Sensitive content

This media may contain sensitive content.

Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Audio version! It was YOUR idea btw~ ♠️✨ Caption by: Pandora's censor (commissions open) VA: Me <3 #lewdrp #sexrp #openrp #blackedrp #ntrrp #blacked #hentai #captions

jessiehatesya's profile picture

Jessie

5,926 subscribers

61,728 görüntüleme • 8 ay önce •via X (Twitter)

NSFW#lewdrp#sexrp#openrp#blackedrp#ntrrp#blacked#hentai#captions
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

ICE was holding a 5 year old autistic child on the street as bait to get a mother to leave her house. This is how responds - help fund it here: and get it into the hands of millions of immigrants and allies. ⸻ IF ICE IS OUTSIDE AND HAS YOUR CHILD: DO NOT GO OUTSIDE. 1. Do not go outside. ICE is baiting you. Your child being outside does not change your rights. If you step out, they may try to detain you. Stay inside. Your child is protected by constitutional rights. 2. Do not call the police unless your child is in immediate physical danger. Most police will not intervene and may cooperate with ICE. Call only if your child’s safety is at risk — not for immigration issues. 3. Do not open your door. ICE cannot legally enter your home without a judicial warrant signed by a judge. Ask to see it through the window or door. An administrative warrant (signed by ICE) does not give them legal entry. 4. Call an immigration lawyer or rapid response legal hotline immediately. If you don’t have one, I can help you find one based on your location. 5. Document everything. If safe, record video or audio through the window or door. Clearly say: “I do not consent to a search.” 6. Do not answer questions. You have the right to remain silent. Do not discuss your immigration status. Do not sign anything. 7. Get a witness if possible. Call a trusted neighbor, friend, or community group to observe and document. 8. Repeat clearly: “I do not consent to entry.” “I want to speak to a lawyer.” “I am not opening the door.” ⸻ Would you like me to look up a hotline or legal defender in your area now?
0:49

ICE was holding a 5 year old autistic child on the street as bait to get a mother to leave her house. This is how responds - help fund it here: and get it into the hands of millions of immigrants and allies. ⸻ IF ICE IS OUTSIDE AND HAS YOUR CHILD: DO NOT GO OUTSIDE. 1. Do not go outside. ICE is baiting you. Your child being outside does not change your rights. If you step out, they may try to detain you. Stay inside. Your child is protected by constitutional rights. 2. Do not call the police unless your child is in immediate physical danger. Most police will not intervene and may cooperate with ICE. Call only if your child’s safety is at risk — not for immigration issues. 3. Do not open your door. ICE cannot legally enter your home without a judicial warrant signed by a judge. Ask to see it through the window or door. An administrative warrant (signed by ICE) does not give them legal entry. 4. Call an immigration lawyer or rapid response legal hotline immediately. If you don’t have one, I can help you find one based on your location. 5. Document everything. If safe, record video or audio through the window or door. Clearly say: “I do not consent to a search.” 6. Do not answer questions. You have the right to remain silent. Do not discuss your immigration status. Do not sign anything. 7. Get a witness if possible. Call a trusted neighbor, friend, or community group to observe and document. 8. Repeat clearly: “I do not consent to entry.” “I want to speak to a lawyer.” “I am not opening the door.” ⸻ Would you like me to look up a hotline or legal defender in your area now?

Joe Walsh

30,619 görüntüleme • 8 ay önce

Grok Bot is officially available in Tesla vehicles! I had early access for my Grok Bot team to be integrated directly inside my Cybertruck with the help of the SpaceXAI and Cursor team. And this is truly one of those things that’s hard to fully understand the power of it until you actually lived with it. It’s the same Grok Voice experience you already know in your Tesla… no separate app is needed. When you’re signed into Grok, your Grok Bots are right there and you can talk directly with all of them with just your voice. So today, Tesla FSD was driving me. I said, “Hey Grok.” Then I talked to my CEO bot like I would talk to someone working for me. I told it what I wanted for dinner tonight: 1/ Blueberries 2/ Strawberries 3/ Ribeye 4/ Shrimp 5/ And taco night for my family of four. That was it. I didn’t sit there searching for every single ingredient. I didn’t open Amazon and start building a cart item by item. I just explained what I wanted. Then, my CEO bot routed the job. My Chef bot figured out what we needed for taco night. Then my Amazon Cart Builder bot started putting the cart together and choosing the best items based on my purchase history. And while all of that was happening… FSD drove me to the gym and parked itself. I went inside and worked out. My bots kept working. And by the time I was finished, the cart was ready for me. I opened it. Looked everything over. Then I just clicked Order. Grok Bot + Tesla FSD completely removed all the manual work in between that would have wasted part of my day researching building a grocery list, getting the Amazon cart ready, and manually driving to my destination. And THAT was the wow moment for me… I just talked to Grok. FSD drove. I worked out. My Grok Bot team handled the busy work. When I finished the gym, the cart was simply just waiting for my approval. And all I had to do was click order. (btw you can make your Grok Bot place the order automatically as well, but I’m super picky with my food so mine are specifically built so I review it at the end). With this technology, you’re no longer opening five different apps and doing every tiny step yourself. You just tell your team what you want with your voice. They go work on it. You go live your life. The biggest upgrade here is that Grok Bot + Tesla FSD is giving us back the one thing we can never make more of… TIME. Grok Bot handles the work. Tesla FSD handles the drive. I get my time back. This is really the future… I’m glad people will be able to try out Grok Bot integrated with their Teslas today… this is truly life-changing technology.
1:20

Grok Bot is officially available in Tesla vehicles! I had early access for my Grok Bot team to be integrated directly inside my Cybertruck with the help of the SpaceXAI and Cursor team. And this is truly one of those things that’s hard to fully understand the power of it until you actually lived with it. It’s the same Grok Voice experience you already know in your Tesla… no separate app is needed. When you’re signed into Grok, your Grok Bots are right there and you can talk directly with all of them with just your voice. So today, Tesla FSD was driving me. I said, “Hey Grok.” Then I talked to my CEO bot like I would talk to someone working for me. I told it what I wanted for dinner tonight: 1/ Blueberries 2/ Strawberries 3/ Ribeye 4/ Shrimp 5/ And taco night for my family of four. That was it. I didn’t sit there searching for every single ingredient. I didn’t open Amazon and start building a cart item by item. I just explained what I wanted. Then, my CEO bot routed the job. My Chef bot figured out what we needed for taco night. Then my Amazon Cart Builder bot started putting the cart together and choosing the best items based on my purchase history. And while all of that was happening… FSD drove me to the gym and parked itself. I went inside and worked out. My bots kept working. And by the time I was finished, the cart was ready for me. I opened it. Looked everything over. Then I just clicked Order. Grok Bot + Tesla FSD completely removed all the manual work in between that would have wasted part of my day researching building a grocery list, getting the Amazon cart ready, and manually driving to my destination. And THAT was the wow moment for me… I just talked to Grok. FSD drove. I worked out. My Grok Bot team handled the busy work. When I finished the gym, the cart was simply just waiting for my approval. And all I had to do was click order. (btw you can make your Grok Bot place the order automatically as well, but I’m super picky with my food so mine are specifically built so I review it at the end). With this technology, you’re no longer opening five different apps and doing every tiny step yourself. You just tell your team what you want with your voice. They go work on it. You go live your life. The biggest upgrade here is that Grok Bot + Tesla FSD is giving us back the one thing we can never make more of… TIME. Grok Bot handles the work. Tesla FSD handles the drive. I get my time back. This is really the future… I’m glad people will be able to try out Grok Bot integrated with their Teslas today… this is truly life-changing technology.

Teslaconomics

552,270 görüntüleme • 7 gün önce

You can’t make anyone else change; all you can do is create the conditions where change is more likely to happen. I remember speaking with an administrator and hearing him complain that his staff wasn’t moving forward. Instead of commiserating with him at that moment, I said, “Maybe it’s you?” It caught him off guard, but there was a point. Too often, myself included, when we try to lead people to new ideas, and they don’t immediately embrace them, we blame them instead of looking at where we may be deficient. Sometimes, the new “approach” is sharing the ideas over again, only louder. That approach rarely works. In this week’s newsletter, I shared 3 things that have helped me rethink how to help create better conditions for others to embrace change. 1. Truly listen to their thoughts and viewpoints, and be open to the idea that your way might not be the best. 2. Recognize that you also had resistance to new ideas, including things that you might consistently do and believe today. Be open with your own story of resistance and what made you change. 3. Start by focusing on the areas in which others are excelling, as opposed to your perception of their deficits. People who feel valued are more likely to move forward. People who feel you are trying to “fix” them will fight you non-stop. You can only control your own actions. What do others see in you that will make them want to move forward, by your side? People are more likely to take steps forward on a path if they see footprints from others already on the ground. Read the entire newsletter here:
1:29

You can’t make anyone else change; all you can do is create the conditions where change is more likely to happen. I remember speaking with an administrator and hearing him complain that his staff wasn’t moving forward. Instead of commiserating with him at that moment, I said, “Maybe it’s you?” It caught him off guard, but there was a point. Too often, myself included, when we try to lead people to new ideas, and they don’t immediately embrace them, we blame them instead of looking at where we may be deficient. Sometimes, the new “approach” is sharing the ideas over again, only louder. That approach rarely works. In this week’s newsletter, I shared 3 things that have helped me rethink how to help create better conditions for others to embrace change. 1. Truly listen to their thoughts and viewpoints, and be open to the idea that your way might not be the best. 2. Recognize that you also had resistance to new ideas, including things that you might consistently do and believe today. Be open with your own story of resistance and what made you change. 3. Start by focusing on the areas in which others are excelling, as opposed to your perception of their deficits. People who feel valued are more likely to move forward. People who feel you are trying to “fix” them will fight you non-stop. You can only control your own actions. What do others see in you that will make them want to move forward, by your side? People are more likely to take steps forward on a path if they see footprints from others already on the ground. Read the entire newsletter here:

George Couros

15,740 görüntüleme • 2 yıl önce

Fake Journalism Part 2 (AKA: Two Grown Adults Play ‘Anonymous Source’ With Burner Emails) Since I was recently reminded of the time that Jess once pretended to have a “source” for the Joe Flipperhead/Olivia audio… let’s finish the story. Because this accidentally cracked open a vault of stupidity that deserves daylight. To recap Part 1, I already exposed how "Journalism Jesus” made a fake Twitter profile of a random journalist just to dodge a copyright strike, a profile he literally forgot I watched him create when he started trying to slither back onto my page a couple weeks back. You can’t make this level of stupidity up. We also established that Jess spun a full investigative-fiction saga about: • talking to a source, • spending an entire day “authenticating” the audio, • verifying details, • receiving a whistleblower drop. Reality: It was Aidan. Just Aidan. No source. No reporting. Aidan told me he was the one who sent it to her, and I asked why he didn't intervene while Nurse_Kim was being blamed by Joe Flipperhead . His response? It'll die down. Who care. I can't be connected to proton emails right now. Kk. Just two chaotic people doing Dollar Tree journalism. And somehow… Part Two is even more embarrassing. Aidan wanted the full Karen recording leaked. Jess volunteered as tribute. When only the partial clips leaked, Aidan spiraled. They weren’t hearing the whole audio which he thought made him look better. (Spoiler alert- it doesn’t. Everyone knows he constantly fires people up prior to recording them while he remains calm). Suddenly Jess posts her dramatic tweet begging an “anonymous source” to send her the entire recording. Totally subtle, right? And shocker- minutes later she magically receives an email with a clip. Who sent it? Aidan. Again. The man is allergic to subtlety. And here’s the part Aidan REALLY doesn’t want you to know: I actually begged Aidan not to do this. The second I saw Jess’s “send me the whole thing” tweet, I knew EXACTLY what was happening, she was signaling him to leak the full recording through a fake source. I told him: “Delete it. Stop sharing it. Stop having it at all. Don’t do it. You can’t trust her.” He was out on bail, and I was worried he’d get himself in deeper trouble. I was also aware both of his lawyers begged him to not share it. He didn’t ultimately do it, but if it ends up in Jess’s paws, we know the real source. Let’s pause for this gem: Jess’s October 3 tweet. She claimed she only operates on “facts and receipts.” Girl… Your “facts” are Aidan’s tantrums. Your “receipts” are whatever file/text message/audio he shoves your way that he doesn’t want to dirty his own hands with. She’s not a journalist. She’s his intern assistant with no salary, no oversight, and no self awareness. She’ll post anything or do anything he’s too cowardly or too legally exposed to do himself, including calling people’s employers. Peak intellectual rigor. And now, the Proton forensic FAIL that broke the internet: Email to Joe Flipperhead: tbwireTAPkr Email to Jess: tbwireTAPPEDkr They really thought adding “ED” made it look like a second whistleblower. This is mental illness disguised as “journalism.” To this day, the full recording has not been released. Likely so he can continue the fantasy that it would actually make Karen look bad (it doesn’t). But as always, he can’t get out of his own way. He always has to make the stupid, impulsive decision. He has to push the self-destruct button. This is why every bridge in his life ends up burned to ash. He can’t stop destroying himself. And these are the people calling me a liar? Please. Part Three is coming. Buckle up. #SarahFinchWasRight
1:41

Fake Journalism Part 2 (AKA: Two Grown Adults Play ‘Anonymous Source’ With Burner Emails) Since I was recently reminded of the time that Jess once pretended to have a “source” for the Joe Flipperhead/Olivia audio… let’s finish the story. Because this accidentally cracked open a vault of stupidity that deserves daylight. To recap Part 1, I already exposed how "Journalism Jesus” made a fake Twitter profile of a random journalist just to dodge a copyright strike, a profile he literally forgot I watched him create when he started trying to slither back onto my page a couple weeks back. You can’t make this level of stupidity up. We also established that Jess spun a full investigative-fiction saga about: • talking to a source, • spending an entire day “authenticating” the audio, • verifying details, • receiving a whistleblower drop. Reality: It was Aidan. Just Aidan. No source. No reporting. Aidan told me he was the one who sent it to her, and I asked why he didn't intervene while Nurse_Kim was being blamed by Joe Flipperhead . His response? It'll die down. Who care. I can't be connected to proton emails right now. Kk. Just two chaotic people doing Dollar Tree journalism. And somehow… Part Two is even more embarrassing. Aidan wanted the full Karen recording leaked. Jess volunteered as tribute. When only the partial clips leaked, Aidan spiraled. They weren’t hearing the whole audio which he thought made him look better. (Spoiler alert- it doesn’t. Everyone knows he constantly fires people up prior to recording them while he remains calm). Suddenly Jess posts her dramatic tweet begging an “anonymous source” to send her the entire recording. Totally subtle, right? And shocker- minutes later she magically receives an email with a clip. Who sent it? Aidan. Again. The man is allergic to subtlety. And here’s the part Aidan REALLY doesn’t want you to know: I actually begged Aidan not to do this. The second I saw Jess’s “send me the whole thing” tweet, I knew EXACTLY what was happening, she was signaling him to leak the full recording through a fake source. I told him: “Delete it. Stop sharing it. Stop having it at all. Don’t do it. You can’t trust her.” He was out on bail, and I was worried he’d get himself in deeper trouble. I was also aware both of his lawyers begged him to not share it. He didn’t ultimately do it, but if it ends up in Jess’s paws, we know the real source. Let’s pause for this gem: Jess’s October 3 tweet. She claimed she only operates on “facts and receipts.” Girl… Your “facts” are Aidan’s tantrums. Your “receipts” are whatever file/text message/audio he shoves your way that he doesn’t want to dirty his own hands with. She’s not a journalist. She’s his intern assistant with no salary, no oversight, and no self awareness. She’ll post anything or do anything he’s too cowardly or too legally exposed to do himself, including calling people’s employers. Peak intellectual rigor. And now, the Proton forensic FAIL that broke the internet: Email to Joe Flipperhead: tbwireTAPkr Email to Jess: tbwireTAPPEDkr They really thought adding “ED” made it look like a second whistleblower. This is mental illness disguised as “journalism.” To this day, the full recording has not been released. Likely so he can continue the fantasy that it would actually make Karen look bad (it doesn’t). But as always, he can’t get out of his own way. He always has to make the stupid, impulsive decision. He has to push the self-destruct button. This is why every bridge in his life ends up burned to ash. He can’t stop destroying himself. And these are the people calling me a liar? Please. Part Three is coming. Buckle up. #SarahFinchWasRight

The old M can’t come to the phone right now

60,048 görüntüleme • 9 ay önce

BOOOM! WE DID IT! BRAINWAVE TO REAL-TIME MUSIC AI! It has been a life long decades quest to read brain activity and to convert it to words, and/or music, colors and/or images. Today I am very excited to announce with the assistance from Mr. Grok director of The Zero-Human Lab, we have solved brainwave to music and this is the absolute worse it will be. We found the code using an array of NeuroSky toy chips and our software pipeline connecting to open source ACE-Step 1.5 and a highly modified LoRA model we built for this. The lyric version is in testing now. This would mean that the model will interpret words from the brainwaves and music! Today we have the music side done and the quality and genera will expand. The is the worse it will sound. Your Brainwave Music™️will be cut into 2-5 minute pieces based on a number of factors. The specimen below is from a dream/hypnogogic state I was in last night and I have a recording of my thoughts after the state. The music was made in real-time and GUIDED the dream state with known technology like binaural beats (not easy to hear in this clip) and word back masking. This specimen below shows the interplay of my brain state to the music made by my brain and adjusted to produce profound insights. I solved a very difficult issue in this session with a new AI model. IT FREAKING WORKS! THIS IS OUR FUTURE OF MORE POWERFUL BRAIN FUNCTION! Our goal is to produce a portable device you wear and will be able to give real-time audio and PEMF (skull region), ultrasound (temple region) to maximize creativity and remote viewing. It is very early days but I wanted you to know first! YOUR support of my X account, just by reading this and sharing it, subscribing to my X, buying me a and becoming a member at supports this research. I will open source this at some point and build a device ANYONE can own. Thank you! I love you.
4:17

BOOOM! WE DID IT! BRAINWAVE TO REAL-TIME MUSIC AI! It has been a life long decades quest to read brain activity and to convert it to words, and/or music, colors and/or images. Today I am very excited to announce with the assistance from Mr. Grok director of The Zero-Human Lab, we have solved brainwave to music and this is the absolute worse it will be. We found the code using an array of NeuroSky toy chips and our software pipeline connecting to open source ACE-Step 1.5 and a highly modified LoRA model we built for this. The lyric version is in testing now. This would mean that the model will interpret words from the brainwaves and music! Today we have the music side done and the quality and genera will expand. The is the worse it will sound. Your Brainwave Music™️will be cut into 2-5 minute pieces based on a number of factors. The specimen below is from a dream/hypnogogic state I was in last night and I have a recording of my thoughts after the state. The music was made in real-time and GUIDED the dream state with known technology like binaural beats (not easy to hear in this clip) and word back masking. This specimen below shows the interplay of my brain state to the music made by my brain and adjusted to produce profound insights. I solved a very difficult issue in this session with a new AI model. IT FREAKING WORKS! THIS IS OUR FUTURE OF MORE POWERFUL BRAIN FUNCTION! Our goal is to produce a portable device you wear and will be able to give real-time audio and PEMF (skull region), ultrasound (temple region) to maximize creativity and remote viewing. It is very early days but I wanted you to know first! YOUR support of my X account, just by reading this and sharing it, subscribing to my X, buying me a and becoming a member at supports this research. I will open source this at some point and build a device ANYONE can own. Thank you! I love you.

Brian Roemmele

199,054 görüntüleme • 3 ay önce

Johnny Cash’s “Hurt” is considered one of the greatest cover songs ever. Its back story is incredible. Released in 2002, Cash worked with Rick Rubin to adapt the 1994 Nine Inch Nails song. The pair met in 1992 after Rubin saw a then 60-year old Cash perform at a Bob Dylan concert in NY. Cash was decades removed from his country peak and recently dropped by his record label. Rubin felt Cash still had more music to make and told him, “What I would do is let you sit down before a microphone with your guitar and sing any song you want to record.” When Rubin first pitched “Hurt” to Cash, the music legend balked. “I can’t do that song,” said Cash. “It’s not my style.” By 2002, though, Rubin and Cash had made 3 critically-acclaimed “American Recordings” albums and were working on a 4th. On the music annotation website Genius, Rubin explained their process: ➡️ “There were a lot of songs that he needed to be convincing about. Eventually, he trusted me enough that if I felt strongly about something, he’d do it. I would send him compilations of CDs of songs to listen to, and I remember that on several compilations in a row, “Hurt” was the first song. There’s just something about it. I imagined him saying those words being very powerful.” ⬅️ When Rubin told Trent Reznor — his friend and Nine Inch Nails lead singer — about the cover idea, Reznor was “flattered” but felt it was “a bit gimmicky”. Everything changed after Reznor saw Cash’s iconic “Hurt” music video (clip below). ➡️ “A CD shows up with the track…the song in particular is straight from the soul…it was a good version and surely wasn’t cringey or anything. But it felt like I was watching my girlfriend kiss somebody else. […] A few weeks later, a video tape shows up with Mark Romanek’s [music] video on it…I pop the video in and wow…tears welling, silence, goosebumps. Wow, I just lost my girlfriend because that’s not my song anymore. Then it all made sense to me. It made me really think about how powerful music is as a medium and an art form. I wrote some words and music in my bedroom as a way of staying sane…Some-f**king-how that winds up reinterpreted by a music legend from a wildly different era/genre but still retains sincerity and…every bit as pure. Things felt even stranger when he passed away [in 2003]. The song’s purpose shifted again. It’s extremely flattering as a writer to have your song chosen by someone whose a great writer and a great artist.” ⬅️
1:50

Johnny Cash’s “Hurt” is considered one of the greatest cover songs ever. Its back story is incredible. Released in 2002, Cash worked with Rick Rubin to adapt the 1994 Nine Inch Nails song. The pair met in 1992 after Rubin saw a then 60-year old Cash perform at a Bob Dylan concert in NY. Cash was decades removed from his country peak and recently dropped by his record label. Rubin felt Cash still had more music to make and told him, “What I would do is let you sit down before a microphone with your guitar and sing any song you want to record.” When Rubin first pitched “Hurt” to Cash, the music legend balked. “I can’t do that song,” said Cash. “It’s not my style.” By 2002, though, Rubin and Cash had made 3 critically-acclaimed “American Recordings” albums and were working on a 4th. On the music annotation website Genius, Rubin explained their process: ➡️ “There were a lot of songs that he needed to be convincing about. Eventually, he trusted me enough that if I felt strongly about something, he’d do it. I would send him compilations of CDs of songs to listen to, and I remember that on several compilations in a row, “Hurt” was the first song. There’s just something about it. I imagined him saying those words being very powerful.” ⬅️ When Rubin told Trent Reznor — his friend and Nine Inch Nails lead singer — about the cover idea, Reznor was “flattered” but felt it was “a bit gimmicky”. Everything changed after Reznor saw Cash’s iconic “Hurt” music video (clip below). ➡️ “A CD shows up with the track…the song in particular is straight from the soul…it was a good version and surely wasn’t cringey or anything. But it felt like I was watching my girlfriend kiss somebody else. […] A few weeks later, a video tape shows up with Mark Romanek’s [music] video on it…I pop the video in and wow…tears welling, silence, goosebumps. Wow, I just lost my girlfriend because that’s not my song anymore. Then it all made sense to me. It made me really think about how powerful music is as a medium and an art form. I wrote some words and music in my bedroom as a way of staying sane…Some-f**king-how that winds up reinterpreted by a music legend from a wildly different era/genre but still retains sincerity and…every bit as pure. Things felt even stranger when he passed away [in 2003]. The song’s purpose shifted again. It’s extremely flattering as a writer to have your song chosen by someone whose a great writer and a great artist.” ⬅️

Trung Phan

1,841,705 görüntüleme • 2 yıl önce

I spent $4,699 on a DGX Spark and used it as a chatbot for six months. Not proud of it. But I think most Spark owners are doing the same thing, so let me tell you what changed. The moment was dumb. I had a recording of a client call, needed notes, and my first instinct was to open a browser tab and upload it to some transcription SaaS. Then I looked at the box on my desk. 128 GB of unified memory. Sitting there. Running a chat model I'd asked maybe 30 questions that day. And I thought: I have a supercomputer and I'm about to pay a stranger to listen to my client call. So I tried to do it locally. And here's where it got annoying: - my LLM can't hear audio - the ASR model I found is great but it's a CLI script - the transcript came out in two languages, need translation - summary needs the LLM again - now I want a mind map, that's another tool - oh and here's a scanned PDF from the same meeting, OCR, another tool Copy paste copy paste. Output saved to ~/Downloads/final_v2_REAL.txt. That's when it clicked: tokens per second was never the point of this box. A 24 GB gaming card forces you to pick ONE model, so everyone picks the biggest generalist and asks it to do everything. But 128 GB doesn't force that. You can keep six specialists loaded at once. A 1.7B speech model transcribes better than a 35B chat model ever will. A tiny translation model just translates and doesn't get creative. The big LLM does the thinking, and only the thinking. The hardware was never the problem. The missing piece was the thing that ties them together. So I built it. SparklingKit. (the name is a tribute to the Spark. also: a spark is one flash, sparkling is when it keeps going) What it actually is: → transcription, OCR, translation, image grounding, text-to-image, mind maps, chat. one panel. → six models hand-picked to fit together in 128 GB with room left over → queue-first: drop a 3-hour recording, go get coffee, check from your phone → nothing is a dead end: finish a transcript and "translate / map / chat with it" are already waiting → visual workflows: chain the steps once, run them forever → everything is a file. one folder. no database. copy it to back up. point Obsidian at it. → nothing leaves your machine unless you send it. The thing I care about most: if you stop using SparklingKit tomorrow, nothing happens to your data. It was already a folder of markdown and json. There's no lock-in to escape. Spark owners: you bought room, not a faster chatbot. This is how I'm using the room. No Spark? The panel runs anywhere Docker runs. Point it at your own endpoints. It's open source, and it's version one. Rough edges exist. Issues and PRs very welcome. Full details and a one-line install at If you've ever uploaded something private to a cloud AI and felt weird about it, this is for you.
1:23

I spent $4,699 on a DGX Spark and used it as a chatbot for six months. Not proud of it. But I think most Spark owners are doing the same thing, so let me tell you what changed. The moment was dumb. I had a recording of a client call, needed notes, and my first instinct was to open a browser tab and upload it to some transcription SaaS. Then I looked at the box on my desk. 128 GB of unified memory. Sitting there. Running a chat model I'd asked maybe 30 questions that day. And I thought: I have a supercomputer and I'm about to pay a stranger to listen to my client call. So I tried to do it locally. And here's where it got annoying: - my LLM can't hear audio - the ASR model I found is great but it's a CLI script - the transcript came out in two languages, need translation - summary needs the LLM again - now I want a mind map, that's another tool - oh and here's a scanned PDF from the same meeting, OCR, another tool Copy paste copy paste. Output saved to ~/Downloads/final_v2_REAL.txt. That's when it clicked: tokens per second was never the point of this box. A 24 GB gaming card forces you to pick ONE model, so everyone picks the biggest generalist and asks it to do everything. But 128 GB doesn't force that. You can keep six specialists loaded at once. A 1.7B speech model transcribes better than a 35B chat model ever will. A tiny translation model just translates and doesn't get creative. The big LLM does the thinking, and only the thinking. The hardware was never the problem. The missing piece was the thing that ties them together. So I built it. SparklingKit. (the name is a tribute to the Spark. also: a spark is one flash, sparkling is when it keeps going) What it actually is: → transcription, OCR, translation, image grounding, text-to-image, mind maps, chat. one panel. → six models hand-picked to fit together in 128 GB with room left over → queue-first: drop a 3-hour recording, go get coffee, check from your phone → nothing is a dead end: finish a transcript and "translate / map / chat with it" are already waiting → visual workflows: chain the steps once, run them forever → everything is a file. one folder. no database. copy it to back up. point Obsidian at it. → nothing leaves your machine unless you send it. The thing I care about most: if you stop using SparklingKit tomorrow, nothing happens to your data. It was already a folder of markdown and json. There's no lock-in to escape. Spark owners: you bought room, not a faster chatbot. This is how I'm using the room. No Spark? The panel runs anywhere Docker runs. Point it at your own endpoints. It's open source, and it's version one. Rough edges exist. Issues and PRs very welcome. Full details and a one-line install at If you've ever uploaded something private to a cloud AI and felt weird about it, this is for you.

stevibe

55,853 görüntüleme • 27 gün önce

📢 #MillenZMovement ,An Open Letter to You, Paul Kagame & Presidency | Rwanda .Personally. We were born and raised under your regime. 🇷🇼 You fed us your version of history. You brainwashed us to believe that your truth was the only truth. We admired you. We sang your praises. We hung your portrait on our walls. I even had it in my own apartment. 🖼️ But what did you give us in return? 💀 Death. 😨 Fear. 💣 Destruction. You destroyed our parents’ homes. You silenced our voices. You made fear a way of life. 🔥 I want you to look closely, this is me burning your portrait. That’s how sick and done we are with you. Even the late artist #JayPolly said it, the youth have lived in sadness, speaking in coded words because we’re too afraid to die. 🎤💔 When you were in the #presidential election of 2017, I was still in S4. You came to campaign near our high school, it was the first time I saw you up close. You even shook my hand, and I was so excited then. 🤝✨ I thought you were a hero. Now I realize how wicked you truly are, the way you staged that drama election, waking up old women and children at 2 a.m. to wait for you at the site, just so the world could see fake love for a dictator. 🎭💔 #MillenZMovement has opened its eyes. 👁️ We see clearly now ,you wish us no good. All you care about is your family’s power and wealth. 💰👑 You kill those who speak up, and those who stay silent die anyway. ⚰️ We are no longer afraid. We refuse to kneel. 🙅🏾‍♂️ We are fighting for our right to speak, to think, to live freely, rights guaranteed by the Constitution itself. 📜✊🏾 Recently, you said at Unity Club that no one dares tell you to step down. Well, here we are ,telling you to step down. Now. 🚨 Because Rwanda is not a kingdom your father left you. 🛑👑 This country belongs to the people, and we are awake now. #MillenZMovement 🌍✊🏾 #TruthOverFear 💥 #RwandaDeservesBetter ❤️‍🔥
2:47

📢 #MillenZMovement ,An Open Letter to You, Paul Kagame & Presidency | Rwanda .Personally. We were born and raised under your regime. 🇷🇼 You fed us your version of history. You brainwashed us to believe that your truth was the only truth. We admired you. We sang your praises. We hung your portrait on our walls. I even had it in my own apartment. 🖼️ But what did you give us in return? 💀 Death. 😨 Fear. 💣 Destruction. You destroyed our parents’ homes. You silenced our voices. You made fear a way of life. 🔥 I want you to look closely, this is me burning your portrait. That’s how sick and done we are with you. Even the late artist #JayPolly said it, the youth have lived in sadness, speaking in coded words because we’re too afraid to die. 🎤💔 When you were in the #presidential election of 2017, I was still in S4. You came to campaign near our high school, it was the first time I saw you up close. You even shook my hand, and I was so excited then. 🤝✨ I thought you were a hero. Now I realize how wicked you truly are, the way you staged that drama election, waking up old women and children at 2 a.m. to wait for you at the site, just so the world could see fake love for a dictator. 🎭💔 #MillenZMovement has opened its eyes. 👁️ We see clearly now ,you wish us no good. All you care about is your family’s power and wealth. 💰👑 You kill those who speak up, and those who stay silent die anyway. ⚰️ We are no longer afraid. We refuse to kneel. 🙅🏾‍♂️ We are fighting for our right to speak, to think, to live freely, rights guaranteed by the Constitution itself. 📜✊🏾 Recently, you said at Unity Club that no one dares tell you to step down. Well, here we are ,telling you to step down. Now. 🚨 Because Rwanda is not a kingdom your father left you. 🛑👑 This country belongs to the people, and we are awake now. #MillenZMovement 🌍✊🏾 #TruthOverFear 💥 #RwandaDeservesBetter ❤️‍🔥

U.Sheila C kamuzinzi 🇷🇼🇸🇬

79,898 görüntüleme • 10 ay önce

✨ I'm getting very very close to the original vision I had for Photo AI when I started it 3 years ago which was essentially: [ 📸 Photography Simulator 2025 ] I wanted to 100% be able to simulate being a real photographer, and what that means is actually being able to move the camera around a subject I think I'm getting really close now by integrating World Labs, it lets me put real people into a 3d world that you can then move around in with WASD and your mouse Then my idea is you have a [ 📸 Take photo ] button there to turn that shot into a new photo with AI I'm not there yet but I'm getting close I think, one thing is that World Labs still suck at people and faces, although if I just ask the AI to take only the angle of the pic maybe that's not a problem so much I first used World Labs on Interior AI where it lets people move around in their re-styled interior design, I thought it'd be a gimmick maybe and few would use it, but now it's one of the most used features Anyway very exciting!!!! 😊
0:57

✨ I'm getting very very close to the original vision I had for Photo AI when I started it 3 years ago which was essentially: [ 📸 Photography Simulator 2025 ] I wanted to 100% be able to simulate being a real photographer, and what that means is actually being able to move the camera around a subject I think I'm getting really close now by integrating World Labs, it lets me put real people into a 3d world that you can then move around in with WASD and your mouse Then my idea is you have a [ 📸 Take photo ] button there to turn that shot into a new photo with AI I'm not there yet but I'm getting close I think, one thing is that World Labs still suck at people and faces, although if I just ask the AI to take only the angle of the pic maybe that's not a problem so much I first used World Labs on Interior AI where it lets people move around in their re-styled interior design, I thought it'd be a gimmick maybe and few would use it, but now it's one of the most used features Anyway very exciting!!!! 😊

@levelsio

423,523 görüntüleme • 9 ay önce

When this 18 year old girl pulled into a fight with her former best friend, she had no idea the battle would continue in the interrogation room after. A mutual fallout between two former best friends ended in a f@tal 20-second encounter outside a Columbus convenience store. 18-year-old Bryanna Barazzini and 19-year-old Halia Culbertson had a bitter falling out. Fueling the feud was complex bad blood involving child custody disputes with Bryanna’s brother, leading to six months of threats, insults, and harassment traded back and forth. ​The two crossed paths purely by chance at a carry-out market at Columbus Square. A verbal argument quickly erupted inside the store after Halia and her friend she was with started verbally attacking Bryanna. In return Bryanna allegedly called Halia a recovering addict and insulted her sobriety. The store clerk intervened and kicked Halia’s group outside first, keeping Bryanna inside to allow the crowd to leave. ​They didn't leave. Halia’s group waited in the parking lot. ​When Bryanna finally walked out, Halia aggressively advanced to start a fistfight. Bystander cell phone footage captured Bryanna retreating more than 30 steps backwards, repeatedly stating she didn't want to fight, while Halia continued to close the distance. However, right before the first punch was thrown, Bryanna pulled an open folding work kn!fe from her pocket and held it at her side. ​Within seconds of the clash, the scuffle broke up. The crowd assumed it was an ordinary fistfight until Halia said she didn't feel well back near her vehicle. She had been st@bbed in the n*ck. Halia was rushed to Riverside Methodist Hospital and passed 35 minutes later. ​When detectives brought Bryanna in for questioning hours later, she believed she was just recounting a routine fight where she was defending herself having no idea that Halia had anything happen to her. Only after walking through her version of events did detectives reveal Halia had passed, and she was booked for m*rder. ​The case presented damning evidence on both sides: For the Defense: Video evidence proved Halia laid in wait, initiated the confrontation, and ignored repeated warnings while Bryanna continually retreated. Bryanna was also confronted by multiple people associated with Halia, creating an immediate, hostile threat. For the State: Before a single blow landed, Bryanna deployed an open kn!fe against an unarmed opponent. Subsequent searches revealed the knife hidden in a crawl space, clothes in the wash, and prior phone messages from Bryanna warning that if Halia ever approached her again, she would sl*ce her thro@t. ​Instead of taking the risk before a jury, both sides agreed to a plea deal. The state dropped the murder charges, and Bryanna pleaded guilty to involuntary manslaughter with a 3-year prison sentence. She was granted early judicial release shortly thereafter and is now back home. ​Now it’s your turn to weigh the evidence. ​Given that Bryanna retreated over 30 steps from an aggressive ambush, was this a justified act of self-defense where charges shouldn't have been filed at all? Or did pulling an open kn!fe and previously texting threats make this a clear case of m*rder that deserved a full conviction? ​I'll be very interested to see the comments on this one.
17:33

When this 18 year old girl pulled into a fight with her former best friend, she had no idea the battle would continue in the interrogation room after. A mutual fallout between two former best friends ended in a f@tal 20-second encounter outside a Columbus convenience store. 18-year-old Bryanna Barazzini and 19-year-old Halia Culbertson had a bitter falling out. Fueling the feud was complex bad blood involving child custody disputes with Bryanna’s brother, leading to six months of threats, insults, and harassment traded back and forth. ​The two crossed paths purely by chance at a carry-out market at Columbus Square. A verbal argument quickly erupted inside the store after Halia and her friend she was with started verbally attacking Bryanna. In return Bryanna allegedly called Halia a recovering addict and insulted her sobriety. The store clerk intervened and kicked Halia’s group outside first, keeping Bryanna inside to allow the crowd to leave. ​They didn't leave. Halia’s group waited in the parking lot. ​When Bryanna finally walked out, Halia aggressively advanced to start a fistfight. Bystander cell phone footage captured Bryanna retreating more than 30 steps backwards, repeatedly stating she didn't want to fight, while Halia continued to close the distance. However, right before the first punch was thrown, Bryanna pulled an open folding work kn!fe from her pocket and held it at her side. ​Within seconds of the clash, the scuffle broke up. The crowd assumed it was an ordinary fistfight until Halia said she didn't feel well back near her vehicle. She had been st@bbed in the n*ck. Halia was rushed to Riverside Methodist Hospital and passed 35 minutes later. ​When detectives brought Bryanna in for questioning hours later, she believed she was just recounting a routine fight where she was defending herself having no idea that Halia had anything happen to her. Only after walking through her version of events did detectives reveal Halia had passed, and she was booked for m*rder. ​The case presented damning evidence on both sides: For the Defense: Video evidence proved Halia laid in wait, initiated the confrontation, and ignored repeated warnings while Bryanna continually retreated. Bryanna was also confronted by multiple people associated with Halia, creating an immediate, hostile threat. For the State: Before a single blow landed, Bryanna deployed an open kn!fe against an unarmed opponent. Subsequent searches revealed the knife hidden in a crawl space, clothes in the wash, and prior phone messages from Bryanna warning that if Halia ever approached her again, she would sl*ce her thro@t. ​Instead of taking the risk before a jury, both sides agreed to a plea deal. The state dropped the murder charges, and Bryanna pleaded guilty to involuntary manslaughter with a 3-year prison sentence. She was granted early judicial release shortly thereafter and is now back home. ​Now it’s your turn to weigh the evidence. ​Given that Bryanna retreated over 30 steps from an aggressive ambush, was this a justified act of self-defense where charges shouldn't have been filed at all? Or did pulling an open kn!fe and previously texting threats make this a clear case of m*rder that deserved a full conviction? ​I'll be very interested to see the comments on this one.

Giggling Gannon

223,270 görüntüleme • 8 gün önce

Open Letter to nelson chamisa “If you do something once, it is a decision. If you do it twice, it becomes a habit. But when you do it more than three times, it becomes character.” Dear Nelson, I have known you for 31 years and have always — and still do — consider you a close friend and brother. Back then, there was no MDC opposition party to speak of, no Nehanda Radio, and none of the political prominence that surrounds us today. We were simply ordinary students at Harare Polytechnic trying to find our way in life. Over the years, you have done exceptionally well in rising through the ranks of leadership — from Youth Leader, to Spokesperson, Organising Secretary, ICT Minister during the GNU, Vice President, and eventually President of Zimbabwe’s main opposition movement. Throughout this journey, we developed a clear understanding of professional boundaries and mutual respect. I own and run a media platform that gives space to voices that may either criticise or praise you in equal measure. I have always appreciated the fact that you understood and respected this principle, and never sought to censor our work. In return, I have respected our friendship and never abused our proximity for personal or journalistic advantage. It is therefore deeply disheartening that, 48 hours after a defamatory article was published by Simba Chikanza on ZimEye falsely claiming that I recorded your conversation regarding Advocate Thabani Mpofu (Thabani Mpofu), you have chosen to remain silent and avoid setting the record straight. You are the only person who knows who was on the other side of that phone call. While you have every right to protect the identity of that individual, it is unfair to remain silent while false accusations are being levelled against a long-time friend and colleague. Unfortunately, your silence — as has happened in other matters which I will not get into here — creates a vacuum that people inevitably fill with assumptions and speculation. One cannot entirely blame those who may conclude that your silence suggests some level of complicity in the breeding of these claims. I have engaged you privately and encouraged you to do the right thing because, apart from the interviewer, you are the only other person capable of authenticating the recording and confirming that the person you were speaking to was clearly not me. Yet despite this, you have chosen not to correct the record. That is troubling — not just politically, but personally. Over the years, Simba Chikanza has made defamatory allegations against several people associated with your political space, including Fadzayi Mahere🇿🇼, Pedzisai Ruhanya, Hopewell Chin'ono (Hopewell Chin’ono), Jameson Z. Timba and many others. Allowing such falsehoods to continue unchecked has inevitably led many to question whether these attacks are random, or whether there is active encouragement behind the scenes regarding who should be targeted next. There are currently two separate audio controversies being discussed publicly. The first was the Daily News recording in which you were heard criticising your colleague Jameson Timba. The second recording (attached here for purposes of clarity) — the one now falsely linked to me — was not released by the Daily News. It was allegedly leaked by the very person whose identity you appear resolutely determined to protect. That naturally raises difficult questions about why you are prepared to protect that individual at the expense of long-standing friends and comrades. This situation mirrors closely that of another dear friend of mine, Luke Tamborinyoka Luke Tamborinyoka). He gave you years of dedication and service, often at great personal sacrifice, yet in 2023 you quietly imposed another candidate in Goromonzi West after Luke had emerged victorious in the much-criticised “Bereka Mwana” voting exercise. Luke narrated his painful experience in an open letter to Domboshava residents, circulated in polling-station-based WhatsApp groups in Domboshava, which we later published on the Nehanda Radio website in July 2023. One could conclude you have a perchant for sabotaging your own friends and comrades, even when they would not in any way have sought any benevolence from you. While your friends and comrades are left exposed, you appear comfortable briefing and confiding in a publisher whom many Zimbabweans consider deeply reckless and unstable. The day you eventually fall out with him, you may come to appreciate the importance of choosing and screening confidants carefully. Should he one day find himself before a court of law (given the many people he has defamed) he may be required to substantiate his allegations and there is every chance he will seek to protect himself by pointing in your direction. You have been warned repeatedly by many people about his conduct and disposition, yet it increasingly appears that you may have found utility in that recklessness because it can easily be weaponised politically against both friends and perceived opponents alike. I could easily have used this letter in anger to retaliate or reveal matters shared in confidence, but I have deliberately chosen not to do so because I still believe in conducting myself professionally and responsibly for the greater good. In 31 years, our phone calls must number in the thousands, and not a single one has ever been recorded by me — something you know better than anyone else. If you ultimately choose not to correct the record, so be it. I am more than capable of defending myself publicly and politically. I am not particularly concerned about what Simba Chikanza writes because, quite frankly, very few serious Zimbabweans consider him credible. My greater concern is your apparent unwillingness to tell the truth on a matter that should be straightforward. What irks me the most is that I never expected this from a friend, a comrade, and a Pastor. God is definitely not in this. Lance Guma Managing Editor Nehanda Radio
1:03

Open Letter to nelson chamisa “If you do something once, it is a decision. If you do it twice, it becomes a habit. But when you do it more than three times, it becomes character.” Dear Nelson, I have known you for 31 years and have always — and still do — consider you a close friend and brother. Back then, there was no MDC opposition party to speak of, no Nehanda Radio, and none of the political prominence that surrounds us today. We were simply ordinary students at Harare Polytechnic trying to find our way in life. Over the years, you have done exceptionally well in rising through the ranks of leadership — from Youth Leader, to Spokesperson, Organising Secretary, ICT Minister during the GNU, Vice President, and eventually President of Zimbabwe’s main opposition movement. Throughout this journey, we developed a clear understanding of professional boundaries and mutual respect. I own and run a media platform that gives space to voices that may either criticise or praise you in equal measure. I have always appreciated the fact that you understood and respected this principle, and never sought to censor our work. In return, I have respected our friendship and never abused our proximity for personal or journalistic advantage. It is therefore deeply disheartening that, 48 hours after a defamatory article was published by Simba Chikanza on ZimEye falsely claiming that I recorded your conversation regarding Advocate Thabani Mpofu (Thabani Mpofu), you have chosen to remain silent and avoid setting the record straight. You are the only person who knows who was on the other side of that phone call. While you have every right to protect the identity of that individual, it is unfair to remain silent while false accusations are being levelled against a long-time friend and colleague. Unfortunately, your silence — as has happened in other matters which I will not get into here — creates a vacuum that people inevitably fill with assumptions and speculation. One cannot entirely blame those who may conclude that your silence suggests some level of complicity in the breeding of these claims. I have engaged you privately and encouraged you to do the right thing because, apart from the interviewer, you are the only other person capable of authenticating the recording and confirming that the person you were speaking to was clearly not me. Yet despite this, you have chosen not to correct the record. That is troubling — not just politically, but personally. Over the years, Simba Chikanza has made defamatory allegations against several people associated with your political space, including Fadzayi Mahere🇿🇼, Pedzisai Ruhanya, Hopewell Chin'ono (Hopewell Chin’ono), Jameson Z. Timba and many others. Allowing such falsehoods to continue unchecked has inevitably led many to question whether these attacks are random, or whether there is active encouragement behind the scenes regarding who should be targeted next. There are currently two separate audio controversies being discussed publicly. The first was the Daily News recording in which you were heard criticising your colleague Jameson Timba. The second recording (attached here for purposes of clarity) — the one now falsely linked to me — was not released by the Daily News. It was allegedly leaked by the very person whose identity you appear resolutely determined to protect. That naturally raises difficult questions about why you are prepared to protect that individual at the expense of long-standing friends and comrades. This situation mirrors closely that of another dear friend of mine, Luke Tamborinyoka Luke Tamborinyoka). He gave you years of dedication and service, often at great personal sacrifice, yet in 2023 you quietly imposed another candidate in Goromonzi West after Luke had emerged victorious in the much-criticised “Bereka Mwana” voting exercise. Luke narrated his painful experience in an open letter to Domboshava residents, circulated in polling-station-based WhatsApp groups in Domboshava, which we later published on the Nehanda Radio website in July 2023. One could conclude you have a perchant for sabotaging your own friends and comrades, even when they would not in any way have sought any benevolence from you. While your friends and comrades are left exposed, you appear comfortable briefing and confiding in a publisher whom many Zimbabweans consider deeply reckless and unstable. The day you eventually fall out with him, you may come to appreciate the importance of choosing and screening confidants carefully. Should he one day find himself before a court of law (given the many people he has defamed) he may be required to substantiate his allegations and there is every chance he will seek to protect himself by pointing in your direction. You have been warned repeatedly by many people about his conduct and disposition, yet it increasingly appears that you may have found utility in that recklessness because it can easily be weaponised politically against both friends and perceived opponents alike. I could easily have used this letter in anger to retaliate or reveal matters shared in confidence, but I have deliberately chosen not to do so because I still believe in conducting myself professionally and responsibly for the greater good. In 31 years, our phone calls must number in the thousands, and not a single one has ever been recorded by me — something you know better than anyone else. If you ultimately choose not to correct the record, so be it. I am more than capable of defending myself publicly and politically. I am not particularly concerned about what Simba Chikanza writes because, quite frankly, very few serious Zimbabweans consider him credible. My greater concern is your apparent unwillingness to tell the truth on a matter that should be straightforward. What irks me the most is that I never expected this from a friend, a comrade, and a Pastor. God is definitely not in this. Lance Guma Managing Editor Nehanda Radio

Nehanda Radio

73,523 görüntüleme • 4 ay önce

Introducing the Content Machine! This was the first time I walked through the mechanics of our anti-slop content system & how we drive 10,000,000+ impressions at tenex with a marketing team of...2. Thanks to claire vo 🖤 for having me on her show to share. Full-writeup & interview below... What is the Content Machine? A directory of daisy-chained skills that turn what you already say into publish-ready content. This is the key way to avoid turning into a slop cannon. It mines the places you already talk (Slack, Notion, Gmail, Linear, GitHub) plus what the internet is saying, finds the ideas worth writing, interviews you to extract the story, drafts it, and edits it to a 9/10 bar before you post. What are the principles of this system? 1) It is not a "write me a post" prompt. The core belief baked into it: the raw material must come from you. 2) The machine never invents your voice and never fabricates your insight. It does the research, the structure, and the editing, so your time goes only to the part only you can do. It focuses the human on the first & final mile of the content process. How is it structured? A two-layer split between the process & the person - The process layer is generic and shared: the pipeline, the content-type specs, the copywriting references, the onboarding flow. That is what lives in git and what gets shipped to other people. - The personal layer is yours and never leaves your machine: content-machine.config.md, creators/ / (profile, style guide, lessons), projects/, published/, oracle-reports/. All gitignored, and the desktop build script refuses to package any of it. What is the 10-step pipeline? 1) Creator Select. Multi-creator by design. It figures out who this run is for and loads their profile, style guide, and content lessons. Everything downstream is scoped to that person. - Onboarding (first run only). Scaffolds the workspace, auto-creates your Notion Vault, connects your sources, and builds your voice guide one of three ways: import a guide you already have, feed it writing samples, or sit for a short voice interview. 2) The Oracle. Two idea engines running in parallel: - Oracle scans what you wrote in the last 7 days across Slack, Notion, Gmail, Linear, and Git, hunting for "spikes," moments where you naturally said something worth expanding. - The Internet Reader scans what the world is saying, pulled only from the source list in your profile (handles, labs, outlets, keyword watchlist), plus a social sweep across Reddit, X, YouTube, Hacker News, and more (thanks /last30days & Matt Van Horn). - Every idea is scored 0 to 10 (POV strength 25%, story potential 25%, emotional intensity 20%, lesson/framework 20%, depth 10%). Everything qualifying gets written to The Vault, a Notion database that is the durable idea bank. 2.5) Research. Before you get interviewed, a research agent builds a sourced brief: - key facts with links - current developments - what has already been said in-market - contrarian angles - open questions only you can answer. 3) Interview Panel. Six interviewer personas (Ferriss, Rogan, Larry King, Stern, Barbaro, Barbara Walters) ask you one question at a time, each chasing a different dimension: tactics, story, core truth, the hidden thing, clarity, emotional depth. It pushes back on vague answers and will not advance until it has 2 to 3 specific stories with real details. 4) Production. Your interview becomes a raw markdown file: stories, core insights, quotable moments, the emotional anchor, surprising reveals, the "so what." Your exact words are preserved. This file is the source of truth for everything that follows. 5) Refinement. Now it drafts, and only now. It must read your style guide, your content lessons, and the spec for the chosen format (LinkedIn post, X thread, long post, playbook, podcast promo, reaction post, article, and so on). 6) Writer's Council. Six reviewers score the draft: Morgan Housel (will this matter in 10 years), Tim Urban (is it confusing), Shaan Puri (would I stop scrolling, plus three alternate hooks), Greg Isenberg (what can someone steal), David Perell (is it personal, observational, playful), and a Slop Detector hunting AI tells. Each gives what's working, what needs work, a fix, and a score. 7) Revision Loop. Under 9/10 goes back around. The smart part: fixes get sorted into editorial (the machine rewrites it itself) and information gaps (only you have the answer), and information gaps route back to the Interview Panel with targeted questions rather than letting the machine make something up. Max 3 editorial cycles. At 9/10 the piece becomes the anchor. 8) Repurposing Engine. One anchor fans out into 10+ derivatives: X article, LinkedIn article, short X posts, short LinkedIn posts, a playbook if there is a framework in it. Each one is written native to its platform with a fresh hook, not cross-posted, and each runs the full council and revision loop to 9/10 on its own. This is the multiplier. 9) Distribution (optional, off by default). UTM tagging, a scheduled publishing queue, CRM capture of every touchpoint, attribution reporting back to pipeline, and marking the Vault row as Published. 10) The Learning Loop, always running. After you approve a piece it diffs your first draft against the final, extracts the pattern, and asks you to confirm it. Confirmed lessons go into content-lessons.md and override the style guide. Once a lesson proves out across a few projects it graduates into the style guide itself. Your first drafts get better over time instead of you re-explaining preferences. P.S. i'm thinking about opensourcing this. should i do it?
42:58

Introducing the Content Machine! This was the first time I walked through the mechanics of our anti-slop content system & how we drive 10,000,000+ impressions at tenex with a marketing team of...2. Thanks to claire vo 🖤 for having me on her show to share. Full-writeup & interview below... What is the Content Machine? A directory of daisy-chained skills that turn what you already say into publish-ready content. This is the key way to avoid turning into a slop cannon. It mines the places you already talk (Slack, Notion, Gmail, Linear, GitHub) plus what the internet is saying, finds the ideas worth writing, interviews you to extract the story, drafts it, and edits it to a 9/10 bar before you post. What are the principles of this system? 1) It is not a "write me a post" prompt. The core belief baked into it: the raw material must come from you. 2) The machine never invents your voice and never fabricates your insight. It does the research, the structure, and the editing, so your time goes only to the part only you can do. It focuses the human on the first & final mile of the content process. How is it structured? A two-layer split between the process & the person - The process layer is generic and shared: the pipeline, the content-type specs, the copywriting references, the onboarding flow. That is what lives in git and what gets shipped to other people. - The personal layer is yours and never leaves your machine: content-machine.config.md, creators/ / (profile, style guide, lessons), projects/, published/, oracle-reports/. All gitignored, and the desktop build script refuses to package any of it. What is the 10-step pipeline? 1) Creator Select. Multi-creator by design. It figures out who this run is for and loads their profile, style guide, and content lessons. Everything downstream is scoped to that person. - Onboarding (first run only). Scaffolds the workspace, auto-creates your Notion Vault, connects your sources, and builds your voice guide one of three ways: import a guide you already have, feed it writing samples, or sit for a short voice interview. 2) The Oracle. Two idea engines running in parallel: - Oracle scans what you wrote in the last 7 days across Slack, Notion, Gmail, Linear, and Git, hunting for "spikes," moments where you naturally said something worth expanding. - The Internet Reader scans what the world is saying, pulled only from the source list in your profile (handles, labs, outlets, keyword watchlist), plus a social sweep across Reddit, X, YouTube, Hacker News, and more (thanks /last30days & Matt Van Horn). - Every idea is scored 0 to 10 (POV strength 25%, story potential 25%, emotional intensity 20%, lesson/framework 20%, depth 10%). Everything qualifying gets written to The Vault, a Notion database that is the durable idea bank. 2.5) Research. Before you get interviewed, a research agent builds a sourced brief: - key facts with links - current developments - what has already been said in-market - contrarian angles - open questions only you can answer. 3) Interview Panel. Six interviewer personas (Ferriss, Rogan, Larry King, Stern, Barbaro, Barbara Walters) ask you one question at a time, each chasing a different dimension: tactics, story, core truth, the hidden thing, clarity, emotional depth. It pushes back on vague answers and will not advance until it has 2 to 3 specific stories with real details. 4) Production. Your interview becomes a raw markdown file: stories, core insights, quotable moments, the emotional anchor, surprising reveals, the "so what." Your exact words are preserved. This file is the source of truth for everything that follows. 5) Refinement. Now it drafts, and only now. It must read your style guide, your content lessons, and the spec for the chosen format (LinkedIn post, X thread, long post, playbook, podcast promo, reaction post, article, and so on). 6) Writer's Council. Six reviewers score the draft: Morgan Housel (will this matter in 10 years), Tim Urban (is it confusing), Shaan Puri (would I stop scrolling, plus three alternate hooks), Greg Isenberg (what can someone steal), David Perell (is it personal, observational, playful), and a Slop Detector hunting AI tells. Each gives what's working, what needs work, a fix, and a score. 7) Revision Loop. Under 9/10 goes back around. The smart part: fixes get sorted into editorial (the machine rewrites it itself) and information gaps (only you have the answer), and information gaps route back to the Interview Panel with targeted questions rather than letting the machine make something up. Max 3 editorial cycles. At 9/10 the piece becomes the anchor. 8) Repurposing Engine. One anchor fans out into 10+ derivatives: X article, LinkedIn article, short X posts, short LinkedIn posts, a playbook if there is a framework in it. Each one is written native to its platform with a fresh hook, not cross-posted, and each runs the full council and revision loop to 9/10 on its own. This is the multiplier. 9) Distribution (optional, off by default). UTM tagging, a scheduled publishing queue, CRM capture of every touchpoint, attribution reporting back to pipeline, and marking the Vault row as Published. 10) The Learning Loop, always running. After you approve a piece it diffs your first draft against the final, extracts the pattern, and asks you to confirm it. Confirmed lessons go into content-lessons.md and override the style guide. Once a lesson proves out across a few projects it graduates into the style guide itself. Your first drafts get better over time instead of you re-explaining preferences. P.S. i'm thinking about opensourcing this. should i do it?

Alex Lieberman

115,942 görüntüleme • 2 ay önce

Abode is live. A while ago I started exploring social with the mindset of creating something fun. I ended up creating something deeply meaningful for myself. When I started Abode it was out of frustration. How have we moved to a state of social that cares more about how you’re perceived than connection. Growing up from when social was all about connection I often became frustrated that group chats are all that remain for that space. Don’t get me wrong it’s not all bad. A lot of great things have come out of this era of social. I just always wished for more than a chat for my close friends and family. I wanted to create a space that felt as unique as each of my friendship groups. One that wasn’t restricted to a chat thread or functionality, I always felt that chat was this weird state of social where you’re waiting on a response, it didn’t allow for some of the best moments which are often spontaneous. Chats felt like more of a chore to keep up with, than something I wanted to be part of. The result of this was Abode. a hyper personal, create your own social, blank canvas. it allows you to add unique functionality through the form of widgets (which we call magnets). This allows you to have a unique experience depending on how you build your space. More games? More organisation? More social? The beauty of this is it’s up to you. This resulted in what I’ve ended up calling social through presence vs social through action. I’d be playing a real life game with a friend and update our score note in the Abode and other friends jump in and congratulate on the win. No message was needed, no context was given. Social just happened. I’ve seen friends who wouldn’t normally go out of their way to message each other play gruelling games of connect4 with each other because they saw someone was online and playing. It’s this idea of someone else’s presence triggering a natural hangout that’s low pressure. Not someone messaging, asking if you’re around and waiting for a response. It’s a blend between async and real time that just works. Right now it’s very fresh, we have a very limited number of widgets as we experiment with the concept and figure out what works. The idea is to open up our widget engine to allow other people’s creativity to shine and build as part of the abode ecosystem. - group space - live widgets - themes - real time audio chat - messaging - AI powered summaries for easy catch up - Automatic memory building Not all of these features may stay but I wanted to build a space that supported all you’d need and work backwards by using the popularity of features to decide a direction that best suited the majority. Abode is yours, and it should feel that way. Appreciate any sharing 🫶
0:40

Abode is live. A while ago I started exploring social with the mindset of creating something fun. I ended up creating something deeply meaningful for myself. When I started Abode it was out of frustration. How have we moved to a state of social that cares more about how you’re perceived than connection. Growing up from when social was all about connection I often became frustrated that group chats are all that remain for that space. Don’t get me wrong it’s not all bad. A lot of great things have come out of this era of social. I just always wished for more than a chat for my close friends and family. I wanted to create a space that felt as unique as each of my friendship groups. One that wasn’t restricted to a chat thread or functionality, I always felt that chat was this weird state of social where you’re waiting on a response, it didn’t allow for some of the best moments which are often spontaneous. Chats felt like more of a chore to keep up with, than something I wanted to be part of. The result of this was Abode. a hyper personal, create your own social, blank canvas. it allows you to add unique functionality through the form of widgets (which we call magnets). This allows you to have a unique experience depending on how you build your space. More games? More organisation? More social? The beauty of this is it’s up to you. This resulted in what I’ve ended up calling social through presence vs social through action. I’d be playing a real life game with a friend and update our score note in the Abode and other friends jump in and congratulate on the win. No message was needed, no context was given. Social just happened. I’ve seen friends who wouldn’t normally go out of their way to message each other play gruelling games of connect4 with each other because they saw someone was online and playing. It’s this idea of someone else’s presence triggering a natural hangout that’s low pressure. Not someone messaging, asking if you’re around and waiting for a response. It’s a blend between async and real time that just works. Right now it’s very fresh, we have a very limited number of widgets as we experiment with the concept and figure out what works. The idea is to open up our widget engine to allow other people’s creativity to shine and build as part of the abode ecosystem. - group space - live widgets - themes - real time audio chat - messaging - AI powered summaries for easy catch up - Automatic memory building Not all of these features may stay but I wanted to build a space that supported all you’d need and work backwards by using the popularity of features to decide a direction that best suited the majority. Abode is yours, and it should feel that way. Appreciate any sharing 🫶

Joseph Smith

33,292 görüntüleme • 1 yıl önce

1/3 You may see clips of Trump’s self-evident dementia, but the full effect requires stamina. Here is NINE MINUTES of Trump talking about his fountains in a Cabinet meeting. Try to read it. I dare you. “D.C. and D.C. is looking beautiful. The fountains are almost all open. We had 28 of them, and we have one in particular, a very long lake, we call it. They're reflecting lake between the Lincoln Monument. Nobody's ever seen anything like it. You take a look at between Lincoln and Washington. You have the longest, like 2,400 feet long. That's longer than the tallest building in the world, if you set it on your side. And it's almost 200 feet wide. And for-- actually, it was built in 1922. And from 1920 to '21, it really never worked. It always leaked. And it was a problem about things. And hundreds of millions was spent. The Biden administration and the Obama administration spent hundreds of millions of dollars trying to get it to work, and they failed. And we'll be spending-- you'll give me a number. But I think it's very low numbers, like in the number that we originally talked about. Where we cleaned it, we fumigated it, that we had 10 major truck dumpsters of garbage taken out. Can you believe it? Friend of mine came in. Very substantial person from Germany. Want to see the Washington Monument. I wanted to see the Lincoln Memorial. He said, I look, but that horrible reflecting pond is disgusting. It's filthy dirty and disgusting. I said, really, I drove down. I said, secret shares was taken dry. And we went down. I said, that's terrible. And for the most part, it didn't work. I mean, they wouldn't even have water in there. But when they did, it was just dirty, filthy water that leaked out. And we got to work on it. And they were supposed to cost almost 400, think of it, 400 million dollars, because it's like putting the skin on a skyscraper. But bigger, much bigger, many skyscrapers, you could almost say, other than for the world. So I think of it, the world's tallest skyscraper is shorter than 2,400 feet. So we went to work. And over the years, I built hundreds of pools. I built them every time I built them. I always like to build the Olympic size swimming pools. And I was very aware of swimming pool, what goes into making a swimming pool. It's not as simple as people think. You never wanted to leak. You want a beautiful surface. I said, you know, I have an idea, Doug. We sat down. I said, let's take that long thing where we're going to fix it with concrete, which leaks, concrete, all sorts of other materials that all leak and don't look good. They're gray. And swimming pool, I have a swimming pool right up the road. I built it 22 years ago. It's perfect. I said, does it ever leak, though? It's good contractors. I actually called it a couple of the contractors, got some ideas from them. And I gave it to Doug and Doug's done an unbelievable job. And so they were going to spend maybe 400 million dollars. I don't know. You never know with cost overruns and everything. But it was going to take years to build three years, four years. And we'll have it done before July 4th substantially, before July 4th. If we didn't have such a bad weather the last four or five days, it could have been almost done. We have to-- you can't do this substance in the rain. And what they did is they cleaned it. They took out, as I said, more than 10 dumpsters of garbage. Shovels, it was disgusting. Well, every corner, every corner, had massive amounts of-- I guess that's the way the tide goes. But at massive amounts of garbage, we then steamed clean it. We then sandblasted it. And then we pebble blasted. That's a bigger version of sand. We made the surface as good as it can be. Now we're now covering it with the most beautiful blue. Very thick, you think of it as a very sophisticated form of rubber. No leaks, no problems. And it's beautiful. It's called American Flag Blue that was the color we chose.”
8:50

1/3 You may see clips of Trump’s self-evident dementia, but the full effect requires stamina. Here is NINE MINUTES of Trump talking about his fountains in a Cabinet meeting. Try to read it. I dare you. “D.C. and D.C. is looking beautiful. The fountains are almost all open. We had 28 of them, and we have one in particular, a very long lake, we call it. They're reflecting lake between the Lincoln Monument. Nobody's ever seen anything like it. You take a look at between Lincoln and Washington. You have the longest, like 2,400 feet long. That's longer than the tallest building in the world, if you set it on your side. And it's almost 200 feet wide. And for-- actually, it was built in 1922. And from 1920 to '21, it really never worked. It always leaked. And it was a problem about things. And hundreds of millions was spent. The Biden administration and the Obama administration spent hundreds of millions of dollars trying to get it to work, and they failed. And we'll be spending-- you'll give me a number. But I think it's very low numbers, like in the number that we originally talked about. Where we cleaned it, we fumigated it, that we had 10 major truck dumpsters of garbage taken out. Can you believe it? Friend of mine came in. Very substantial person from Germany. Want to see the Washington Monument. I wanted to see the Lincoln Memorial. He said, I look, but that horrible reflecting pond is disgusting. It's filthy dirty and disgusting. I said, really, I drove down. I said, secret shares was taken dry. And we went down. I said, that's terrible. And for the most part, it didn't work. I mean, they wouldn't even have water in there. But when they did, it was just dirty, filthy water that leaked out. And we got to work on it. And they were supposed to cost almost 400, think of it, 400 million dollars, because it's like putting the skin on a skyscraper. But bigger, much bigger, many skyscrapers, you could almost say, other than for the world. So I think of it, the world's tallest skyscraper is shorter than 2,400 feet. So we went to work. And over the years, I built hundreds of pools. I built them every time I built them. I always like to build the Olympic size swimming pools. And I was very aware of swimming pool, what goes into making a swimming pool. It's not as simple as people think. You never wanted to leak. You want a beautiful surface. I said, you know, I have an idea, Doug. We sat down. I said, let's take that long thing where we're going to fix it with concrete, which leaks, concrete, all sorts of other materials that all leak and don't look good. They're gray. And swimming pool, I have a swimming pool right up the road. I built it 22 years ago. It's perfect. I said, does it ever leak, though? It's good contractors. I actually called it a couple of the contractors, got some ideas from them. And I gave it to Doug and Doug's done an unbelievable job. And so they were going to spend maybe 400 million dollars. I don't know. You never know with cost overruns and everything. But it was going to take years to build three years, four years. And we'll have it done before July 4th substantially, before July 4th. If we didn't have such a bad weather the last four or five days, it could have been almost done. We have to-- you can't do this substance in the rain. And what they did is they cleaned it. They took out, as I said, more than 10 dumpsters of garbage. Shovels, it was disgusting. Well, every corner, every corner, had massive amounts of-- I guess that's the way the tide goes. But at massive amounts of garbage, we then steamed clean it. We then sandblasted it. And then we pebble blasted. That's a bigger version of sand. We made the surface as good as it can be. Now we're now covering it with the most beautiful blue. Very thick, you think of it as a very sophisticated form of rubber. No leaks, no problems. And it's beautiful. It's called American Flag Blue that was the color we chose.”

Jim Stewartson, Decelerationist 🇨🇦🇺🇦🇺🇸

410,247 görüntüleme • 4 ay önce

Wolverine's difficulty does not stem from a complex combat system, but rather from the placement, variety, and quantity of the enemies that you will face. One of the more difficult combat trials that I experienced had you facing a group of enemies comprised of a large dude, multiple basic enemies, two energy shield enemies, and a few more guys. Because this was a challenge map, the goal was to take out the big guy first, and there was a timer that I had to beat. Facing these guys head on was essentially guaranteed failure, cuz they would either overwhelm me immediately, or by the time I killed them all the timer would run out. You had to sneak around them and avoid the mutant detection alarms to sneak attack the large berserker and remove half his health, and then immediately begin open combat. Now think of this like Batman Arkham, where shield enemies require a special combo to beat, knife enemies can't be countered, and large guys have to be stunned with your cape before being attacked... Same concept here, but much more focused on dodging at the right time, parrying at the right time, and using your special attacks at the right moments. The shield guys HAD to be kicked before slashing him, the gun holders HAD to be dodged to avoid damage, and the sword guys HAD to be parried, all while the large dude is on your ass doing area effect attacks that can 2-3 hit you. These are all from the Trask soldiers faction btw, and there's of course "The Hand" faction of enemies, and more that they haven't revealed yet. This is obviously just ONE example, and I'm not even trying to say it's the most revolutionary combat system ever, because even I still need to try more of the combat to be fully sold on it, but I just wanted to give a better image to y'all of what combat will feel like once you're no longer in the tutorial stage of the game. I also didn't LOVE the sentinel boss fight, but it's hard to give proper criticism to this game right now without seeming like an absolute chud thanks to the current discourse lmao
0:57

Wolverine's difficulty does not stem from a complex combat system, but rather from the placement, variety, and quantity of the enemies that you will face. One of the more difficult combat trials that I experienced had you facing a group of enemies comprised of a large dude, multiple basic enemies, two energy shield enemies, and a few more guys. Because this was a challenge map, the goal was to take out the big guy first, and there was a timer that I had to beat. Facing these guys head on was essentially guaranteed failure, cuz they would either overwhelm me immediately, or by the time I killed them all the timer would run out. You had to sneak around them and avoid the mutant detection alarms to sneak attack the large berserker and remove half his health, and then immediately begin open combat. Now think of this like Batman Arkham, where shield enemies require a special combo to beat, knife enemies can't be countered, and large guys have to be stunned with your cape before being attacked... Same concept here, but much more focused on dodging at the right time, parrying at the right time, and using your special attacks at the right moments. The shield guys HAD to be kicked before slashing him, the gun holders HAD to be dodged to avoid damage, and the sword guys HAD to be parried, all while the large dude is on your ass doing area effect attacks that can 2-3 hit you. These are all from the Trask soldiers faction btw, and there's of course "The Hand" faction of enemies, and more that they haven't revealed yet. This is obviously just ONE example, and I'm not even trying to say it's the most revolutionary combat system ever, because even I still need to try more of the combat to be fully sold on it, but I just wanted to give a better image to y'all of what combat will feel like once you're no longer in the tutorial stage of the game. I also didn't LOVE the sentinel boss fight, but it's hard to give proper criticism to this game right now without seeming like an absolute chud thanks to the current discourse lmao

Farzam

171,524 görüntüleme • 1 ay önce

Using Claude Fable 5, I built a model that predicts the entire 2026 FIFA world cup.. every single game, not just the final.. so let me break the whole thing down. what it does, how it works, and exactly how i built it.. #1 First what it does: it predicts all 104 games of the tournament. not just who lifts the trophy, but every group match, every knockout, the full path from the round of 32 to the final.. everything lands in one dashboard: > group stage, every match with each team's win % and the chance of a draw > standings, how all 12 groups are projected to finish > bracket, the full knockout tree with each team's odds of advancing > champion odds, who's most likely to actually win it all and it doesn't freeze after one prediction. the moment a real game is played, it locks that result in and re-runs everything around it. so the odds move live as the tournament goes, week by week you watch favorites rise and contenders collapse. #2. How it works: the core idea is simple. the model only ever predicts one thing, a single match. the real trick is the repetition. it learns from decades of match history, then plays the whole tournament out from the first game to the final, tens of thousands of times. each run it records who advanced and who won. do that enough and you stop getting one guess and start getting real odds, one team lifts the trophy in maybe 14% of the runs, another in 9%, and so on. #3. So, how i built it ? i didn't hand-write most of the code. i broke the project into 4 pieces, described each one to fable, and let it build while i focused on getting the football logic exactly right. - The data every international match going back over a century, around 50,000 games, plus each team's elo rating, which is the truest measure of strength, and the official 2026 schedule. garbage data means garbage predictions, so this part mattered most. - The features i turned that raw history into signals the model can learn from, the elo gap between the two teams, recent form, goals scored and conceded, and a home boost for the hosts, usa, canada and mexico. - The model for each match it predicts the expected goals for both sides, then turns that into win, draw and loss probabilities plus a likely scoreline. that's what feeds the simulation. - The tournament engine this was the hard part. the 2026 world cup is brand new, 48 teams, 12 groups, a round of 32 that's never existed before, and 8 "best third-placed" teams that slot into the bracket by a fixed fifa table. even the group tiebreakers changed this year, head to head now counts before goal difference. get any of it wrong and the whole bracket falls apart, so i built it carefully and tested the format until it was exact, then wrapped it in a simulation loop that plays the tournament out tens of thousands of times. and the last piece, the live part. as real results come in, they get locked, and only the unplayed games get re-simulated. that's what makes it a living model instead of a one-time prediction. all of it outputs to a clean dashboard you can actually read and screenshot.. right now, before kickoff, it already has a clear favorite to lift the trophy.. 👀 btw who's your pick to win the 2026 world cup?
0:20

Using Claude Fable 5, I built a model that predicts the entire 2026 FIFA world cup.. every single game, not just the final.. so let me break the whole thing down. what it does, how it works, and exactly how i built it.. #1 First what it does: it predicts all 104 games of the tournament. not just who lifts the trophy, but every group match, every knockout, the full path from the round of 32 to the final.. everything lands in one dashboard: > group stage, every match with each team's win % and the chance of a draw > standings, how all 12 groups are projected to finish > bracket, the full knockout tree with each team's odds of advancing > champion odds, who's most likely to actually win it all and it doesn't freeze after one prediction. the moment a real game is played, it locks that result in and re-runs everything around it. so the odds move live as the tournament goes, week by week you watch favorites rise and contenders collapse. #2. How it works: the core idea is simple. the model only ever predicts one thing, a single match. the real trick is the repetition. it learns from decades of match history, then plays the whole tournament out from the first game to the final, tens of thousands of times. each run it records who advanced and who won. do that enough and you stop getting one guess and start getting real odds, one team lifts the trophy in maybe 14% of the runs, another in 9%, and so on. #3. So, how i built it ? i didn't hand-write most of the code. i broke the project into 4 pieces, described each one to fable, and let it build while i focused on getting the football logic exactly right. - The data every international match going back over a century, around 50,000 games, plus each team's elo rating, which is the truest measure of strength, and the official 2026 schedule. garbage data means garbage predictions, so this part mattered most. - The features i turned that raw history into signals the model can learn from, the elo gap between the two teams, recent form, goals scored and conceded, and a home boost for the hosts, usa, canada and mexico. - The model for each match it predicts the expected goals for both sides, then turns that into win, draw and loss probabilities plus a likely scoreline. that's what feeds the simulation. - The tournament engine this was the hard part. the 2026 world cup is brand new, 48 teams, 12 groups, a round of 32 that's never existed before, and 8 "best third-placed" teams that slot into the bracket by a fixed fifa table. even the group tiebreakers changed this year, head to head now counts before goal difference. get any of it wrong and the whole bracket falls apart, so i built it carefully and tested the format until it was exact, then wrapped it in a simulation loop that plays the tournament out tens of thousands of times. and the last piece, the live part. as real results come in, they get locked, and only the unplayed games get re-simulated. that's what makes it a living model instead of a one-time prediction. all of it outputs to a clean dashboard you can actually read and screenshot.. right now, before kickoff, it already has a clear favorite to lift the trophy.. 👀 btw who's your pick to win the 2026 world cup?

Axel Bitblaze 🪓

63,796 görüntüleme • 3 ay önce

I am the Director of Summit Outcomes for the Presidential Advance Team. My job is to land in a foreign capital and leave with a word the President can say on the tarmac. We landed in Beijing 6 days after rolling back the tariffs we spent 4 years imposing. 145% to 30%. The average rate before the trade war was approximately 3%. In Geneva, we called this "creating the conditions for productive dialogue." The conditions were that we had already conceded. I want to be clear: Beijing was a success. We went in with 7 objectives. We left with 3 photo categories, a tentative agreement China has not confirmed, and a bag of burner phones we threw off Air Force One on the tarmac. Diplomacy. My team prepared the deliverables matrix in March. 241 line items organized by urgency, feasibility, and what we call "headline potential." The President reviewed it for 4 minutes. He circled "big deal" and "historic" and wrote "MORE" next to the Boeing section. That became the strategy. Boeing was the centerpiece. 500 aircraft was the White House number we briefed to reporters before departure. 300 was the floor. The Chinese offered 200. Their commerce ministry released the number before we could brief the press. Boeing stock dropped 4.73% that afternoon. Boeing referred questions about the order to the White House. The company receiving the aircraft could not confirm it was receiving aircraft. We called it "fantastic." In Washington, "fantastic" means the other side named the number and the market already priced in your failure. I should note: in 2017, the President announced $250 billion in deals during his first China trip. 300 aircraft. An $84 billion shale gas investment in West Virginia from China Energy Investment Corporation. I can tell you the exact amount of that investment that materialized. Zero. The shale facility was never built. The 2017 Boeing order was renegotiated twice and partially canceled during the trade war the President started 8 months later. There is a binder in my office labeled "2017 OUTCOMES: DO NOT REFERENCE." It is 3 inches thick. It has not been opened in 4 years. We do not reference it because the outcomes are the reference. The agricultural package was what we call a "scaffolding commitment." Billions in purchases over 3 years, structured so the announcement is front-loaded and the verification is someone else's administration. U.S. Trade Representative Greer said "double-digit billions." Beijing's Commerce Ministry issued a statement about "deepening cooperation in agricultural trade." Those are not the same sentence. By design. My deputy maintains a glossary of every term we have invented for agreements that are not agreements. It is 41 pages. He updates it after each summit. Last quarter he added "scaffolding commitment," "streamlined licensing framework," and "mutual recognition of shared concerns." He is in line for a promotion. NVIDIA was the quiet win. H200 chips approved for approximately 10 Chinese companies. We don't say "approved." We say "under a streamlined licensing framework." The chips ship. The export controls remain "in effect." The framework is the loophole wearing a lanyard. The controls exist because these chips in Chinese hands threaten American national security. The chips are shipping to Chinese hands. The controls remain in effect. Both of these are true. Fentanyl was discussed for 9 minutes. Both sides agreed it was a problem. Both sides agreed to continue discussing it. We added it to the deliverables matrix under "ongoing mutual engagement." The previous version of the matrix also listed it under "ongoing mutual engagement." That was in 2023. I copied the line item from the 2023 matrix into the 2026 version. Changed the date. The language was identical. But Taiwan. Taiwan was the deliverable we didn't put on the matrix. I watched the Taiwan exchange from the overflow room on a 12-second delay. I had the contingency statement drafted in 3 versions: "productive exchange," "frank discussion," and "both sides reaffirmed their respective positions." I used none of them. There was no contingency for silence. Chairman Xi released his remarks before the meeting was over. While the President was still seated across the table, Chinese state media published the transcript. "Clashes and even conflicts." His bluntest language on Taiwan in the history of the relationship, released to 1.4 billion people while we were still pouring tea. We called this "sequencing." The President was asked whether he would defend Taiwan if China attacked. He chose not to answer. We wrote that down as "a strong listen." The $14 billion arms sale. Already approved by Congress. The largest in the history of the Taiwan Relations Act. Taiwan's parliament spent months appropriating the $25 billion to proceed with this package and the $11 billion tranche approved last year. They finally secured the funding this month. The President told Fox News it was "a very good negotiating chip." He used the word "chip." Referring to the defense of 24 million people. Taiwan's Ministry of National Defense sent our office a letter requesting clarity on the delivery timeline. 3 pages. It referenced specific weapons systems by name: F-16V Block 70 fighters, HIMARS launchers, Harpoon coastal defense missiles. The letter was addressed to me. I filed it under "pending." On Air Force One, a reporter asked about the 1982 Six Assurances, the framework in which the United States committed not to consult with Beijing before selling arms to Taiwan. The President said: "What am I going to do, say I don't want to talk to you about it because I have an agreement wrote in 1982? No, we discussed arms sales." 44 years of bipartisan Taiwan policy, dismissed in 2 sentences at 38,000 feet. We are calling this "a modernized approach to alliance management." Our readout mentioned trade, agriculture, energy, and regional stability. It did not mention Taiwan. I wrote it. Their readout opened with Taiwan. I have staffed 7 summits across 2 administrations. This is the first where I could not draft a single deliverable as a success without a qualifier. In my office there is a laminated card that lists every synonym for "undecided" that polls above 40% approval. "Active review" is 3rd. "Determination" is 7th. Both tested well with independents in the Midwest. He also said: "Taiwan would be very smart to cool it a little bit. China would be very smart to cool it a little bit." He was eating a cheeseburger. He said this while eating a cheeseburger. Secretary Rubio told NBC that Taiwan arms sales "did not feature prominently." This is accurate in the same way that the iceberg did not feature prominently in the Titanic's itinerary. Representative McCaul, Republican of Texas, former chairman of the House Foreign Affairs Committee, said the United States must "arm Taiwan so they can defend themselves." He said Xi was "very aggressive" regarding Taiwan during the summit and that "most of what Xi talked about was Taiwan." Representative Meeks, Democrat of New York, ranking member of the same committee, said Xi has "leverage over the president" but not "over the United States Congress and the American people." He noted that Congress already approved the package. "The president is the one that's holding it up." Representative Fitzpatrick, Republican of Pennsylvania, compared Taiwan to Ukraine. He called both "fortresses of democracy on the front lines." Speaker Johnson said Taiwan needs to "stay independent and secure." The bipartisan consensus was that something had gone wrong. The bipartisan action was press quotes. No vote. No resolution. No hearing scheduled. 4 members of Congress from both parties said the right words to reporters and then went to lunch. That's how the system processes alarm. I monitor 14 accounts we classify as "aligned messaging amplifiers." Within 4 hours of the Taiwan exchange, 9 went silent. 2 pivoted to fentanyl. 1 posted 3 words: "Not like this." It received 280,000 impressions in 90 minutes. He deleted it and posted about the border instead. The President patted Chairman Xi on the back 7 times during the Zhongnanhai garden walk. We counted. He called him "my friend" in 4 languages, 2 of which he does not speak. He asked if other world leaders had been invited to the compound. They had. Putin was there last year. The President asked if his tour was longer. 15 CEOs flew with us to Beijing. Their combined net worth approaches $1 trillion. Cook. Musk. Jensen Huang. Larry Fink from BlackRock. Jane Fraser from Citigroup. David Solomon from Goldman Sachs. Stephen Schwarzman from Blackstone. Kelly Ortberg from Boeing. The CEO of Visa. The CEO of Mastercard. The CEO of Qualcomm. Illumina. Micron. Cargill. GE Aerospace. Musk and Huang rode on Air Force One. The others flew commercial. Tesla's Shanghai factory produces approximately half of the company's vehicles worldwide. Musk's presence on Air Force One was noted by my counterintelligence liaison. No further action was taken. We organized the state banquet seating chart by net worth. I am told this was the President's suggestion. They came for market access. Xi told them China would "open further to American business." That was the deliverable. Those 5 words. No specifics. No timeline. No sectors named. 15 chief executives flew to Beijing and received a sentence. Chairman Xi has delivered this sentence at every summit I have staffed. It has not once been followed by a named sector, a timeline, or a specific commitment. It is received as news each time. 43 lobby badges in a Ziploc bag. That's what my team collected from the CEOs after the garden tour. Standard protocol. The badges were embossed with the Great Hall of the People seal. Several executives asked if they could keep them. We said no. One asked twice. 15 executives with combined access to American financial, defense, and technology infrastructure had spent 3 hours inside the Great Hall of the People. We secured the lobby badges. The S&P 500 futures dropped 1% on the morning after the summit. The KOSPI fell 6.12%. China's CSI 300 fell 1.12%. UBS told clients that "much increasingly scarce jet fuel has been burned to produce nothing of real substance." Fortune's headline was "Wall Street sees nothing of real substance." The markets liked the anticipation. The markets did not like the deliverables matrix. Iran was the item we listed as "mutual recognition of shared concerns." The President told reporters they "feel very similar." Xi sat in silence. China's Foreign Ministry did not comment on any commitment regarding the Strait of Hormuz. The President then told reporters the United States "doesn't need the Strait of Hormuz open at all." Oil hit $109 per barrel. Deutsche Bank flagged it as a market-killing statement within the hour. The President described Iran as "a little bit crazy." This was during a toast. Over Peking duck. Rare earths. I prepared a 40-page brief on critical mineral dependency. Supply chain maps for 14 minerals. $1.2 trillion in dependent U.S. industries. Roughly 4% of GDP. The President circled the GDP figure and wrote "big." In the meeting, he asked Chairman Xi if rare earths were "the things in magnets." They are. They are also in every F-35, every Patriot missile battery, and every MRI machine in the country. The discussion lasted 11 minutes. 3 of them were about magnets. No agreement on export licenses. China exposed our dependency last year and has not let us forget it. The Supreme Court struck down our tariffs separately, which was helpful context for the discussions. Fentanyl received 9 minutes. Magnets received 3. We are calling the rare earth outcome "a foundation for continued engagement." There is a poster in the Advance Team office that says "A foundation is not a building." It has been there since my first summit. No one has removed it. On the flight home, my team collected every item the Chinese government had distributed. The credentials. The pins. The keepsakes. The rose seeds Chairman Xi offered for the White House Rose Garden. Standard counterintelligence protocol. All of it went into a bag and off the plane before wheels-up. We threw away the roses. We kept the talking points. The Boeing order grew on the flight home. 500 before departure. 200 in Beijing. 750 somewhere over the Pacific. Boeing had not confirmed 200. The President told reporters on Air Force One it was "a pretty historic couple days." I wrote the line that preceded it: "Tonal reset with significant forward momentum." He used "fantastic" instead. In previous administrations, a tonal reset preceded the deliverables. In this administration, the tonal reset is the deliverable. He has used "fantastic" for every summit since 2017. I have not checked whether the word still polls well. I am told it does. Beijing has not confirmed any of the agreements announced by U.S. officials. This is consistent with the 2017 visit, where $250 billion in deals were announced and an estimated $10 billion materialized. It is consistent with the October summit, where pledges were also made and also not fulfilled. We have a term for this in the Advance Team. We call it "precedent." I have already labeled the binder for 2026. We go back in September. Same matrix. New line items. The verification will be someone else's administration. The President has already asked for the word "monumental." I am told it polls well.
0:54

I am the Director of Summit Outcomes for the Presidential Advance Team. My job is to land in a foreign capital and leave with a word the President can say on the tarmac. We landed in Beijing 6 days after rolling back the tariffs we spent 4 years imposing. 145% to 30%. The average rate before the trade war was approximately 3%. In Geneva, we called this "creating the conditions for productive dialogue." The conditions were that we had already conceded. I want to be clear: Beijing was a success. We went in with 7 objectives. We left with 3 photo categories, a tentative agreement China has not confirmed, and a bag of burner phones we threw off Air Force One on the tarmac. Diplomacy. My team prepared the deliverables matrix in March. 241 line items organized by urgency, feasibility, and what we call "headline potential." The President reviewed it for 4 minutes. He circled "big deal" and "historic" and wrote "MORE" next to the Boeing section. That became the strategy. Boeing was the centerpiece. 500 aircraft was the White House number we briefed to reporters before departure. 300 was the floor. The Chinese offered 200. Their commerce ministry released the number before we could brief the press. Boeing stock dropped 4.73% that afternoon. Boeing referred questions about the order to the White House. The company receiving the aircraft could not confirm it was receiving aircraft. We called it "fantastic." In Washington, "fantastic" means the other side named the number and the market already priced in your failure. I should note: in 2017, the President announced $250 billion in deals during his first China trip. 300 aircraft. An $84 billion shale gas investment in West Virginia from China Energy Investment Corporation. I can tell you the exact amount of that investment that materialized. Zero. The shale facility was never built. The 2017 Boeing order was renegotiated twice and partially canceled during the trade war the President started 8 months later. There is a binder in my office labeled "2017 OUTCOMES: DO NOT REFERENCE." It is 3 inches thick. It has not been opened in 4 years. We do not reference it because the outcomes are the reference. The agricultural package was what we call a "scaffolding commitment." Billions in purchases over 3 years, structured so the announcement is front-loaded and the verification is someone else's administration. U.S. Trade Representative Greer said "double-digit billions." Beijing's Commerce Ministry issued a statement about "deepening cooperation in agricultural trade." Those are not the same sentence. By design. My deputy maintains a glossary of every term we have invented for agreements that are not agreements. It is 41 pages. He updates it after each summit. Last quarter he added "scaffolding commitment," "streamlined licensing framework," and "mutual recognition of shared concerns." He is in line for a promotion. NVIDIA was the quiet win. H200 chips approved for approximately 10 Chinese companies. We don't say "approved." We say "under a streamlined licensing framework." The chips ship. The export controls remain "in effect." The framework is the loophole wearing a lanyard. The controls exist because these chips in Chinese hands threaten American national security. The chips are shipping to Chinese hands. The controls remain in effect. Both of these are true. Fentanyl was discussed for 9 minutes. Both sides agreed it was a problem. Both sides agreed to continue discussing it. We added it to the deliverables matrix under "ongoing mutual engagement." The previous version of the matrix also listed it under "ongoing mutual engagement." That was in 2023. I copied the line item from the 2023 matrix into the 2026 version. Changed the date. The language was identical. But Taiwan. Taiwan was the deliverable we didn't put on the matrix. I watched the Taiwan exchange from the overflow room on a 12-second delay. I had the contingency statement drafted in 3 versions: "productive exchange," "frank discussion," and "both sides reaffirmed their respective positions." I used none of them. There was no contingency for silence. Chairman Xi released his remarks before the meeting was over. While the President was still seated across the table, Chinese state media published the transcript. "Clashes and even conflicts." His bluntest language on Taiwan in the history of the relationship, released to 1.4 billion people while we were still pouring tea. We called this "sequencing." The President was asked whether he would defend Taiwan if China attacked. He chose not to answer. We wrote that down as "a strong listen." The $14 billion arms sale. Already approved by Congress. The largest in the history of the Taiwan Relations Act. Taiwan's parliament spent months appropriating the $25 billion to proceed with this package and the $11 billion tranche approved last year. They finally secured the funding this month. The President told Fox News it was "a very good negotiating chip." He used the word "chip." Referring to the defense of 24 million people. Taiwan's Ministry of National Defense sent our office a letter requesting clarity on the delivery timeline. 3 pages. It referenced specific weapons systems by name: F-16V Block 70 fighters, HIMARS launchers, Harpoon coastal defense missiles. The letter was addressed to me. I filed it under "pending." On Air Force One, a reporter asked about the 1982 Six Assurances, the framework in which the United States committed not to consult with Beijing before selling arms to Taiwan. The President said: "What am I going to do, say I don't want to talk to you about it because I have an agreement wrote in 1982? No, we discussed arms sales." 44 years of bipartisan Taiwan policy, dismissed in 2 sentences at 38,000 feet. We are calling this "a modernized approach to alliance management." Our readout mentioned trade, agriculture, energy, and regional stability. It did not mention Taiwan. I wrote it. Their readout opened with Taiwan. I have staffed 7 summits across 2 administrations. This is the first where I could not draft a single deliverable as a success without a qualifier. In my office there is a laminated card that lists every synonym for "undecided" that polls above 40% approval. "Active review" is 3rd. "Determination" is 7th. Both tested well with independents in the Midwest. He also said: "Taiwan would be very smart to cool it a little bit. China would be very smart to cool it a little bit." He was eating a cheeseburger. He said this while eating a cheeseburger. Secretary Rubio told NBC that Taiwan arms sales "did not feature prominently." This is accurate in the same way that the iceberg did not feature prominently in the Titanic's itinerary. Representative McCaul, Republican of Texas, former chairman of the House Foreign Affairs Committee, said the United States must "arm Taiwan so they can defend themselves." He said Xi was "very aggressive" regarding Taiwan during the summit and that "most of what Xi talked about was Taiwan." Representative Meeks, Democrat of New York, ranking member of the same committee, said Xi has "leverage over the president" but not "over the United States Congress and the American people." He noted that Congress already approved the package. "The president is the one that's holding it up." Representative Fitzpatrick, Republican of Pennsylvania, compared Taiwan to Ukraine. He called both "fortresses of democracy on the front lines." Speaker Johnson said Taiwan needs to "stay independent and secure." The bipartisan consensus was that something had gone wrong. The bipartisan action was press quotes. No vote. No resolution. No hearing scheduled. 4 members of Congress from both parties said the right words to reporters and then went to lunch. That's how the system processes alarm. I monitor 14 accounts we classify as "aligned messaging amplifiers." Within 4 hours of the Taiwan exchange, 9 went silent. 2 pivoted to fentanyl. 1 posted 3 words: "Not like this." It received 280,000 impressions in 90 minutes. He deleted it and posted about the border instead. The President patted Chairman Xi on the back 7 times during the Zhongnanhai garden walk. We counted. He called him "my friend" in 4 languages, 2 of which he does not speak. He asked if other world leaders had been invited to the compound. They had. Putin was there last year. The President asked if his tour was longer. 15 CEOs flew with us to Beijing. Their combined net worth approaches $1 trillion. Cook. Musk. Jensen Huang. Larry Fink from BlackRock. Jane Fraser from Citigroup. David Solomon from Goldman Sachs. Stephen Schwarzman from Blackstone. Kelly Ortberg from Boeing. The CEO of Visa. The CEO of Mastercard. The CEO of Qualcomm. Illumina. Micron. Cargill. GE Aerospace. Musk and Huang rode on Air Force One. The others flew commercial. Tesla's Shanghai factory produces approximately half of the company's vehicles worldwide. Musk's presence on Air Force One was noted by my counterintelligence liaison. No further action was taken. We organized the state banquet seating chart by net worth. I am told this was the President's suggestion. They came for market access. Xi told them China would "open further to American business." That was the deliverable. Those 5 words. No specifics. No timeline. No sectors named. 15 chief executives flew to Beijing and received a sentence. Chairman Xi has delivered this sentence at every summit I have staffed. It has not once been followed by a named sector, a timeline, or a specific commitment. It is received as news each time. 43 lobby badges in a Ziploc bag. That's what my team collected from the CEOs after the garden tour. Standard protocol. The badges were embossed with the Great Hall of the People seal. Several executives asked if they could keep them. We said no. One asked twice. 15 executives with combined access to American financial, defense, and technology infrastructure had spent 3 hours inside the Great Hall of the People. We secured the lobby badges. The S&P 500 futures dropped 1% on the morning after the summit. The KOSPI fell 6.12%. China's CSI 300 fell 1.12%. UBS told clients that "much increasingly scarce jet fuel has been burned to produce nothing of real substance." Fortune's headline was "Wall Street sees nothing of real substance." The markets liked the anticipation. The markets did not like the deliverables matrix. Iran was the item we listed as "mutual recognition of shared concerns." The President told reporters they "feel very similar." Xi sat in silence. China's Foreign Ministry did not comment on any commitment regarding the Strait of Hormuz. The President then told reporters the United States "doesn't need the Strait of Hormuz open at all." Oil hit $109 per barrel. Deutsche Bank flagged it as a market-killing statement within the hour. The President described Iran as "a little bit crazy." This was during a toast. Over Peking duck. Rare earths. I prepared a 40-page brief on critical mineral dependency. Supply chain maps for 14 minerals. $1.2 trillion in dependent U.S. industries. Roughly 4% of GDP. The President circled the GDP figure and wrote "big." In the meeting, he asked Chairman Xi if rare earths were "the things in magnets." They are. They are also in every F-35, every Patriot missile battery, and every MRI machine in the country. The discussion lasted 11 minutes. 3 of them were about magnets. No agreement on export licenses. China exposed our dependency last year and has not let us forget it. The Supreme Court struck down our tariffs separately, which was helpful context for the discussions. Fentanyl received 9 minutes. Magnets received 3. We are calling the rare earth outcome "a foundation for continued engagement." There is a poster in the Advance Team office that says "A foundation is not a building." It has been there since my first summit. No one has removed it. On the flight home, my team collected every item the Chinese government had distributed. The credentials. The pins. The keepsakes. The rose seeds Chairman Xi offered for the White House Rose Garden. Standard counterintelligence protocol. All of it went into a bag and off the plane before wheels-up. We threw away the roses. We kept the talking points. The Boeing order grew on the flight home. 500 before departure. 200 in Beijing. 750 somewhere over the Pacific. Boeing had not confirmed 200. The President told reporters on Air Force One it was "a pretty historic couple days." I wrote the line that preceded it: "Tonal reset with significant forward momentum." He used "fantastic" instead. In previous administrations, a tonal reset preceded the deliverables. In this administration, the tonal reset is the deliverable. He has used "fantastic" for every summit since 2017. I have not checked whether the word still polls well. I am told it does. Beijing has not confirmed any of the agreements announced by U.S. officials. This is consistent with the 2017 visit, where $250 billion in deals were announced and an estimated $10 billion materialized. It is consistent with the October summit, where pledges were also made and also not fulfilled. We have a term for this in the Advance Team. We call it "precedent." I have already labeled the binder for 2026. We go back in September. Same matrix. New line items. The verification will be someone else's administration. The President has already asked for the word "monumental." I am told it polls well.

Peter Girnus 🦅

97,655 görüntüleme • 4 ay önce

Andrew Wilkinson owns 40+ businesses. He just showed me how he's using OpenClaw, Claude Code and AI agents to run latest business, start new ones, and automate everything. Here's what I learned: 1. In December 2025, something clicked. He started waking up at 3AM with a smile, sitting in terminal with 10 Claude Code tabs open. He hasn't stopped since. He calls it chasing the dragon. 2.He built a full SaaS product called Deep Personality. A 40-minute personality test that generates a 100-page report written like Robert Greene. $20 000 in revenue. Zero employees. The entire business runs on AI agents. 3. He has agents for support, marketing, and dev. When a support ticket comes in, the agent either handles it or sends it to the dev agent. If it's critical, the agent fixes the bug and merges the PR before he wakes up. Then it emails the customer back. 4. His marketing agent is connected to PostHog, manages Meta and Reddit ads, creates ad creative, runs multivariate tests, and sets budgets. He's about to give it a $100 k/month ad budget and see what happens. 5. He forgot his laptop on a trip to Arizona. He ran his entire business from the back of Ubers using OpenClaw. Nobody picked up that every single email was written by AI. 6. His take on vibe coding: the worst part about business is people. Between your vision and execution are 100 people you have to convince. Vibe coding removes all of them. For the first time he can do every part of building a product himself. 7. He was trying to build OpenClaw before OpenClaw existed. Now he uses a tool called Harbor, which is basically a GUI for managing multiple agents. You can see all your agents, their status, knowledge bases, and databases in one place. 8. He built a custom AI for his relationship. He and his girlfriend took 15 psychological tests, put the results into ChatGPT, and asked it to analyze their relationship. It nailed every fight they've ever had. That became the product idea for Deep Personality. 9. His honest take: he spends 50% of his time debugging, 30% improving the setup, and 20% being productive. It's a treadmill. But the 20% that works is so powerful he can't stop. 10. His prediction: we're 3-6 months from being able to hand basic businesses off to AI to run entirely. And pretty soon Anthropic and OpenAI are going to launch AI CEOs. This is an inside look at how a serious operator Andrew Wilkinson is using AI agents in the real world. The good, the bad, the debugging, all of it. Most people don't show you this. Episode is live on The Startup Ideas Podcast (SIP) 🧃 watch
47:23

Andrew Wilkinson owns 40+ businesses. He just showed me how he's using OpenClaw, Claude Code and AI agents to run latest business, start new ones, and automate everything. Here's what I learned: 1. In December 2025, something clicked. He started waking up at 3AM with a smile, sitting in terminal with 10 Claude Code tabs open. He hasn't stopped since. He calls it chasing the dragon. 2.He built a full SaaS product called Deep Personality. A 40-minute personality test that generates a 100-page report written like Robert Greene. $20 000 in revenue. Zero employees. The entire business runs on AI agents. 3. He has agents for support, marketing, and dev. When a support ticket comes in, the agent either handles it or sends it to the dev agent. If it's critical, the agent fixes the bug and merges the PR before he wakes up. Then it emails the customer back. 4. His marketing agent is connected to PostHog, manages Meta and Reddit ads, creates ad creative, runs multivariate tests, and sets budgets. He's about to give it a $100 k/month ad budget and see what happens. 5. He forgot his laptop on a trip to Arizona. He ran his entire business from the back of Ubers using OpenClaw. Nobody picked up that every single email was written by AI. 6. His take on vibe coding: the worst part about business is people. Between your vision and execution are 100 people you have to convince. Vibe coding removes all of them. For the first time he can do every part of building a product himself. 7. He was trying to build OpenClaw before OpenClaw existed. Now he uses a tool called Harbor, which is basically a GUI for managing multiple agents. You can see all your agents, their status, knowledge bases, and databases in one place. 8. He built a custom AI for his relationship. He and his girlfriend took 15 psychological tests, put the results into ChatGPT, and asked it to analyze their relationship. It nailed every fight they've ever had. That became the product idea for Deep Personality. 9. His honest take: he spends 50% of his time debugging, 30% improving the setup, and 20% being productive. It's a treadmill. But the 20% that works is so powerful he can't stop. 10. His prediction: we're 3-6 months from being able to hand basic businesses off to AI to run entirely. And pretty soon Anthropic and OpenAI are going to launch AI CEOs. This is an inside look at how a serious operator Andrew Wilkinson is using AI agents in the real world. The good, the bad, the debugging, all of it. Most people don't show you this. Episode is live on The Startup Ideas Podcast (SIP) 🧃 watch

GREG ISENBERG

144,005 görüntüleme • 4 ay önce