sitting. And yes, my butt hurts. Pain makes motivation really hard. The dogs help with that. On the days when I just want to give up, wet noses nudge me on. 15,000 km is a lot for a power chair. I got this one in the fall of 2020. Techs are always shocked at the mileage. They typically see about 5,000 km over the life of a chair. This one will probably need replacing in the next year or so. Thankfully there is some government funding but they only fund base models. Good chairs that are easier on the body and actually work well cost a lot more. A full-time power wheelchair in Canada often starts around $15,000. Add tilt to help prevent pressure sores and you can easily add another $10,000. When you live in the chair it has to be customized. One of the hardest parts of disability is the cost. I am fortunate to have a consistent disability income that covers monthly bills but I don't have extended medical insurance to get upgraded equipment. Prescriptions, physical therapy, and a lot of disability related equipment is all out of pocket. I am not asking for money. I just want people to understand that power mobility is ridiculously expensive, and that a couple of giant-hearted Rottweilers made 15,000 km possible. Today I took the chair off the pavement and did a little 4x4ing so the dogs could run off-leash on a beautiful summer day. Watching them play, I thought about every hard kilometre that got us here. That still makes my heart smile. ❤️🐾❤️ Admin

Team Servicerottie🇨🇦🐕‍🦺🦽

21,919 görüntüleme • 27 gün önce

The July 4th weekend All-In The All-In Podcast turned into a long argument about who owns the intelligence layer. The besties think enterprises just woke up to a trap they had been walking into, here's how the conversation went (save this): ◽️ The Palantir-Nvidia deal is a bet against the model-layer duopoly. Palantir will use Nvidia's Nemotron open models to build a custom frontier-quality model for US government agencies, and the agencies own the hardware, the data, and the weights. Sacks framed it as structural: an application company and a chip company both want a competitive model layer, so they are natural partners against a two-provider middle. ◽️ Alex Karp's CNBC "crashout" was actually the thesis. Karp argued enterprises have lost trust in the frontier labs and want to own their compute, models, data, and alpha. Sacks translated it as a new definition of enterprise AI safety: safety means the model provider cannot hoover up your proprietary knowledge and turn it into its next product. ◽️ Figma is the cautionary tale that made it real. Anthropic launched Claude Design into Figma's category, its chief product officer sat on Figma's board and resigned only 3 days before launch, and Figma's stock is down about 50% this year while Anthropic's valuation surged. Sacks listed Claude Science, Security, Legal, Financial, and Code as the same move: dominate the model layer, then take the lucrative verticals. ◽️ The playbook has a name, and it is Microsoft and Google. Sacks argued Anthropic is running the operating-system strategy: own the layer everyone builds on, then walk up the stack. His Google receipt is that fewer than half of searches now send you off-site, versus an early Google that prided itself on how fast it kicked you away. ◽️ The BCG number is what raises the stakes. Chamath cited a BCG return-on-capital-employed study: the cost of capital is back to its long-run 8 to 11%, and half of large US companies cannot earn returns above it. If you are already teetering on your cost of capital, handing your alpha to a provider that may compete with you is not a luxury risk, it is fatal. ◽️ The 16.4x number is the whole argument in one data point. Chamath ran a code-migration task through 8090's harness. Wrapping Claude was 1.4x cheaper and 1.5x faster than Claude Opus alone. Wrapping the best open-source model was 16.4x cheaper, at about 3x slower. For a background task, three extra hours to cut cost by 16x is not a close call. ◽️ Even at 100x cheaper, enterprises were saying no for the wrong reason. Chamath relayed an ex-Meta PM's point that companies reject open models over China and safety fears, when they could host those same open weights on their own GPUs in US data centers with nothing flowing back. The safety objection, she argued, is backwards: the leak is the data you hand the frontier labs. ◽️ Friedberg says the frontier labs are trying to commoditize their own customers. Anthropic has been signing up life-sciences companies to feed a new life-focused model in exchange for early access, and nearly everyone he has talked to now refuses, recognizing that data they spent billions generating becomes worthless once it is pooled with everyone else's. ◽️ The deployment topology is shifting from big hubs to distributed spokes. Friedberg's map: the old assumption was a few capital-advantaged mega-clusters plus inference clouds. The new one is large hubs, medium hubs (enterprise training clusters), and distributed spokes, including on-prem inference in your own building. Owning your weights is the point. ◽️ Chamath's endgame is running GLM himself. An industry contact told him that with harness post-training and telemetry, an open Chinese model like GLM could get as good as Anthropic's Mythos. His conclusion: take GLM, control it soup-to-nuts on US hardware with only US citizens touching it, and pay a fraction. ◽️ The Apple analogy sharpens why renting intelligence is different from renting distribution. Chamath argued Apple is the only platform that respected developers, deliberately keeping its stock apps basic to protect the ecosystem and collect its 30% tax. There is no 30% tax on open models, and worse, you cannot rent intelligence from the same place that rents it to your competitor without ending up identical to them. ◽️ Nvidia's open model is now good enough to matter. Calacanis claimed you cannot tell Jensen Huang's Nemotron from Claude on 95% of searches, and that Nvidia downplayed the model until now to avoid alarming its top customers. The gloves came off once OpenAI, Anthropic, and Elon all signaled their own silicon ambitions. ◽️ Sacks sized the duopoly: roughly $60B and $40B in ARR. Anthropic is around ~$60 billion of ARR, OpenAI at ~$40 billion, and no one else generates meaningful model-layer revenue. Sacks's policy line: the US does not ban monopolies, only anti-competitive tactics, but the government should do nothing to make the duopoly more likely. ◽️ The token deflation call: 90% a year for three years. Calacanis predicted token costs fall 90% annually for three years, putting the price of intelligence near free and making it rational to waste tokens on hardware you already own. Friedberg's version is a 70/20/10 split between big cloud, local, and other clouds. ◽️ A wave of platform lock-in spending is already landing. Calacanis flagged Microsoft standing up a roughly $2.5 billion forward-deployed-engineer effort and Amazon spending about $1 billion on the same, plus OpenAI's version. His read: enterprises will slam the door, because letting a provider's engineers study your business is how it ends up in their model. ◽️ The server-per-employee prediction. Calacanis expects every employee to get $10,000 to $20,000 of local compute, a Mac Studio or a high-RAM Dell, running a personal local model that syncs to a thin laptop. A server per person, so nothing leaks. ◽️ On jobs, the data does not show present-tense loss. Sacks cited a RAMP and Revelio Labs study of over 21,000 US firms: the heaviest AI spenders grew headcount about 10% over two years, and entry-level headcount grew even faster at 12%. Friedberg's harder claim: there is no AI job loss yet, only clunky, gradual value creation, and the media will not reverse its narrative because that destroys its credibility. ◽️ The displacement case is real but forward-dated. The counterpoint on the show was that customer support, entry-level data entry and BPO, and driving are the near-term displacements, with Waymo cited as present-tense evidence: in markets where it hits critical mass, Uber and Lyft stop recruiting drivers. Sacks noted most US entry-level support was already offshored, so the acute risk sits in those countries first. ◽️ The human-premium counternarrative. Friedberg argued that as automation spreads, human interaction gets a premium: the skilled bartender, the real driver, the human-in-the-loop tier. He cited the company (referenced as Klarna) that hyped replacing its whole support team with AI, then reversed a year later on brand grounds. ◽️ The export-control episode needed three conditions, and Sacks says do not over-read it. Commerce lifted controls on Anthropic's Fable 5 after two weeks, with Mythos 5 restored to US customers around June 26 once co-founder Tom Brown replaced Dario as lead negotiator. Sacks's three conditions: Dario boasting for months about a cyber weapon, Amazon reporting failed guardrails in testing, and Dario refusing to roll Fable back. His message to allies: this was a particular set of circumstances rather than the debut of a standing lever. ◽️ The import question nobody answered cleanly. Calacanis pressed on why the US blocks Chinese cars and drones but not Chinese open models like DeepSeek and Kimi. Sacks's answer: a forked open model run on US hardware stops being Chinese, and banning open source would isolate the US and impose a token tax on American enterprises, so let the market decide if American open models win. ◽️ The California fiscal story is a business-climate story. Friedberg walked through the numbers behind Newsom's "balanced" $351B budget: expenses exceed revenue and $20-40B is borrowed to close the gap, the budget grew 65% in six years ($215B to $355B), personal income tax is $142B of ~$211B revenue with the top 1% (150,000 people) paying $70B of it, and the corporate rate of 8.9% sits far above Texas at zero. ◽️ The tax base is leaving, and the state is now taxing everyone else. Friedberg cited 1 to 1.5% of adjusted gross income leaving each year (about 15% over a decade), at least 15 Fortune 500 HQs and ~2,100 firms gone since 2019, and a new 8% software sales tax hitting Word, Gmail, and ChatGPT subscriptions plus a health-insurance tax, on top of a now-permanent 14.4% top bracket. The liabilities behind it run $1.4T in debt, up to $1.5T in unfunded pensions senior to state bonds, and ~$40B/year in out-year deficits. Lastly, the line that framed the whole show: "You can't rent intelligence from the same place that rents it to your competitor." That is the sovereignty thesis in one sentence, and every number in this episode is an argument for it. ____ Follow Fireside Alpha for more summaries on key business and technology conversations.
1:33

The July 4th weekend All-In The All-In Podcast turned into a long argument about who owns the intelligence layer. The besties think enterprises just woke up to a trap they had been walking into, here's how the conversation went (save this): ◽️ The Palantir-Nvidia deal is a bet against the model-layer duopoly. Palantir will use Nvidia's Nemotron open models to build a custom frontier-quality model for US government agencies, and the agencies own the hardware, the data, and the weights. Sacks framed it as structural: an application company and a chip company both want a competitive model layer, so they are natural partners against a two-provider middle. ◽️ Alex Karp's CNBC "crashout" was actually the thesis. Karp argued enterprises have lost trust in the frontier labs and want to own their compute, models, data, and alpha. Sacks translated it as a new definition of enterprise AI safety: safety means the model provider cannot hoover up your proprietary knowledge and turn it into its next product. ◽️ Figma is the cautionary tale that made it real. Anthropic launched Claude Design into Figma's category, its chief product officer sat on Figma's board and resigned only 3 days before launch, and Figma's stock is down about 50% this year while Anthropic's valuation surged. Sacks listed Claude Science, Security, Legal, Financial, and Code as the same move: dominate the model layer, then take the lucrative verticals. ◽️ The playbook has a name, and it is Microsoft and Google. Sacks argued Anthropic is running the operating-system strategy: own the layer everyone builds on, then walk up the stack. His Google receipt is that fewer than half of searches now send you off-site, versus an early Google that prided itself on how fast it kicked you away. ◽️ The BCG number is what raises the stakes. Chamath cited a BCG return-on-capital-employed study: the cost of capital is back to its long-run 8 to 11%, and half of large US companies cannot earn returns above it. If you are already teetering on your cost of capital, handing your alpha to a provider that may compete with you is not a luxury risk, it is fatal. ◽️ The 16.4x number is the whole argument in one data point. Chamath ran a code-migration task through 8090's harness. Wrapping Claude was 1.4x cheaper and 1.5x faster than Claude Opus alone. Wrapping the best open-source model was 16.4x cheaper, at about 3x slower. For a background task, three extra hours to cut cost by 16x is not a close call. ◽️ Even at 100x cheaper, enterprises were saying no for the wrong reason. Chamath relayed an ex-Meta PM's point that companies reject open models over China and safety fears, when they could host those same open weights on their own GPUs in US data centers with nothing flowing back. The safety objection, she argued, is backwards: the leak is the data you hand the frontier labs. ◽️ Friedberg says the frontier labs are trying to commoditize their own customers. Anthropic has been signing up life-sciences companies to feed a new life-focused model in exchange for early access, and nearly everyone he has talked to now refuses, recognizing that data they spent billions generating becomes worthless once it is pooled with everyone else's. ◽️ The deployment topology is shifting from big hubs to distributed spokes. Friedberg's map: the old assumption was a few capital-advantaged mega-clusters plus inference clouds. The new one is large hubs, medium hubs (enterprise training clusters), and distributed spokes, including on-prem inference in your own building. Owning your weights is the point. ◽️ Chamath's endgame is running GLM himself. An industry contact told him that with harness post-training and telemetry, an open Chinese model like GLM could get as good as Anthropic's Mythos. His conclusion: take GLM, control it soup-to-nuts on US hardware with only US citizens touching it, and pay a fraction. ◽️ The Apple analogy sharpens why renting intelligence is different from renting distribution. Chamath argued Apple is the only platform that respected developers, deliberately keeping its stock apps basic to protect the ecosystem and collect its 30% tax. There is no 30% tax on open models, and worse, you cannot rent intelligence from the same place that rents it to your competitor without ending up identical to them. ◽️ Nvidia's open model is now good enough to matter. Calacanis claimed you cannot tell Jensen Huang's Nemotron from Claude on 95% of searches, and that Nvidia downplayed the model until now to avoid alarming its top customers. The gloves came off once OpenAI, Anthropic, and Elon all signaled their own silicon ambitions. ◽️ Sacks sized the duopoly: roughly $60B and $40B in ARR. Anthropic is around ~$60 billion of ARR, OpenAI at ~$40 billion, and no one else generates meaningful model-layer revenue. Sacks's policy line: the US does not ban monopolies, only anti-competitive tactics, but the government should do nothing to make the duopoly more likely. ◽️ The token deflation call: 90% a year for three years. Calacanis predicted token costs fall 90% annually for three years, putting the price of intelligence near free and making it rational to waste tokens on hardware you already own. Friedberg's version is a 70/20/10 split between big cloud, local, and other clouds. ◽️ A wave of platform lock-in spending is already landing. Calacanis flagged Microsoft standing up a roughly $2.5 billion forward-deployed-engineer effort and Amazon spending about $1 billion on the same, plus OpenAI's version. His read: enterprises will slam the door, because letting a provider's engineers study your business is how it ends up in their model. ◽️ The server-per-employee prediction. Calacanis expects every employee to get $10,000 to $20,000 of local compute, a Mac Studio or a high-RAM Dell, running a personal local model that syncs to a thin laptop. A server per person, so nothing leaks. ◽️ On jobs, the data does not show present-tense loss. Sacks cited a RAMP and Revelio Labs study of over 21,000 US firms: the heaviest AI spenders grew headcount about 10% over two years, and entry-level headcount grew even faster at 12%. Friedberg's harder claim: there is no AI job loss yet, only clunky, gradual value creation, and the media will not reverse its narrative because that destroys its credibility. ◽️ The displacement case is real but forward-dated. The counterpoint on the show was that customer support, entry-level data entry and BPO, and driving are the near-term displacements, with Waymo cited as present-tense evidence: in markets where it hits critical mass, Uber and Lyft stop recruiting drivers. Sacks noted most US entry-level support was already offshored, so the acute risk sits in those countries first. ◽️ The human-premium counternarrative. Friedberg argued that as automation spreads, human interaction gets a premium: the skilled bartender, the real driver, the human-in-the-loop tier. He cited the company (referenced as Klarna) that hyped replacing its whole support team with AI, then reversed a year later on brand grounds. ◽️ The export-control episode needed three conditions, and Sacks says do not over-read it. Commerce lifted controls on Anthropic's Fable 5 after two weeks, with Mythos 5 restored to US customers around June 26 once co-founder Tom Brown replaced Dario as lead negotiator. Sacks's three conditions: Dario boasting for months about a cyber weapon, Amazon reporting failed guardrails in testing, and Dario refusing to roll Fable back. His message to allies: this was a particular set of circumstances rather than the debut of a standing lever. ◽️ The import question nobody answered cleanly. Calacanis pressed on why the US blocks Chinese cars and drones but not Chinese open models like DeepSeek and Kimi. Sacks's answer: a forked open model run on US hardware stops being Chinese, and banning open source would isolate the US and impose a token tax on American enterprises, so let the market decide if American open models win. ◽️ The California fiscal story is a business-climate story. Friedberg walked through the numbers behind Newsom's "balanced" $351B budget: expenses exceed revenue and $20-40B is borrowed to close the gap, the budget grew 65% in six years ($215B to $355B), personal income tax is $142B of ~$211B revenue with the top 1% (150,000 people) paying $70B of it, and the corporate rate of 8.9% sits far above Texas at zero. ◽️ The tax base is leaving, and the state is now taxing everyone else. Friedberg cited 1 to 1.5% of adjusted gross income leaving each year (about 15% over a decade), at least 15 Fortune 500 HQs and ~2,100 firms gone since 2019, and a new 8% software sales tax hitting Word, Gmail, and ChatGPT subscriptions plus a health-insurance tax, on top of a now-permanent 14.4% top bracket. The liabilities behind it run $1.4T in debt, up to $1.5T in unfunded pensions senior to state bonds, and ~$40B/year in out-year deficits. Lastly, the line that framed the whole show: "You can't rent intelligence from the same place that rents it to your competitor." That is the sovereignty thesis in one sentence, and every number in this episode is an argument for it. ____ Follow Fireside Alpha for more summaries on key business and technology conversations.

Fireside Alpha

55,816 görüntüleme • 2 ay önce