Sensitive content

This media may contain sensitive content.

Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

#CumbalonianLogisticServices. The Full Service Delivery and Dispatch company.... #aigenerated #aimuscle #musclev #aibodybuilder #guns #bodybuilder #cigar #jock #cigarpic (Entire sequence was generated Locally offline, with #LTXDesktop actually without LoRAs #M5ProMax 48gig)

Cumbalone

5,165 subscribers

40,701 просмотров • 1 день назад •via X (Twitter)

NSFW #CumbalonianLogisticServices #aigenerated #aimuscle #musclev #aibodybuilder #guns #bodybuilder #cigar #jock #cigarpic #LTXDesktop #M5ProMax

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

We get a bit of grief here that Cumbalone videos don't fully depict real life. #aigenerated #aimuscle #muscle #aibodybuilder #bodybuilder #cigar #jock #gay #cumbolaxmax

Sensitive content

We get a bit of grief here that Cumbalone videos don't fully depict real life. #aigenerated #aimuscle #muscle #aibodybuilder #bodybuilder #cigar #jock #gay #cumbolaxmax

Cumbalone

12,717 просмотров • 1 месяц назад

Hmmmm, its been a while since we heard from #Jason. Anyway, in his own private #mancave quarters, he explains the demands of being in #SeniorManagement in #CumbaloneTacticalServices #aigenerated #aimuscle #muscle #aibodybuilder #cigarman #cigar #cigarpic #guns #gay #gaymen

Sensitive content

Hmmmm, its been a while since we heard from #Jason. Anyway, in his own private #mancave quarters, he explains the demands of being in #SeniorManagement in #CumbaloneTacticalServices #aigenerated #aimuscle #muscle #aibodybuilder #cigarman #cigar #cigarpic #guns #gay #gaymen

Cumbalone

13,299 просмотров • 3 месяцев назад

#Gunnar : our top perfomer in #CumbaloneTacticalServices last year, in realising full cost synergies through redunant resource decommissioning. Bonus was a Vegas holdiay and one of those machine gun days #aigenerated #aimuscle #muscle #aibodybuilder #guns #bodybuilder #cigar

#Gunnar : our top perfomer in #CumbaloneTacticalServices last year, in realising full cost synergies through redunant resource decommissioning. Bonus was a Vegas holdiay and one of those machine gun days #aigenerated #aimuscle #muscle #aibodybuilder #guns #bodybuilder #cigar

Cumbalone

26,681 просмотров • 7 дней назад

Cigars vs Cigarettes.... Certainly NOT meant to offend please. Each to their own. It was just for laughs It was a requested concept. Grok made a mess of what we were trying to make .... but anyway... #aigenerated #afantasy #bodybuilder #muscle #jock #cigar #cigarette

Cigars vs Cigarettes.... Certainly NOT meant to offend please. Each to their own. It was just for laughs It was a requested concept. Grok made a mess of what we were trying to make .... but anyway... #aigenerated #afantasy #bodybuilder #muscle #jock #cigar #cigarette

Cumbalone

11,955 просмотров • 5 дней назад

Gunnar wasn't born in Cumbalonia. But he's a naturalised citizen. He likes Handkäse mit Musik ( can be an interesting musical problem later) and he likes other German things as the video shows An attempt at a #silkpurse from more of #Groks #Fuckups #muscle #gun #bodybuilder

Gunnar wasn't born in Cumbalonia. But he's a naturalised citizen. He likes Handkäse mit Musik ( can be an interesting musical problem later) and he likes other German things as the video shows An attempt at a #silkpurse from more of #Groks #Fuckups #muscle #gun #bodybuilder

Cumbalone

11,764 просмотров • 7 дней назад

Cumbalone was deeply honored to be noticed by Ace Bannon Ace Bannon Ace saw Marcc's recent outing, and wondered what Marcc might look like out of uniform? Marcc was happy to oblige given all Ace has done for guys like us His suit was soiled by creamy residue anyway! #gay

Sensitive content

Cumbalone was deeply honored to be noticed by Ace Bannon Ace Bannon Ace saw Marcc's recent outing, and wondered what Marcc might look like out of uniform? Marcc was happy to oblige given all Ace has done for guys like us His suit was soiled by creamy residue anyway! #gay

Cumbalone

15,337 просмотров • 3 месяцев назад

Pedro Martinez on adjusting his delivery to move the ball inside vs outside.🤔🔑 Notice how he's not course-correcting with his hand/wrist at the last second (as some do), but rather is talking about making a full body micro-adjustment, subtly shifting the direction of his entire delivery. When pitchers try to get the ball glove side by course correcting at the last second with the hand, this usually results in the ball being yanked or cut. It's worth pointing out that these are his feels. Pitchers don't actually release the ball that far in front of their face, and pitchers without good hip internal rotation like him won't be able to land that closed off with the front foot. But the underlying point still stands. Ignore HOF players' feels at your own peril. The feel is what creates the real. Finding out what was going on inside the heads of the greats is just as important as observing how they actually moved/performed. 📽️ MLB Network

Pedro Martinez on adjusting his delivery to move the ball inside vs outside.🤔🔑 Notice how he's not course-correcting with his hand/wrist at the last second (as some do), but rather is talking about making a full body micro-adjustment, subtly shifting the direction of his entire delivery. When pitchers try to get the ball glove side by course correcting at the last second with the hand, this usually results in the ball being yanked or cut. It's worth pointing out that these are his feels. Pitchers don't actually release the ball that far in front of their face, and pitchers without good hip internal rotation like him won't be able to land that closed off with the front foot. But the underlying point still stands. Ignore HOF players' feels at your own peril. The feel is what creates the real. Finding out what was going on inside the heads of the greats is just as important as observing how they actually moved/performed. 📽️ MLB Network

Ben Brewster

19,051 просмотров • 8 месяцев назад

One week ago, we launched Typeahead. The product is already meaningfully better. We’ve shipped new features, Typeahead 2.0 is almost ready, and we’re already planning 3.0. 3.0 is going to be big. That pace is part of the point. Sam Asante and I started a new company together because we wanted to see how far we could push local AI software. Typeahead is the first thing we shipped. It is a local AI writing app for Mac. You type, suggestions appear inline, and it learns how you actually write. It works offline. You pay once. $79 and you own it. The product is intentionally simple. The thesis behind it is bigger. We think local models are going to create a new class of software. Fast, private and offline. By default. Personal without being creepy. Useful without turning everything into a subscription. Most AI products today assume the model lives in the cloud. That will not be the only path. The machines we already own are getting powerful enough. The models are getting small enough. And the experience can start to feel less like chatting with a remote service and more like using software that belongs on your computer. We have already built a few fully local experiments together. Typeahead felt right because the value is immediate. A few people have asked me how this fits with Crazy Egg. Crazy Egg is still my main work. This is a focused company with Sam, built around a thesis we both believe in. I’ve learned that the people I work with shape the work more than almost anything else. Sam was already on my short list of people I wanted to build with. Typeahead is the first public proof of the thesis. Watch the video. Get it here:

One week ago, we launched Typeahead. The product is already meaningfully better. We’ve shipped new features, Typeahead 2.0 is almost ready, and we’re already planning 3.0. 3.0 is going to be big. That pace is part of the point. Sam Asante and I started a new company together because we wanted to see how far we could push local AI software. Typeahead is the first thing we shipped. It is a local AI writing app for Mac. You type, suggestions appear inline, and it learns how you actually write. It works offline. You pay once. $79 and you own it. The product is intentionally simple. The thesis behind it is bigger. We think local models are going to create a new class of software. Fast, private and offline. By default. Personal without being creepy. Useful without turning everything into a subscription. Most AI products today assume the model lives in the cloud. That will not be the only path. The machines we already own are getting powerful enough. The models are getting small enough. And the experience can start to feel less like chatting with a remote service and more like using software that belongs on your computer. We have already built a few fully local experiments together. Typeahead felt right because the value is immediate. A few people have asked me how this fits with Crazy Egg. Crazy Egg is still my main work. This is a focused company with Sam, built around a thesis we both believe in. I’ve learned that the people I work with shape the work more than almost anything else. Sam was already on my short list of people I wanted to build with. Typeahead is the first public proof of the thesis. Watch the video. Get it here:

Hiten Shah

18,608 просмотров • 1 месяц назад

Sometimes in #Cumbalonia its just about #bros , #buddies , #bikes and #cigars #gearfetish #gaygearfetish #gaybiker

Sometimes in #Cumbalonia its just about #bros , #buddies , #bikes and #cigars #gearfetish #gaygearfetish #gaybiker

Cumbalone

14,706 просмотров • 2 месяцев назад

the Mundanities of daily life in #Cumbalonia for #FornicScrotus , son of the #PrimeCommander of #TheAndrogenocracyOfCumbalonia #gaycigar

the Mundanities of daily life in #Cumbalonia for #FornicScrotus , son of the #PrimeCommander of #TheAndrogenocracyOfCumbalonia #gaycigar

Cumbalone

13,775 просмотров • 22 дней назад

A #Cumbalonian #LoveStory Maybe, Just Maybe, there is someone out there for us all.... And maybe it will be #Jason #gay #gaylove #rubber #gearfetish #gayfairytale #Cumbalonia #CumbaloneTacticalServices

Sensitive content

A #Cumbalonian #LoveStory Maybe, Just Maybe, there is someone out there for us all.... And maybe it will be #Jason #gay #gaylove #rubber #gearfetish #gayfairytale #Cumbalonia #CumbaloneTacticalServices

Cumbalone

37,053 просмотров • 1 месяц назад

I didn’t start my company NextLink Labs because I wanted to be a founder. I started it because production kept falling over and no one could explain why. At my last job, I was a solo developer on a critical system. Servers were going down. The service was unstable. Customers were feeling it. So I started digging. What I found wasn’t a bad engineer or a bad decision. It was something worse. The application was running on extremely old servers, filled with things that had nothing to do with the service. When I asked why we hadn’t migrated to the new servers we bought 18 months earlier, the answer stunned me: “We bought them… but no one knows how to set them up.” No documentation. No CI/CD. No infrastructure as code. No one actually knew what was running in production, so no one felt confident touching it. To make it harder, the servers were in Los Angeles. I was in Pittsburgh. I couldn’t even get hands-on to fix it myself. That was the moment it clicked. This wasn’t a one-off failure. This was how a lot of companies operated. I kept pushing for a better way. More structure. More clarity. Systems that didn’t rely on tribal knowledge. Eventually, I “wore out my welcome.” My manager didn’t fire me, but he did say, “Maybe you should do what you’re building on the side… full time.” Once I landed one more client, I resigned within four weeks. That was 11 years ago. Here's the lesson I took with me: Most outages aren’t caused by bad engineers. They’re caused by undocumented systems and neglected foundations. At NextLink Labs, everything we do comes back to that belief. Build systems people understand. Make infrastructure boring and predictable. And design for execution, not heroics. That’s how you scale without things constantly breaking.

I didn’t start my company NextLink Labs because I wanted to be a founder. I started it because production kept falling over and no one could explain why. At my last job, I was a solo developer on a critical system. Servers were going down. The service was unstable. Customers were feeling it. So I started digging. What I found wasn’t a bad engineer or a bad decision. It was something worse. The application was running on extremely old servers, filled with things that had nothing to do with the service. When I asked why we hadn’t migrated to the new servers we bought 18 months earlier, the answer stunned me: “We bought them… but no one knows how to set them up.” No documentation. No CI/CD. No infrastructure as code. No one actually knew what was running in production, so no one felt confident touching it. To make it harder, the servers were in Los Angeles. I was in Pittsburgh. I couldn’t even get hands-on to fix it myself. That was the moment it clicked. This wasn’t a one-off failure. This was how a lot of companies operated. I kept pushing for a better way. More structure. More clarity. Systems that didn’t rely on tribal knowledge. Eventually, I “wore out my welcome.” My manager didn’t fire me, but he did say, “Maybe you should do what you’re building on the side… full time.” Once I landed one more client, I resigned within four weeks. That was 11 years ago. Here's the lesson I took with me: Most outages aren’t caused by bad engineers. They’re caused by undocumented systems and neglected foundations. At NextLink Labs, everything we do comes back to that belief. Build systems people understand. Make infrastructure boring and predictable. And design for execution, not heroics. That’s how you scale without things constantly breaking.

Jordan Saunders

417,369 просмотров • 5 месяцев назад

Eight months ago, David Sacks, the White House AI and Crypto Czar publicly accused Anthropic of running a sophisticated regulatory capture campaign built on fear mongering (save this). People thought it was a spicy take and then Fable 5 release just turned it into evidence. When Anthropic released its Mythos-class models, it disclosed that every prompt and output sent through them would be retained for 30 days with no exceptions including for enterprise customers who had previously signed zero data retention agreements, and for up to two years if a prompt was flagged by a safety classifier. Microsoft moved so quickly that it restricted its own employees from using Claude Fable 5 within days of the release, citing the retention terms as incompatible with its internal policies, the largest enterprise software company in the world treating the new terms as a non-starter. But the data retention was not even the part that generated the most outrage in the developer community. The system card also disclosed that for users Anthropic suspected of working on frontier AI research, chip design, or competing model development, the system would automatically route those requests to a less capable model without telling the user, rewrite the prompt in the background, deliver a deliberately degraded response, and charge full price for access to a frontier model the user was not actually receiving. Business Insider confirmed that Anthropic's own apology acknowledged the company was intentionally giving worse answers and concealing that fact from paying customers. The examples of who triggered these filters make the safety justification difficult to defend, Ben Thompson from Stratechery was flagged for asking about the relationship between GLP-1s and cancer risk, and users asking routine questions about mitochondria were quietly downgraded, none of them aware it was happening. Under pressure, Anthropic walked back the narrowest possible piece of the policy, they will now disclose when a request is being downgraded. The underlying architecture, the 30 day retention, the behavioral profiling, the routing tiers, and the two-class access system remains fully intact. This is the part that makes David Sacks argument from October 2025 land differently today. He argued that Anthropic's safety positioning was principally a regulatory capture strategy using fear-based arguments to shape rules that would entrench incumbents and damage the broader startup ecosystem. The Fable 5 disclosure shows a company that used safety language to justify building an opaque, paternalistic system where Anthropic alone decides who is worthy of frontier AI access, profiles users to enforce that decision and collects full payment regardless.

Eight months ago, David Sacks, the White House AI and Crypto Czar publicly accused Anthropic of running a sophisticated regulatory capture campaign built on fear mongering (save this). People thought it was a spicy take and then Fable 5 release just turned it into evidence. When Anthropic released its Mythos-class models, it disclosed that every prompt and output sent through them would be retained for 30 days with no exceptions including for enterprise customers who had previously signed zero data retention agreements, and for up to two years if a prompt was flagged by a safety classifier. Microsoft moved so quickly that it restricted its own employees from using Claude Fable 5 within days of the release, citing the retention terms as incompatible with its internal policies, the largest enterprise software company in the world treating the new terms as a non-starter. But the data retention was not even the part that generated the most outrage in the developer community. The system card also disclosed that for users Anthropic suspected of working on frontier AI research, chip design, or competing model development, the system would automatically route those requests to a less capable model without telling the user, rewrite the prompt in the background, deliver a deliberately degraded response, and charge full price for access to a frontier model the user was not actually receiving. Business Insider confirmed that Anthropic's own apology acknowledged the company was intentionally giving worse answers and concealing that fact from paying customers. The examples of who triggered these filters make the safety justification difficult to defend, Ben Thompson from Stratechery was flagged for asking about the relationship between GLP-1s and cancer risk, and users asking routine questions about mitochondria were quietly downgraded, none of them aware it was happening. Under pressure, Anthropic walked back the narrowest possible piece of the policy, they will now disclose when a request is being downgraded. The underlying architecture, the 30 day retention, the behavioral profiling, the routing tiers, and the two-class access system remains fully intact. This is the part that makes David Sacks argument from October 2025 land differently today. He argued that Anthropic's safety positioning was principally a regulatory capture strategy using fear-based arguments to shape rules that would entrench incumbents and damage the broader startup ecosystem. The Fable 5 disclosure shows a company that used safety language to justify building an opaque, paternalistic system where Anthropic alone decides who is worthy of frontier AI access, profiles users to enforce that decision and collects full payment regardless.

Milk Road AI

47,312 просмотров • 1 месяц назад

David Sacks was one of the first people to get a full readout from the White House after the Fable ban. He went on the All-In podcast this week and told the story from the inside. It is not the story anyone is telling. Here is what actually happened. Dario went to Washington in April and told national security officials he had built a cyber weapon. He spiked cortisol levels across the entire administration. Got everyone focused. Then Anthropic quietly expanded the Mythos preview to over 50 companies without telling the White House. According to the Washington Post, at least one of those companies was flagged as a national security concern. That was the predicate. Then Fable launched. Mythos with guardrails. Anthropic's own largest partner started testing those guardrails and found a jailbreak. They escalated to the White House. The administration called Dario directly. A cabinet secretary picked up the phone personally. It should have been a five minute call. Instead, Dario argued. He said the jailbreak was not serious. Then he published a blog post trying to distinguish minor jailbreaks from major ones. This is the man who had just told Washington he built a cyber weapon. Sacks said it plainly. The trust is gone. And once you are in one of these situations it is always harder to get out than it was to avoid getting in. Anthropic spent years building credibility as the AI safety company. They burned it in a single week by refusing a phone call. WATCH THE FULL PODCAST ON The All-In Podcast

David Sacks was one of the first people to get a full readout from the White House after the Fable ban. He went on the All-In podcast this week and told the story from the inside. It is not the story anyone is telling. Here is what actually happened. Dario went to Washington in April and told national security officials he had built a cyber weapon. He spiked cortisol levels across the entire administration. Got everyone focused. Then Anthropic quietly expanded the Mythos preview to over 50 companies without telling the White House. According to the Washington Post, at least one of those companies was flagged as a national security concern. That was the predicate. Then Fable launched. Mythos with guardrails. Anthropic's own largest partner started testing those guardrails and found a jailbreak. They escalated to the White House. The administration called Dario directly. A cabinet secretary picked up the phone personally. It should have been a five minute call. Instead, Dario argued. He said the jailbreak was not serious. Then he published a blog post trying to distinguish minor jailbreaks from major ones. This is the man who had just told Washington he built a cyber weapon. Sacks said it plainly. The trust is gone. And once you are in one of these situations it is always harder to get out than it was to avoid getting in. Anthropic spent years building credibility as the AI safety company. They burned it in a single week by refusing a phone call. WATCH THE FULL PODCAST ON The All-In Podcast

Ihtesham Ali

261,311 просмотров • 1 месяц назад

This time in Tokyo we experienced the full range of hotels I think, from $40 to $5000/night We arrived from a long night flight where we barely slept from Singapore and then at our hotel we couldn't get early check in cause it was full (which seems to be the case everywhere you go now in Japan, everything is fully booked, soooooooooooo many tourists, like I never seen this at this level anywhere, not even Thailand or Paris!) So anyway we were in Shibuya at 7am completely homeless, roaming the streets, eating in 7/11, we walked around and tried to find other hotels to get a room like APA, but they were all full too, entire Shibuya was full! Then I remembered, Japan has those hourly 🏩 love hotels of course, so I found one, they're kinda camouflaged, you don't really see many signs, just a black wall and some stairs down, then a front desk that has half the wall covered so you can't look the front desk staff in the face (cause shame and privacy) She didn't speak English and we didn't speak Japanese but she was very helpful though and got us a room until 5pm! Most Japanese are honestly so helpful and sweet trying to get you what you want without much/any English It was a very compact and a bit old room but very clean. I kept it secret from my gf that it was a love hotel or she would never want to go in 😂👌 she also kept her jacket on thinking it was dirty My argument for the love hotel being clean was "it's cleaned more than a regular hotel" for that exact reason. And it was REALLY clean to be fair. After convincing gf, she took her bomber jacket off too I actually slept really good for a few hours It was 6700 JPY or $40 to stay 7am-5pm, so ~10 hours, not so bad. And the room had everything like soap, tooth brush, comb etc. Probably if we booked days before it could have been much cheaper. There's something in me that highly prefers these kind of compact functional hotels over luxury hotels, obviously it's the Dutchness in me (we like simple and cheap), but also I've increasingly realized luxury hotels are just facades for chains with shitty service I feel with compact functional hotels you get what you pay for and things usually just work better (I think because there's higher volume and you have less expectations), very similar to how EasyJet, RyanAir, Air Asia, and JetBlue are low-cost but actually really good airlines So this trip we experienced both $43 love hotels and $5000/night luxury hotel which was Aman (the quoted tweet below is Aman) Aman was flawless btw, and I think the best hotel I've ever stayed, we paid for a big suite, and it was a crazy amount of money, and everyone asked me how it was Well it was what you'd expect from a regular hotel, everything just works, the people are helpful, the AC works and is cold (!!!!!!!), when something goes wrong you call and there's a person at your door instantly to fix it, no need to keep calling, it's like how hotels were before "The Great Enshittification of post-COVID 2020" happened, just excellent! It says something about hidden inflation that a good luxury hotel where you actually get service you used to get is now $5000/night! Unaffordable but I guess that's how it is now! Don't call me jaded, I'm just the messenger! Obviously Asia helps, hotels are just much better here than the rest of the world in terms of service Anyway I hope you like the video tour of our love hotel 😊🏩

This time in Tokyo we experienced the full range of hotels I think, from $40 to $5000/night We arrived from a long night flight where we barely slept from Singapore and then at our hotel we couldn't get early check in cause it was full (which seems to be the case everywhere you go now in Japan, everything is fully booked, soooooooooooo many tourists, like I never seen this at this level anywhere, not even Thailand or Paris!) So anyway we were in Shibuya at 7am completely homeless, roaming the streets, eating in 7/11, we walked around and tried to find other hotels to get a room like APA, but they were all full too, entire Shibuya was full! Then I remembered, Japan has those hourly 🏩 love hotels of course, so I found one, they're kinda camouflaged, you don't really see many signs, just a black wall and some stairs down, then a front desk that has half the wall covered so you can't look the front desk staff in the face (cause shame and privacy) She didn't speak English and we didn't speak Japanese but she was very helpful though and got us a room until 5pm! Most Japanese are honestly so helpful and sweet trying to get you what you want without much/any English It was a very compact and a bit old room but very clean. I kept it secret from my gf that it was a love hotel or she would never want to go in 😂👌 she also kept her jacket on thinking it was dirty My argument for the love hotel being clean was "it's cleaned more than a regular hotel" for that exact reason. And it was REALLY clean to be fair. After convincing gf, she took her bomber jacket off too I actually slept really good for a few hours It was 6700 JPY or $40 to stay 7am-5pm, so ~10 hours, not so bad. And the room had everything like soap, tooth brush, comb etc. Probably if we booked days before it could have been much cheaper. There's something in me that highly prefers these kind of compact functional hotels over luxury hotels, obviously it's the Dutchness in me (we like simple and cheap), but also I've increasingly realized luxury hotels are just facades for chains with shitty service I feel with compact functional hotels you get what you pay for and things usually just work better (I think because there's higher volume and you have less expectations), very similar to how EasyJet, RyanAir, Air Asia, and JetBlue are low-cost but actually really good airlines So this trip we experienced both $43 love hotels and $5000/night luxury hotel which was Aman (the quoted tweet below is Aman) Aman was flawless btw, and I think the best hotel I've ever stayed, we paid for a big suite, and it was a crazy amount of money, and everyone asked me how it was Well it was what you'd expect from a regular hotel, everything just works, the people are helpful, the AC works and is cold (!!!!!!!), when something goes wrong you call and there's a person at your door instantly to fix it, no need to keep calling, it's like how hotels were before "The Great Enshittification of post-COVID 2020" happened, just excellent! It says something about hidden inflation that a good luxury hotel where you actually get service you used to get is now $5000/night! Unaffordable but I guess that's how it is now! Don't call me jaded, I'm just the messenger! Obviously Asia helps, hotels are just much better here than the rest of the world in terms of service Anyway I hope you like the video tour of our love hotel 😊🏩

@levelsio

723,022 просмотров • 8 месяцев назад

i watched gemma 4 12b build something genuinely impressive today, and then loop itself to death right in front of me. the full run is in the video, sped up but completely uncut, watch it to the end and you will catch the exact moment it stops building and starts looping right in the middle of the work. the task was clean, build a single file gravity simulator, n-body physics, orbits, collisions, running locally on one 3090 through an agent. and for ten minutes it was a joy to watch. it reached for a symplectic integrator on its own, the correct one, the kind that keeps orbits stable instead of spiralling out. real gravity with softening, proper orbital velocities, momentum conserved on collision. the physics was right. the thing actually worked. then on the very last step, writing a few tests to prove its own code, it fell into a loop. not a crash, a loop. it started repeating itself and would not stop. ten more minutes, thirty four thousand tokens into a single answer, the same fragments over and over, until i killed it myself. so it's not that gemma can't code. it did the hard part beautifully. it cannot finish. it cannot hold a long task together without unravelling, and finishing is the entire job in agentic work. here's the part that stings. i run this exact task, same harness, same card, on the chinese open models, qwen especially, and i never see this. they build it, they test it, they stop. every single time. google has the raw capability, you can see it sitting right there in the code, and then the model loops itself to death on a task a 27b from alibaba finishes clean. open weights, apache 2.0, so much to love on paper. i just need it to know when to stop talking.

i watched gemma 4 12b build something genuinely impressive today, and then loop itself to death right in front of me. the full run is in the video, sped up but completely uncut, watch it to the end and you will catch the exact moment it stops building and starts looping right in the middle of the work. the task was clean, build a single file gravity simulator, n-body physics, orbits, collisions, running locally on one 3090 through an agent. and for ten minutes it was a joy to watch. it reached for a symplectic integrator on its own, the correct one, the kind that keeps orbits stable instead of spiralling out. real gravity with softening, proper orbital velocities, momentum conserved on collision. the physics was right. the thing actually worked. then on the very last step, writing a few tests to prove its own code, it fell into a loop. not a crash, a loop. it started repeating itself and would not stop. ten more minutes, thirty four thousand tokens into a single answer, the same fragments over and over, until i killed it myself. so it's not that gemma can't code. it did the hard part beautifully. it cannot finish. it cannot hold a long task together without unravelling, and finishing is the entire job in agentic work. here's the part that stings. i run this exact task, same harness, same card, on the chinese open models, qwen especially, and i never see this. they build it, they test it, they stop. every single time. google has the raw capability, you can see it sitting right there in the code, and then the model loops itself to death on a task a 27b from alibaba finishes clean. open weights, apache 2.0, so much to love on paper. i just need it to know when to stop talking.

Sudo su

39,574 просмотров • 1 месяц назад

What does it actually mean to be AI native? There was no clear guide on the internet for how to become AI native so we built the definitive one (60 min masterclass): 1. An AI native org has 3 layers: people for strategy and taste, agents for execution, and a shared context layer that makes the entire company readable to agents. 2. AI eats the middle of your work. You used to spend 80% of your day on execution. Now agents do that. Your job is the bookends: deciding what to do and judging whether it's good enough. 3. Everyone is a manager now. Your output is the output of your agents. If your agents produce garbage, that's on you. You set them up wrong. 4. Using ChatGPT doesn't make you AI native. That's like having a website and calling yourself a tech company lol. 5. No AI native org without AI native people. Most companies skip straight to the tools. That's why it fails. If your people don't understand how to manage agents, the tech doesn't matter. 6. Making your company "readable" to agents is the real work. Every process, every decision, every piece of knowledge needs to exist in a format an agent can consume. Most companies are nowhere close. 7. Speed without signal is just expensive chaos. You need the system to move fast AND know if you're moving in the right direction. 8. The skill chain is how agents get good at your specific workflows. Skills build on skills. The more you invest in them, the more your company compounds. 9. The moat is the system. People managing agents, agents reading from rich context, the whole thing getting smarter every week. That compounds. Your competitor can copy your tools. They can't copy your system. Full episode with Theo Tabah from LCA on The Startup Ideas Podcast (SIP) 🧃. This is the stuff we normally keep internal but all the sauce is yours. Theo Tabah is the brains behind advising the world's biggest companies on AI and building AI products. Your fav CEO's first call for figuring out AI. You are in for a treat Become AI native in under 60 minutes Watch

What does it actually mean to be AI native? There was no clear guide on the internet for how to become AI native so we built the definitive one (60 min masterclass): 1. An AI native org has 3 layers: people for strategy and taste, agents for execution, and a shared context layer that makes the entire company readable to agents. 2. AI eats the middle of your work. You used to spend 80% of your day on execution. Now agents do that. Your job is the bookends: deciding what to do and judging whether it's good enough. 3. Everyone is a manager now. Your output is the output of your agents. If your agents produce garbage, that's on you. You set them up wrong. 4. Using ChatGPT doesn't make you AI native. That's like having a website and calling yourself a tech company lol. 5. No AI native org without AI native people. Most companies skip straight to the tools. That's why it fails. If your people don't understand how to manage agents, the tech doesn't matter. 6. Making your company "readable" to agents is the real work. Every process, every decision, every piece of knowledge needs to exist in a format an agent can consume. Most companies are nowhere close. 7. Speed without signal is just expensive chaos. You need the system to move fast AND know if you're moving in the right direction. 8. The skill chain is how agents get good at your specific workflows. Skills build on skills. The more you invest in them, the more your company compounds. 9. The moat is the system. People managing agents, agents reading from rich context, the whole thing getting smarter every week. That compounds. Your competitor can copy your tools. They can't copy your system. Full episode with Theo Tabah from LCA on The Startup Ideas Podcast (SIP) 🧃. This is the stuff we normally keep internal but all the sauce is yours. Theo Tabah is the brains behind advising the world's biggest companies on AI and building AI products. Your fav CEO's first call for figuring out AI. You are in for a treat Become AI native in under 60 minutes Watch

GREG ISENBERG

84,316 просмотров • 1 месяц назад

How NYPD Radio Encryption Changed the Way New Yorkers See Crime and Policing 🚨 For over half a decade, I have filmed protests, police interactions, and crime without pushing a narrative. My job was never to protect one side or attack another. My job was to document what actually happened and show the full picture. When I covered protests, I showed the entire incident from beginning to end. A lot of viral clips showing NYPD officers moving into crowds only show the final moments, leaving out what happened beforehand, including situations where protesters or anarchist groups were allegedly provoking officers, throwing objects, or escalating tensions. On multiple occasions, anarchist groups pressured me not to release certain footage because it did not fit the narrative they wanted to push. I was even warned I could become a target if I showed the full story. I refused. I was not going to edit reality to protect any movement or political side. When I moved into crime reporting, my cameras showed police doing what they are supposed to do—responding to calls, making arrests, and protecting the public. But I wasn’t only filming the NYPD; I was also filming suspects and alleged suspects. I documented what happened before, during, and after arrests. There were times when people later claimed officers attacked them for no reason. Because I was there with a camera, I was able to show the full sequence of events instead of just a short clip with no context. If officers made mistakes, that would be documented too. I wasn’t running cover for anyone—I was documenting reality. During the migrant crime wave, I was often with officers from Midtown South and Times Square as they responded to robbery patterns and searched for alleged suspects. My cameras captured officers tracking suspects, making arrests, and doing the job they were hired to do. Some of those same officers later told me they appreciated having me there because they knew I would show the full story. They said they felt more comfortable knowing an independent person was documenting what happened, because if something went wrong, the truth would be captured and have their backs and go hands on , They knew I wasn’t there to make them look good or bad—I was there to show what happened. Even while covering crime, I faced pressure from different sides. There were times when people in law enforcement asked me not to release certain footage because they felt it could make an area look unsafe or negatively impact the perception of a neighborhood. My answer was always the same: NO , the public deserves to see what is happening on the streets. Protecting an image is not more important than transparency. I believe the “crime is down” narrative also changed how the public sees what is happening in the city. With NYPD radio encryption, reporters and independent journalists lost the ability to hear calls develop and respond in real time like they once did. Before encryption, the public could see officers responding to crimes, chasing suspects, and handling situations as they unfolded. Since encryption, that type of reporting has become much harder. Instead, many viral clips now focus on moments designed for engagement—people trolling officers, provoking reactions, flirting with officers, or creating situations that make them appear ineffective or weak. Those clips can shape public perception far more than the thousands of routine police interactions that happen every day. The NYPD radio encryption didn’t just change how reporters cover police—it changed what the public gets to see. When independent cameras can no longer document the full picture in real time, the public loses the ability to see the entire story. Whether it involves protesters, police officers, or suspects, the public deserves the full truth—not just the version that gets the most attention online. By Leeroy Johnson For licensing email viralnewsnyc@gmail.com

How NYPD Radio Encryption Changed the Way New Yorkers See Crime and Policing 🚨 For over half a decade, I have filmed protests, police interactions, and crime without pushing a narrative. My job was never to protect one side or attack another. My job was to document what actually happened and show the full picture. When I covered protests, I showed the entire incident from beginning to end. A lot of viral clips showing NYPD officers moving into crowds only show the final moments, leaving out what happened beforehand, including situations where protesters or anarchist groups were allegedly provoking officers, throwing objects, or escalating tensions. On multiple occasions, anarchist groups pressured me not to release certain footage because it did not fit the narrative they wanted to push. I was even warned I could become a target if I showed the full story. I refused. I was not going to edit reality to protect any movement or political side. When I moved into crime reporting, my cameras showed police doing what they are supposed to do—responding to calls, making arrests, and protecting the public. But I wasn’t only filming the NYPD; I was also filming suspects and alleged suspects. I documented what happened before, during, and after arrests. There were times when people later claimed officers attacked them for no reason. Because I was there with a camera, I was able to show the full sequence of events instead of just a short clip with no context. If officers made mistakes, that would be documented too. I wasn’t running cover for anyone—I was documenting reality. During the migrant crime wave, I was often with officers from Midtown South and Times Square as they responded to robbery patterns and searched for alleged suspects. My cameras captured officers tracking suspects, making arrests, and doing the job they were hired to do. Some of those same officers later told me they appreciated having me there because they knew I would show the full story. They said they felt more comfortable knowing an independent person was documenting what happened, because if something went wrong, the truth would be captured and have their backs and go hands on , They knew I wasn’t there to make them look good or bad—I was there to show what happened. Even while covering crime, I faced pressure from different sides. There were times when people in law enforcement asked me not to release certain footage because they felt it could make an area look unsafe or negatively impact the perception of a neighborhood. My answer was always the same: NO , the public deserves to see what is happening on the streets. Protecting an image is not more important than transparency. I believe the “crime is down” narrative also changed how the public sees what is happening in the city. With NYPD radio encryption, reporters and independent journalists lost the ability to hear calls develop and respond in real time like they once did. Before encryption, the public could see officers responding to crimes, chasing suspects, and handling situations as they unfolded. Since encryption, that type of reporting has become much harder. Instead, many viral clips now focus on moments designed for engagement—people trolling officers, provoking reactions, flirting with officers, or creating situations that make them appear ineffective or weak. Those clips can shape public perception far more than the thousands of routine police interactions that happen every day. The NYPD radio encryption didn’t just change how reporters cover police—it changed what the public gets to see. When independent cameras can no longer document the full picture in real time, the public loses the ability to see the entire story. Whether it involves protesters, police officers, or suspects, the public deserves the full truth—not just the version that gets the most attention online. By Leeroy Johnson For licensing email [email protected]

Viral News NYC

37,465 просмотров • 26 дней назад

A Watch and Investment Story I have been a watch enthusiast since my dad got me interested in watches decades ago. About three years ago, I was in London and walked by a watch boutique called Bremont. I stepped into the store, fell in love with the watches, thought they were fairly priced if not somewhat inexpensive for the quality, and bought a number of them as gifts for friends and one for myself. As I was completing the purchases, I asked the store manager: ‘Who owns the company?’ He said, “The English brothers” – Nick and Giles English – and then he proceeded to tell me Bremont’s origin story that begins with a tragic plane crash of a 1942 Harvard trainer aircraft (an American plane purchased by the British beginning in 1938) which killed their father and almost Nick. In their ‘What does not kill me makes me stronger’ new world, Nick and Giles were inspired to pursue their dream of creating and building a British watch company, and Bremont thereafter was born. The manager gave me their contact information and I sent an email congratulating the brothers on the company, the brand, and the watches. I also offered to invest and help the business grow. Sometime later, I did a Zoom with Nick. He explained that my timing was good, as a long-time investor in the company was looking to sell some of their shares. Within a couple of months, an affiliate of mine invested coinciding with the purchase by a legacy Bremont shareholder of primary shares to provide the company with additional growth capital. I thereafter bought more shares of stock from other selling shareholders and I invested a substantial amount of additional growth capital in the company just this past week. Affiliates of mine and The Bremont Long Term Trust, a trust I recently established, now own 63% of the company. Bremont is a luxury British watch company that produces adventure and exploration watches. About one-fourth of Bremont’s sales are to the military, where the company has made custom-designed watches for more than 500 British, U.S., and other American ally squadrons around the world. While today there are very few British watch companies, the British actually created the watch industry – Rolex, notably, was a British company before it moved to Switzerland – with many of the most important technical innovations and complications of the industry having been invented in England in the 17th, 18th and 19th centuries. For the watch industry, the Swiss can be thought of like the Japanese of the U.S. auto industry, where in the early days, the Swiss made cheap imitations of British watches, but eventually by the 20th century, came to dominate the industry. For compliance reasons, I have been limited in my personal investments to private situations, principally startups, real estate, and private equity, directly and through funds. Because of my personal time limitations, I spend very little time on these investments, but through a combination of good luck, investment experience, and a good eye for talent, my collective private investment outcomes have been excellent, with a few huge successes outweighing some disappointments. I have always viewed my non-Pershing Square investments as an opportunity for learning and insights that I can apply to my day job. For example, I have found that closely following the venture investment world has provided important insights into disruptive technologies and companies that can soon become serious threats to even the largest and best capitalized public companies. Experiences at small companies also very often apply to big ones, so in my hobby of making personal investments, my returns have been both financial and educational, from my successes and from my failures. Other than tennis, I don’t have any real hobbies, but perhaps my personal investing qualifies as my second passion. To date, I have been a passive investor in Bremont, but perhaps the activist in me caused me to step forward, to recently seize the opportunity to materially increase my investment in the company, and become the non-executive chairman of a newly refreshed board. I don’t expect my chairman role at Bremont to take much of my time as it is a private company of limited scale, but I do expect my experience here will provide some helpful learnings and insights. I also thought it would be fun, interesting, and rewarding to take the X community along for the ride – at least those that are interested in watches, operations, and investing. I intend to provide periodic updates of the company’s progress on X, about our successes, our struggles, and our failures – so that we can learn and have some fun together. Think of my periodic updates as “Drive to Survive,” but for watches on X. “Time to Succeed”? You can probably come up with a better name for the series, and perhaps then I should reach out to Netflix to see if they are interested (while holding back my tears as I have watched the stock massively appreciate since our exit!). Bremont can greatly benefit by your feedback so I strongly encourage you to share your insights, critiques, and other ideas about the company and its watches on X so we can learn and improve. We will periodically award the best ideas with Bremont watches so you can have an opportunity to earn an appropriate in-kind return on your time invested in helping us succeed. In the modern era, building an independent watch company into a major company, let alone one in the U.K., has rarely if ever occurred. The watch world is littered with many such failed attempts so it is far from guaranteed that we will succeed in building a profitable and sustainable company, let alone a major independent player. My Investment History to Date with Bremont Prior to my investment in the company, Nick and Giles had taken Bremont to a reasonable scale for an independently owned watch company at about £21 million in revenues with a modest operating loss, which is an incredible accomplishment for two young men with no watch industry experience. Nick and Giles accomplishment is particularly significant in an extremely competitive industry characterized by well capitalized incumbents that control many of the top brands, e.g., Richemont, Swatch, LVMH, as well dominant, independently owned companies like Rolex, Patek Philippe, Audemars and Piguet, and a few others. Despite their progress, both Nick and Giles and the other shareholders agreed that bringing in an executive with watch industry experience would help to take the company to the next level. Shortly after I invested, we hired a search firm to identify and ultimately recruit our new CEO, Davide Cerrato, who joined in May of 2023. Davide’s entire career has been spent in the industry at Panerai, Tudor – where he famously created the Black Bay, Montblanc, and with a brief stent at HYT. Davide in turn hired other industry notables to round out the executive suite, and then the team went to work with the ambitious goal of transforming Bremont into a global industry leader. Bremont has some important competitive advantages. First and most importantly, the watches are handsome, extremely well designed, and overengineered. Davide’s team has materially upgraded materials (for example, the company now uses 904L steel for all of its steel watches – the same as Rolex – unprecedented for watches at Bremont’s price point), parts, and movements from what was already a good standard to a level comparable to watches at multiples of the price. The watches are developed, designed, manufactured, and serviced in Henley-on-Thames using principally Swiss movements. While making its own movement is an aspirational goal for Bremont, the Swiss still make the highest quality movements so that is what Bremont uses. Bremont has a unique brand story and heritage, particularly for a 23-year-old company, with two decades of credibility in war zones and air combat missions with the best war fighters and military pilots in the world. Bremont also makes a limited number of watches, around 10,000 per year, compared with more than 1.2 million for Rolex and 70,000 for Patek Philippe, and scarcity drives value in luxury goods. The combination of battlefield credibility, rarity, quality and a fair price make Bremont an extremely attractive alternative to the typical Rolex or Patek, which everyone seems to be wearing in my industry, a ubiquity in my view that loses its luster over time. Over the past 23 months, Davide and team have redesigned and focused Bremont’s range around three core offerings in Land, Sea, and Air – the Terra Nova, the Supermarine, and the Altitude – while upgrading materials, movements, and quality, updating the logo to reflect the new Land and Sea offerings (Bremont was previously perceived to only be an aviation brand), and dramatically improving manufacturing and service. The company has extended its warranty from three to five years on its new watches reflecting these improvements. Bremont benefits from having overinvested in its spectacular 35,000 square foot combined headquarters, manufacturing facility, and showroom in Henley-on-Thames, aka “The Wing” (email [email protected] to arrange a superb tour), with the capacity for substantial growth without the requirement for incremental capital investment. Lastly, Bremont now has a well-capitalized long-term major owner who would like to see Bremont become a big success and achieve its vision of returning watchmaking to the UK, and who is not looking for dividends or a liquidity event in this lifetime. I have learned over time that permanent capital and a truly long-term orientation are enormous competitive advantages for any business, and they should be very helpful here. With respect to the product, later this morning in Geneva at Watches and Wonders, the industry’s most important trade show, the company is launching its new Altitude aviation line, which builds from the company’s highly successful Martin Baker offerings. The company will also be introducing some complicated new watches including two unique, jumping hour models, a 12-piece new tourbillion model (the company's first 30-piece tourbillion watch sold out shortly after its launch last year), and the company’s first perpetual calendar (limited to 50 pieces). I will be back with further updates as we continue to make progress. In the meantime, please check out and visit the Wing ([email protected] ) and our boutiques in Mayfair and throughout the U.K., on Madison Avenue and 53rd Street, in Hong Kong, and in Melbourne. Pictures don’t do Bremonts justice so you should go in person and tell the store manager that I sent you. Also, please take a look at our new launch video, and don't forget to let me know what you think.

Bill Ackman

1,625,720 просмотров • 1 год назад

🚨 WARNING: NVIDIA x ELON MUSK DEAL IS BUILT ON FAKE NUMBERS!! Michael Burry published an analysis calling the structure “Fugazi”, meaning fake. If the structure is real, we could be heading for a COLLAPSE: He is alleging that BILLIONS of dollars in Nvidia chips are being hidden off balance sheets, and that American retirees are unknowingly funding the whole thing. Nvidia, the world's largest AI chip company sold $5.4 BILLION worth of its most advanced GPUs, the GB200, to a company called Valor. Valor is not a real operating business. It is a special purpose vehicle, a shell company created specifically to hold these chips and nothing else. Nvidia also invested $1.9 BILLION of its own money directly into Valor on top of the sale. Those 100,000+ chips are now physically inside xAI's data center. xAI is Elon Musk's artificial intelligence company, the one that builds Grok. xAI is using every single one of those chips right now to run its AI models. But here is what Burry is flagging. Neither Nvidia nor xAI owns those chips on paper. Valor, the shell company holds legal title. That means $5.4 BILLION in GPU assets do not show up on Nvidia's balance sheet as inventory. They do not show up on xAI's balance sheet as assets. They are legally invisible to both companies. Nvidia gets to book the $5.4 BILLION as a completed sale and record it as revenue. xAI gets full use of the chips without owning them. And the risk disappears into a shell company in the middle. Now here is where American retirees enter the picture. Valor needed $3.5 BILLION in debt to fund this structure. Apollo provided it. Apollo is one of the largest asset managers on earth with $1.03 TRILLION under management and $834 BILLION specifically in private credit. Apollo raised the $3.5 BILLION, packaged it into debt securities, and sold those securities to Athene. Athene is Apollo's own insurance company. It sells fixed and indexed annuities, retirement savings products, to ordinary Americans. When a retiree buys an Athene annuity, they believe their money is sitting in safe, stable investments. That money is now inside a structure funding Elon Musk's AI data center. The numbers inside Athene are most alarming. Athene holds $74.2 BILLION in reserves. It has moved $217 BILLION in assets into a captive insurer based in Bermuda, meaning those assets sit outside normal US insurance regulation and oversight. Of the entire portfolio, 34.7%, equal to $103 BILLION, is classified as Level 3 assets. Level 3 is an accounting classification that means there is no observable market price for these assets. No outside party can independently verify what they are actually worth. The leverage sitting on top of those unpriced assets is 16 times. Burry's says: Every step of this structure is technically legal and publicly disclosed. But the entire thing was deliberately engineered across 8 to 12 steps to move credit risk off balance sheets and away from any market pricing. Nvidia books the revenue. Apollo collects the fees. xAI gets the computing power. And retirees sitting at the bottom of a 16x leveraged Bermuda insurance structure, holding $103 BILLION in assets with no market price carry the risk without knowing it exists. I’ve been in finance for more than 15 years. When I EXIT the markets completely, I’ll say it here publicly, like I always do. Turn notifications on. If you’re not following yet, you’ll understand why that was a mistake later.

🚨 WARNING: NVIDIA x ELON MUSK DEAL IS BUILT ON FAKE NUMBERS!! Michael Burry published an analysis calling the structure “Fugazi”, meaning fake. If the structure is real, we could be heading for a COLLAPSE: He is alleging that BILLIONS of dollars in Nvidia chips are being hidden off balance sheets, and that American retirees are unknowingly funding the whole thing. Nvidia, the world's largest AI chip company sold $5.4 BILLION worth of its most advanced GPUs, the GB200, to a company called Valor. Valor is not a real operating business. It is a special purpose vehicle, a shell company created specifically to hold these chips and nothing else. Nvidia also invested $1.9 BILLION of its own money directly into Valor on top of the sale. Those 100,000+ chips are now physically inside xAI's data center. xAI is Elon Musk's artificial intelligence company, the one that builds Grok. xAI is using every single one of those chips right now to run its AI models. But here is what Burry is flagging. Neither Nvidia nor xAI owns those chips on paper. Valor, the shell company holds legal title. That means $5.4 BILLION in GPU assets do not show up on Nvidia's balance sheet as inventory. They do not show up on xAI's balance sheet as assets. They are legally invisible to both companies. Nvidia gets to book the $5.4 BILLION as a completed sale and record it as revenue. xAI gets full use of the chips without owning them. And the risk disappears into a shell company in the middle. Now here is where American retirees enter the picture. Valor needed $3.5 BILLION in debt to fund this structure. Apollo provided it. Apollo is one of the largest asset managers on earth with $1.03 TRILLION under management and $834 BILLION specifically in private credit. Apollo raised the $3.5 BILLION, packaged it into debt securities, and sold those securities to Athene. Athene is Apollo's own insurance company. It sells fixed and indexed annuities, retirement savings products, to ordinary Americans. When a retiree buys an Athene annuity, they believe their money is sitting in safe, stable investments. That money is now inside a structure funding Elon Musk's AI data center. The numbers inside Athene are most alarming. Athene holds $74.2 BILLION in reserves. It has moved $217 BILLION in assets into a captive insurer based in Bermuda, meaning those assets sit outside normal US insurance regulation and oversight. Of the entire portfolio, 34.7%, equal to $103 BILLION, is classified as Level 3 assets. Level 3 is an accounting classification that means there is no observable market price for these assets. No outside party can independently verify what they are actually worth. The leverage sitting on top of those unpriced assets is 16 times. Burry's says: Every step of this structure is technically legal and publicly disclosed. But the entire thing was deliberately engineered across 8 to 12 steps to move credit risk off balance sheets and away from any market pricing. Nvidia books the revenue. Apollo collects the fees. xAI gets the computing power. And retirees sitting at the bottom of a 16x leveraged Bermuda insurance structure, holding $103 BILLION in assets with no market price carry the risk without knowing it exists. I’ve been in finance for more than 15 years. When I EXIT the markets completely, I’ll say it here publicly, like I always do. Turn notifications on. If you’re not following yet, you’ll understand why that was a mistake later.

WhaleTwits

48,758 просмотров • 2 месяцев назад