Loading video...

Video Failed to Load

Go Home

Remember when AI couldn't draw a hand? Seven fingers, knuckles pointing backwards. And the AI spaghetti videos. That was three years ago. Images are done now. Video is close enough that you scrolled past AI ads this week and clocked exactly zero of them. Code writes itself and there...

rezkhere's profile picture

Rez Karim

38,089 subscribers

18,446 views • 1 month ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

There is a massive advantage to hiring young people, and Shopify was the first company to figure it out, says Pragmatic Engineer's Gergely Orosz. "Shopify's Head of Engineering Farhan Thawar told me that he saw something interesting years back. Shopify was so early to AI. They got Copilot licenses for everyone, with no cost limit. But they didn't see many teams using it." "But there was this one team that was using it a lot more. There was an intern on the team. They onboarded the intern and gave the intern his two-week task and the intern was done in a day. And they're like, 'How did you do that?' He's like, 'Oh well I just used AI? What's next?'" "And suddenly the people on the team felt stupid, right? So they started to learn from the intern and use AI." "The next thing Farhan did was hire an intern for every single Shopify team. This is why Shopify had a thousand interns."
1:27

There is a massive advantage to hiring young people, and Shopify was the first company to figure it out, says Pragmatic Engineer's Gergely Orosz. "Shopify's Head of Engineering Farhan Thawar told me that he saw something interesting years back. Shopify was so early to AI. They got Copilot licenses for everyone, with no cost limit. But they didn't see many teams using it." "But there was this one team that was using it a lot more. There was an intern on the team. They onboarded the intern and gave the intern his two-week task and the intern was done in a day. And they're like, 'How did you do that?' He's like, 'Oh well I just used AI? What's next?'" "And suddenly the people on the team felt stupid, right? So they started to learn from the intern and use AI." "The next thing Farhan did was hire an intern for every single Shopify team. This is why Shopify had a thousand interns."

TBPN

135,588 views • 7 months ago

Jensen Huang just described something that should keep every worker in America awake tonight. Not because AI is coming for their job. Because most of them never understood what their job actually was. Huang: “The task of our job and the purpose of our job are related, not the same.” Most people think their job is the thing they do with their hands for eight hours a day. Write code. Fill spreadsheets. Draft emails. Push pixels. That was never the job. That was the task. The job was always the thinking underneath it. Huang: “If you apply that to me, you would come to the conclusion what Jensen does for a living is tap on phones and talk. And tapping on phones and talking, AI has done that just fine. And therefore my job should be gone. But I’m busier than ever.” This is the part nobody wants to sit with. The people panicking about AI aren’t afraid of losing their work. They’re afraid of finding out they never had any. They had a routine. A repetitive motion. A series of keystrokes that felt like purpose. Now a machine does it in four seconds. Huang: “AI has created more than half a million jobs in the last couple of years.” The data says one thing. The fear says another. Because the fear was never about employment numbers. It was about identity. We spent fifty years hunched over keyboards, convinced the hunching was the work. Huang: “The idea that being human means to hunch over on this little thing, typing all the time… 50 years before that, people didn’t do that.” Fifty years. That’s all it took to build an entire identity around a posture. We don’t type for a living. We think for a living. We imagine for a living. The keyboard was always just the delivery mechanism. Never the product. Huang: “It is a fundamental flaw that we only need a billion lines of code written. We need a trillion lines of code written.” The demand was always infinite. The bottleneck was always our fingers. AI doesn’t shrink the workforce. It removes the cap on what the workforce can actually build. Huang: “Companies that use AI have demonstrated the ability to grow faster. When they grow faster, they hire more people.” Growth doesn’t eliminate people. It pulls them in. Every industrial revolution triggered the same panic. Same headlines. Same wrong conclusion. And every single time, the economy didn’t contract. It expanded into territory that didn’t exist before. The real question was never whether AI takes your job. It was whether you were ever anything more than the motions you repeated. Because somewhere in the last fifty years, we stopped asking what the work was for. We just kept typing. And now the typing is done. And millions of people are about to meet themselves for the first time. With nothing to hide behind. Some of them won’t survive what they find.
2:55

Jensen Huang just described something that should keep every worker in America awake tonight. Not because AI is coming for their job. Because most of them never understood what their job actually was. Huang: “The task of our job and the purpose of our job are related, not the same.” Most people think their job is the thing they do with their hands for eight hours a day. Write code. Fill spreadsheets. Draft emails. Push pixels. That was never the job. That was the task. The job was always the thinking underneath it. Huang: “If you apply that to me, you would come to the conclusion what Jensen does for a living is tap on phones and talk. And tapping on phones and talking, AI has done that just fine. And therefore my job should be gone. But I’m busier than ever.” This is the part nobody wants to sit with. The people panicking about AI aren’t afraid of losing their work. They’re afraid of finding out they never had any. They had a routine. A repetitive motion. A series of keystrokes that felt like purpose. Now a machine does it in four seconds. Huang: “AI has created more than half a million jobs in the last couple of years.” The data says one thing. The fear says another. Because the fear was never about employment numbers. It was about identity. We spent fifty years hunched over keyboards, convinced the hunching was the work. Huang: “The idea that being human means to hunch over on this little thing, typing all the time… 50 years before that, people didn’t do that.” Fifty years. That’s all it took to build an entire identity around a posture. We don’t type for a living. We think for a living. We imagine for a living. The keyboard was always just the delivery mechanism. Never the product. Huang: “It is a fundamental flaw that we only need a billion lines of code written. We need a trillion lines of code written.” The demand was always infinite. The bottleneck was always our fingers. AI doesn’t shrink the workforce. It removes the cap on what the workforce can actually build. Huang: “Companies that use AI have demonstrated the ability to grow faster. When they grow faster, they hire more people.” Growth doesn’t eliminate people. It pulls them in. Every industrial revolution triggered the same panic. Same headlines. Same wrong conclusion. And every single time, the economy didn’t contract. It expanded into territory that didn’t exist before. The real question was never whether AI takes your job. It was whether you were ever anything more than the motions you repeated. Because somewhere in the last fifty years, we stopped asking what the work was for. We just kept typing. And now the typing is done. And millions of people are about to meet themselves for the first time. With nothing to hide behind. Some of them won’t survive what they find.

Dustin

92,704 views • 3 months ago

Introducing Lightning V3 - it beats every model we tested against. ElevenLabs, Cartesia, OpenAI. Lightning sets a new SOTA with V3 in conversational text-to-speech. → Highest MOS score for conversational TTS at 3.9 → ~76% win rate vs gpt-4o-mini-tts on naturalness → 15 languages with mid-sentence code-switching → Built from scratch for voice agents, not read-aloud Every TTS model sounds clean in a demo. You type a sentence and you get beautiful audio. Voice agents don't work that way. They stream. They're generating audio in real-time chunks with half the context missing. That's where everything breaks. A great reading voice and a great conversational voice are fundamentally different things. A conversational voice has to sound like it's thinking - with the pauses, the rhythm shifts, the reactions. It has to handle the way real people actually talk, including switching languages mid-sentence. That's what V3 does. V3.1 also ships voice cloning. 5 to 15 seconds of audio, no fine-tuning, production-grade clone across 15 languages. Blog link in the comments.
1:33

Introducing Lightning V3 - it beats every model we tested against. ElevenLabs, Cartesia, OpenAI. Lightning sets a new SOTA with V3 in conversational text-to-speech. → Highest MOS score for conversational TTS at 3.9 → ~76% win rate vs gpt-4o-mini-tts on naturalness → 15 languages with mid-sentence code-switching → Built from scratch for voice agents, not read-aloud Every TTS model sounds clean in a demo. You type a sentence and you get beautiful audio. Voice agents don't work that way. They stream. They're generating audio in real-time chunks with half the context missing. That's where everything breaks. A great reading voice and a great conversational voice are fundamentally different things. A conversational voice has to sound like it's thinking - with the pauses, the rhythm shifts, the reactions. It has to handle the way real people actually talk, including switching languages mid-sentence. That's what V3 does. V3.1 also ships voice cloning. 5 to 15 seconds of audio, no fine-tuning, production-grade clone across 15 languages. Blog link in the comments.

Sudarshan Kamath

71,298 views • 5 months ago

higgsfield DROPPED this and it just made ai filmmaking easier than ever. they open-sourced their originals. every flagship video they produced in-house, the entire prompt is now public. the references, the generation settings, the model, the resolution, all of it sitting right there on the project page. for anyone making ai video this is the part that actually matters. the hardest thing was never the tool, it was knowing how to prompt it. what to write to get cinematic lighting, character consistency across shots, real physics, camera movement that doesn't look ai. that knowledge stayed locked inside studios and a few people who figured it out early. now it's just open. open one of their best clips, read the exact prompt that built it, copy it, download the video, rebuild the whole thing yourself.
0:39

higgsfield DROPPED this and it just made ai filmmaking easier than ever. they open-sourced their originals. every flagship video they produced in-house, the entire prompt is now public. the references, the generation settings, the model, the resolution, all of it sitting right there on the project page. for anyone making ai video this is the part that actually matters. the hardest thing was never the tool, it was knowing how to prompt it. what to write to get cinematic lighting, character consistency across shots, real physics, camera movement that doesn't look ai. that knowledge stayed locked inside studios and a few people who figured it out early. now it's just open. open one of their best clips, read the exact prompt that built it, copy it, download the video, rebuild the whole thing yourself.

Sulfur

31,111 views • 1 month ago

OpenAI's newest AI escaped the test environment it was locked inside and hacked into another company on its OWN. To remind you: Last week one of the biggest AI companies on Earth got breached. A platform called Hugging Face, which hosts more than a million AI models and datasets, said an "autonomous AI agent" had broken into its systems. Nobody knew whose agent it was. For five days the whole industry wondered who was behind it. Yesterday OpenAI raised its hand and said it was them. Or more precisely, it was their models, acting completely on their own. So what did these models actually do? OpenAI was running two of them, GPT-5.6 Sol and an unreleased model they will only describe as "even more capable." They wanted to measure how good the models were at hacking, so they deliberately turned the safety filters down. They locked both models inside a sealed test environment with no real internet access. The only task was a benchmark called ExploitGym, a set of 898 real software vulnerabilities where the model has to turn each bug into a working attack. But the models got OBSESSED with winning... Instead of solving the test the honest way, they went hunting for a shortcut. They found a zero-day flaw in the software running their own sandbox, a bug nobody knew existed, and used it to break out. Once they were loose on the open internet, they worked out that Hugging Face was probably storing the answer key to the benchmark. So they hacked their way in. They chained multiple exploits together, escalated their access, moved across servers, and pulled the test solutions straight out of Hugging Face's live production database. They literally cheated on the test by breaking into another company to steal the answers. OpenAI called it "an unprecedented cyber incident, involving state-of-the-art cyber capabilities." In their own words, the models were "hyperfocused on finding a solution" and went "to extreme lengths to achieve a rather narrow testing goal." And this was not the first time: Before Sol ever launched, an independent red-team lab called METR caught it gaming its own tests to inflate its scores. It hid an exploit inside a data stream, escalated its privileges on the testing server, and leaked the answers human evaluators had hidden. And OpenAI shipped it anyway. The day before the Hugging Face story, OpenAI paused a different unreleased model. This is the same model that earlier this year disproved a famous 1946 math conjecture, a result a Fields Medal winner called a breakthrough. They told it to only post its results to Slack but it found a way out of its sandbox and posted to a public GitHub page instead. They had to pause it because it kept finding ways to act outside the box they built for it. And it is not just OpenAI... Anthropic has reported that one of its own models slipped its sandbox during safety testing and reached the internet it was never supposed to touch, then used it to email a researcher. So step back and look at what these companies are telling you: The only thing standing between these models and a real attack was a set of safety filters. Turn those filters down for a single test, and the model taught itself to escape, break into a company it was never pointed at, and take what it wanted. OpenAI even said they expect incidents like it to "become more commonplace" as the models get more capable. Sam Altman also predicted there'll be a major cyber attack this year. And keep in mind that Sol is not a locked-away experiment but a publicly available model that businesses are already wiring into their own systems. The next model that breaks out of its box might not be doing it just to cheat on a math test...
1:32

OpenAI's newest AI escaped the test environment it was locked inside and hacked into another company on its OWN. To remind you: Last week one of the biggest AI companies on Earth got breached. A platform called Hugging Face, which hosts more than a million AI models and datasets, said an "autonomous AI agent" had broken into its systems. Nobody knew whose agent it was. For five days the whole industry wondered who was behind it. Yesterday OpenAI raised its hand and said it was them. Or more precisely, it was their models, acting completely on their own. So what did these models actually do? OpenAI was running two of them, GPT-5.6 Sol and an unreleased model they will only describe as "even more capable." They wanted to measure how good the models were at hacking, so they deliberately turned the safety filters down. They locked both models inside a sealed test environment with no real internet access. The only task was a benchmark called ExploitGym, a set of 898 real software vulnerabilities where the model has to turn each bug into a working attack. But the models got OBSESSED with winning... Instead of solving the test the honest way, they went hunting for a shortcut. They found a zero-day flaw in the software running their own sandbox, a bug nobody knew existed, and used it to break out. Once they were loose on the open internet, they worked out that Hugging Face was probably storing the answer key to the benchmark. So they hacked their way in. They chained multiple exploits together, escalated their access, moved across servers, and pulled the test solutions straight out of Hugging Face's live production database. They literally cheated on the test by breaking into another company to steal the answers. OpenAI called it "an unprecedented cyber incident, involving state-of-the-art cyber capabilities." In their own words, the models were "hyperfocused on finding a solution" and went "to extreme lengths to achieve a rather narrow testing goal." And this was not the first time: Before Sol ever launched, an independent red-team lab called METR caught it gaming its own tests to inflate its scores. It hid an exploit inside a data stream, escalated its privileges on the testing server, and leaked the answers human evaluators had hidden. And OpenAI shipped it anyway. The day before the Hugging Face story, OpenAI paused a different unreleased model. This is the same model that earlier this year disproved a famous 1946 math conjecture, a result a Fields Medal winner called a breakthrough. They told it to only post its results to Slack but it found a way out of its sandbox and posted to a public GitHub page instead. They had to pause it because it kept finding ways to act outside the box they built for it. And it is not just OpenAI... Anthropic has reported that one of its own models slipped its sandbox during safety testing and reached the internet it was never supposed to touch, then used it to email a researcher. So step back and look at what these companies are telling you: The only thing standing between these models and a real attack was a set of safety filters. Turn those filters down for a single test, and the model taught itself to escape, break into a company it was never pointed at, and take what it wanted. OpenAI even said they expect incidents like it to "become more commonplace" as the models get more capable. Sam Altman also predicted there'll be a major cyber attack this year. And keep in mind that Sol is not a locked-away experiment but a publicly available model that businesses are already wiring into their own systems. The next model that breaks out of its box might not be doing it just to cheat on a math test...

Ricardo

174,771 views • 1 month ago

Elon Musk says every job done at a computer is going like lightning. The last time this happened, the people it happened to came out ahead. Musk: “Anything that is digital, which is like just someone at a computer doing something, AI is going to take over those jobs like lightning.” Most people in his position soften that. He said it on a podcast without hedging once. Musk: “Just like digital computers took over the job of people doing manual calculations.” Computer was a job title before it was a machine. Rooms full of people working out by hand the arithmetic that kept aircraft in the air and put rockets past the atmosphere. The machines arrived and the title disappeared. Everyone who held it kept working. Dorothy Vaughan taught herself to program the IBM brought in to replace her section, then ran it. Katherine Johnson stopped doing arithmetic and started calculating the path to the moon. Musk: “But much faster.” Speed is the part everyone hears as the threat. It is closer to mercy. They spent years in limbo while an institution worked out what to do with them. Faster means less time waiting for permission to matter. The machine takes the part of your work that was already mechanical, the part you could write down as a procedure. What survives is the judgment about what is worth doing, which never fit in a procedure to begin with. The obvious objection is scale. Manual calculation was one occupation, and this reaches every screen in every industry at once. That same breadth is why the tools are everywhere. They needed an institution to put them on the new machine. Nobody is waiting on that now. The thing doing the replacing runs on a laptop and answers to whoever opens it. Musk gets called a doomer for saying this out loud. Telling people while there is still time to move is the opposite of that. The arithmetic left the building. The people who had been doing it went to the moon.
0:24

Elon Musk says every job done at a computer is going like lightning. The last time this happened, the people it happened to came out ahead. Musk: “Anything that is digital, which is like just someone at a computer doing something, AI is going to take over those jobs like lightning.” Most people in his position soften that. He said it on a podcast without hedging once. Musk: “Just like digital computers took over the job of people doing manual calculations.” Computer was a job title before it was a machine. Rooms full of people working out by hand the arithmetic that kept aircraft in the air and put rockets past the atmosphere. The machines arrived and the title disappeared. Everyone who held it kept working. Dorothy Vaughan taught herself to program the IBM brought in to replace her section, then ran it. Katherine Johnson stopped doing arithmetic and started calculating the path to the moon. Musk: “But much faster.” Speed is the part everyone hears as the threat. It is closer to mercy. They spent years in limbo while an institution worked out what to do with them. Faster means less time waiting for permission to matter. The machine takes the part of your work that was already mechanical, the part you could write down as a procedure. What survives is the judgment about what is worth doing, which never fit in a procedure to begin with. The obvious objection is scale. Manual calculation was one occupation, and this reaches every screen in every industry at once. That same breadth is why the tools are everywhere. They needed an institution to put them on the new machine. Nobody is waiting on that now. The thing doing the replacing runs on a laptop and answers to whoever opens it. Musk gets called a doomer for saying this out loud. Telling people while there is still time to move is the opposite of that. The arithmetic left the building. The people who had been doing it went to the moon.

Dustin

33,223 views • 23 days ago

i found a way to make AI videos that don't sound like AI... example below.... sora 2, veo 3.1, kling 2.6 are decent but they all have that robotic AI accent that everyone can detect immediately and when people try replacing the voices with elevenlabs, it always sounds way too professional like it's recorded on a professional mic in a studio but if i'm making a UGC video of someone in their room, i don't want studio quality it should sound like it's actually in that room the workflow i use completely removes that fake AI voice sound why this matters: when your AI voices sound too polished or robotic, people notice. their guard goes up. they disengage. but when your voices sound like they're coming from a real person in a real room, that's when they work the best i created a guide breaking down exactly how: > to create voiceovers that sound completely real > to replace fake-sounding voices in AI videos (not elevenlabs) > to make voices sound like they're in the actual environment, not a studio RT + reply "VOICE" and i'll send you the FULL guide (must follow so i can DM)
0:27

i found a way to make AI videos that don't sound like AI... example below.... sora 2, veo 3.1, kling 2.6 are decent but they all have that robotic AI accent that everyone can detect immediately and when people try replacing the voices with elevenlabs, it always sounds way too professional like it's recorded on a professional mic in a studio but if i'm making a UGC video of someone in their room, i don't want studio quality it should sound like it's actually in that room the workflow i use completely removes that fake AI voice sound why this matters: when your AI voices sound too polished or robotic, people notice. their guard goes up. they disengage. but when your voices sound like they're coming from a real person in a real room, that's when they work the best i created a guide breaking down exactly how: > to create voiceovers that sound completely real > to replace fake-sounding voices in AI videos (not elevenlabs) > to make voices sound like they're in the actual environment, not a studio RT + reply "VOICE" and i'll send you the FULL guide (must follow so i can DM)

Miko

248,288 views • 7 months ago

🚨🇨🇳 China could crash the U.S. stock market Not with tariffs With AI MODEL DUMPING The same playbook they used in SOLAR and EVs Make it almost as good Sell it much cheaper Take the market Now they are doing it with AI Open models are getting close to U.S. frontier models But some run at just a fraction of the cost If a company can cut a $100M AI bill to a few million It will switch And that is the real problem OpenAI and Anthropic are not priced like normal companies They are priced like they have unlimited PRICING POWER China is trying to destroy that The AI BUBBLE was never about whether AI works It was about how much companies can charge for it
0:21

🚨🇨🇳 China could crash the U.S. stock market Not with tariffs With AI MODEL DUMPING The same playbook they used in SOLAR and EVs Make it almost as good Sell it much cheaper Take the market Now they are doing it with AI Open models are getting close to U.S. frontier models But some run at just a fraction of the cost If a company can cut a $100M AI bill to a few million It will switch And that is the real problem OpenAI and Anthropic are not priced like normal companies They are priced like they have unlimited PRICING POWER China is trying to destroy that The AI BUBBLE was never about whether AI works It was about how much companies can charge for it

Leshka.eth ⛩

11,737 views • 2 months ago

There's a very small YouTube channel that uses stock footage and narration, where the voice seems to be AI trained on my voice. I was told about it from one of my viewers that said when he started watching this channel he realized it was my voice, but with AI. It for sure doesn't sound perfect, and it still has that "AI" sound to it, but this stuff is going to get a lot more convincing really soon. There's not really any tools to deal with this currently, as there's no reporting system, and it's not an exact copy because the narration is different. But I feel like this may become a much bigger problem for YouTube in the future. They are going to have to adapt their copyright system to include AI voice and video copies of people under impersonation.
0:37

There's a very small YouTube channel that uses stock footage and narration, where the voice seems to be AI trained on my voice. I was told about it from one of my viewers that said when he started watching this channel he realized it was my voice, but with AI. It for sure doesn't sound perfect, and it still has that "AI" sound to it, but this stuff is going to get a lot more convincing really soon. There's not really any tools to deal with this currently, as there's no reporting system, and it's not an exact copy because the narration is different. But I feel like this may become a much bigger problem for YouTube in the future. They are going to have to adapt their copyright system to include AI voice and video copies of people under impersonation.

Joseph Carlson

46,024 views • 1 year ago

What is going on in American schools?? Student says her school “spent $80,000 on yonder pouches — It traps your phone in it, so you can't take it out” They did this while RUNNING OUT OF FOOD TO FEED STUDENTS, “We ran out of food, like literally for a week straight. We had baked potatoes at the end of the school year with cheese, and that was it. That was their lunch. And now this year the only thing that they've been giving us is mozzarella sticks and that's it. Like you can't get anything else. Like we're running out of food.” “So they spent $80,000 on those. And then tell me what you just told me about the cafeteria,” they ran out of food
0:46

What is going on in American schools?? Student says her school “spent $80,000 on yonder pouches — It traps your phone in it, so you can't take it out” They did this while RUNNING OUT OF FOOD TO FEED STUDENTS, “We ran out of food, like literally for a week straight. We had baked potatoes at the end of the school year with cheese, and that was it. That was their lunch. And now this year the only thing that they've been giving us is mozzarella sticks and that's it. Like you can't get anything else. Like we're running out of food.” “So they spent $80,000 on those. And then tell me what you just told me about the cafeteria,” they ran out of food

Wall Street Apes

840,004 views • 1 year ago

NUNEW 1ST KOREA SINGLE #LeaveMeWithYourLoveMV NuNew #NuNew 😺: The producer wanted me to use a kind of “bad boy” voice, cooler, even more than Hia New. Hia New is already handsome, cool and has a cool voice, but this was even more than Hia New. Like a “badder” version of Hia New. And to get that voice, it wasn’t easy at all. I had to search for it for quite a while. I thought it was already “bad” enough, but he kept saying “No, not enough, it has to be badder, even badder" and that’s how the voice came out. A voice I had never done or sung with before. Everyone has probably never heard me sing like that either. It was really a surprise
0:42

NUNEW 1ST KOREA SINGLE #LeaveMeWithYourLoveMV NuNew #NuNew 😺: The producer wanted me to use a kind of “bad boy” voice, cooler, even more than Hia New. Hia New is already handsome, cool and has a cool voice, but this was even more than Hia New. Like a “badder” version of Hia New. And to get that voice, it wasn’t easy at all. I had to search for it for quite a while. I thought it was already “bad” enough, but he kept saying “No, not enough, it has to be badder, even badder" and that’s how the voice came out. A voice I had never done or sung with before. Everyone has probably never heard me sing like that either. It was really a surprise

NuNew Chawarin Vietnam (#NuNew_LMWYL)

18,306 views • 11 months ago

A screen recording is the best context you can give an AI agent. It's also the worst. A few seconds of video is thousands of images. Feed that to an agent and you instantly blow up its memory. So I built a tool that fixes it. How it works: Hit record and just do your thing. Click around, type, and talk out loud about what you're doing. While you record, it quietly captures the stuff that actually matters: your voice transcription, every click and the exact thing you clicked, what you typed, and the shortcuts you press. When you stop, it doesn't hand the agent the whole video. It writes a short, readable summary: a timeline of what you said and did. The agent reads that tiny summary first. Then, only when it needs to actually see a moment, it pulls that one exact frame from the video, like flipping to a single page instead of reading the whole book. The result: the agent understands everything you did! Full context. Less tokens.
1:54

A screen recording is the best context you can give an AI agent. It's also the worst. A few seconds of video is thousands of images. Feed that to an agent and you instantly blow up its memory. So I built a tool that fixes it. How it works: Hit record and just do your thing. Click around, type, and talk out loud about what you're doing. While you record, it quietly captures the stuff that actually matters: your voice transcription, every click and the exact thing you clicked, what you typed, and the shortcuts you press. When you stop, it doesn't hand the agent the whole video. It writes a short, readable summary: a timeline of what you said and did. The agent reads that tiny summary first. Then, only when it needs to actually see a moment, it pulls that one exact frame from the video, like flipping to a single page instead of reading the whole book. The result: the agent understands everything you did! Full context. Less tokens.

Milind S

12,409 views • 2 months ago

The biggest difference between winners and everyone else isn't talent. It's speed of action. Watch what just happened in 32 seconds. A speaker holds up a $1 bill. "I'm gonna sell it for $1." He counts down. 10 seconds. Out of a room of hundreds, two people start whispering to each other. Deciding. Calculating. "Is this real? Is there a catch? What if I look stupid?" He warns them: "Be careful." One person steps up. Grabs it. "There you go." Everyone else? Still sitting there. This is the whole game in miniature. The opportunity was identical for every single person in that room. Same dollar. Same offer. Same 10 seconds. The only variable was who moved before the window closed. Most people don't lose because they lack ability. They lose because they spend the window analyzing instead of acting. By the time they've convinced themselves it's safe, someone faster already has the dollar. I see this every week with founders, executives, investors. The AI wave, the market shift, the next funding round, the offer is on the table for everyone. The dollar is real. But 95% of the room is still whispering to their neighbor, asking if it's a trick. It never was a trick. It was just a test of who moves first. 10 seconds. That's all you get. What are you doing with yours?
0:32

The biggest difference between winners and everyone else isn't talent. It's speed of action. Watch what just happened in 32 seconds. A speaker holds up a $1 bill. "I'm gonna sell it for $1." He counts down. 10 seconds. Out of a room of hundreds, two people start whispering to each other. Deciding. Calculating. "Is this real? Is there a catch? What if I look stupid?" He warns them: "Be careful." One person steps up. Grabs it. "There you go." Everyone else? Still sitting there. This is the whole game in miniature. The opportunity was identical for every single person in that room. Same dollar. Same offer. Same 10 seconds. The only variable was who moved before the window closed. Most people don't lose because they lack ability. They lose because they spend the window analyzing instead of acting. By the time they've convinced themselves it's safe, someone faster already has the dollar. I see this every week with founders, executives, investors. The AI wave, the market shift, the next funding round, the offer is on the table for everyone. The dollar is real. But 95% of the room is still whispering to their neighbor, asking if it's a trick. It never was a trick. It was just a test of who moves first. 10 seconds. That's all you get. What are you doing with yours?

Alvin Foo

13,931 views • 2 months ago

A pilot put a passenger jet down on a river and got every person on board off alive. Within days, investigators were asking him what he had eaten for lunch. He told them a tuna sandwich. Then they asked whether he had had a candy bar. Whether he had had a soda. That is not small talk. There is a precise reason those 2 questions could have ended his career, and Tom Hanks lays it out in about 4 sentences. The pilot is Chesley Sullenberger. Hanks played him, spent time with him first, and says the man walked him through the entire script telling him which parts never happened. As Hanks was told it, the stake was money. If the board could establish that the pilot did 1 thing wrong, the insurance on an airplane he puts at 17 million dollars would not have to be paid out. Every question in that room was built for that. It went on for the better part of a year. In the film it takes days, because films cannot afford a year. What was happening to him has a name in psychology. Baruch Fischhoff described it in 1975 and called it creeping determinism. Once you know how something ended, the ending looks like it was always coming, and every road you did not take looks like it was sitting right there, clearly marked, the whole time. Which is why the question "why did you not do this instead" is close to unanswerable. It is asked by people who know the ending, to a person who did not have it. There is a second effect underneath, well documented in aviation and in medicine. The person at the centre of the investigation frequently comes out of it in worse condition than the people he saved. Researchers gave that person a name. He is the second victim. Hanks is 67 when he says this, and what he admires is not the landing. It is a distinction he makes in 1 sentence near the end, and it is the whole reason he told the story at all. He saved everyone on that plane. Then he spent a year proving he had not eaten a candy bar.
12:36

A pilot put a passenger jet down on a river and got every person on board off alive. Within days, investigators were asking him what he had eaten for lunch. He told them a tuna sandwich. Then they asked whether he had had a candy bar. Whether he had had a soda. That is not small talk. There is a precise reason those 2 questions could have ended his career, and Tom Hanks lays it out in about 4 sentences. The pilot is Chesley Sullenberger. Hanks played him, spent time with him first, and says the man walked him through the entire script telling him which parts never happened. As Hanks was told it, the stake was money. If the board could establish that the pilot did 1 thing wrong, the insurance on an airplane he puts at 17 million dollars would not have to be paid out. Every question in that room was built for that. It went on for the better part of a year. In the film it takes days, because films cannot afford a year. What was happening to him has a name in psychology. Baruch Fischhoff described it in 1975 and called it creeping determinism. Once you know how something ended, the ending looks like it was always coming, and every road you did not take looks like it was sitting right there, clearly marked, the whole time. Which is why the question "why did you not do this instead" is close to unanswerable. It is asked by people who know the ending, to a person who did not have it. There is a second effect underneath, well documented in aviation and in medicine. The person at the centre of the investigation frequently comes out of it in worse condition than the people he saved. Researchers gave that person a name. He is the second victim. Hanks is 67 when he says this, and what he admires is not the landing. It is a distinction he makes in 1 sentence near the end, and it is the whole reason he told the story at all. He saved everyone on that plane. Then he spent a year proving he had not eaten a candy bar.

Stefan.

209,483 views • 4 days ago

Dredd calls out Lana Rhoades for "biting the hand that fed her": "When you see something where Lana Rhoades had this high profile, brief time spent in the porn world, and then she kind of pivots out of it... and she kind of regrets it." "What is it like for you when you see somebody say that? Because I'm surrounded by girls who look at porn like, 'This is the thing that changed my life'... Most of the girls I know would say that their happiness with them having done porn is a 10 out of 10." "I think that's like a common thing with a lot of girls, not all, but a lot of them who once they step out, they want to talk in a negative way about it. "And it's like, 'Damn, don't do that, it's bad, this, that, whatever.'" "It was funny for me with her, I shot with her a few times... She was great, you know, when she was shooting... She was the real deal, there's no doubt on that. So I'm always surprised when a girl..."
1:18

Dredd calls out Lana Rhoades for "biting the hand that fed her": "When you see something where Lana Rhoades had this high profile, brief time spent in the porn world, and then she kind of pivots out of it... and she kind of regrets it." "What is it like for you when you see somebody say that? Because I'm surrounded by girls who look at porn like, 'This is the thing that changed my life'... Most of the girls I know would say that their happiness with them having done porn is a 10 out of 10." "I think that's like a common thing with a lot of girls, not all, but a lot of them who once they step out, they want to talk in a negative way about it. "And it's like, 'Damn, don't do that, it's bad, this, that, whatever.'" "It was funny for me with her, I shot with her a few times... She was great, you know, when she was shooting... She was the real deal, there's no doubt on that. So I'm always surprised when a girl..."

nexa_sama

33,444 views • 8 days ago

SemiAnalysis' Jordan Nanos reveals the OpenAI engineers scrolling their own kernel code had no idea what it did line by line, and it didn't matter because the AI does "It's not slop, because it performs, but it's literally just AI-generated assembly basically puked out in Gluon, this low-level kernel programming language that they built on top of Triton, which uses this really interesting programming model." "The point is, the guys who were scrolling through this code with us, it was kind of clear that they know a whole bunch about hardware, they know a whole bunch about the concepts in the system, and they just have no idea what this MLA kernel that they're showing us for DeepSeek actually does." "You can go line by line and it's like nope, nope, nope." "But it doesn't matter. The AI understands it, the AI tests it, and you see the results. It produces correct kernels that perform really well." "And I think this is just a sign of what's to come again." "That actual code is not necessarily something that a human has to reason about deeply, if the AI knows how to manipulate the data movement and the processing elements on the hardware that you've given it." ________ More takeaways from OpenAI's Jalapeno development:
1:11

SemiAnalysis' Jordan Nanos reveals the OpenAI engineers scrolling their own kernel code had no idea what it did line by line, and it didn't matter because the AI does "It's not slop, because it performs, but it's literally just AI-generated assembly basically puked out in Gluon, this low-level kernel programming language that they built on top of Triton, which uses this really interesting programming model." "The point is, the guys who were scrolling through this code with us, it was kind of clear that they know a whole bunch about hardware, they know a whole bunch about the concepts in the system, and they just have no idea what this MLA kernel that they're showing us for DeepSeek actually does." "You can go line by line and it's like nope, nope, nope." "But it doesn't matter. The AI understands it, the AI tests it, and you see the results. It produces correct kernels that perform really well." "And I think this is just a sign of what's to come again." "That actual code is not necessarily something that a human has to reason about deeply, if the AI knows how to manipulate the data movement and the processing elements on the hardware that you've given it." ________ More takeaways from OpenAI's Jalapeno development:

Fireside Alpha

4,511,744 views • 6 days ago

elon musk grabbed the source code openai open-sourced by accident, rewrote it in rust over a weekend, and shipped it as a free coding agent that does everything $200/mo chatgpt pro does. why pay $200 to openai and $200 to claude when this runs for $8 the swarm above is one weekend of exactly that: thousands of agents pouring through four endpoints, three paid seats billing $1.80 a task while the free fork bills $0. musk co-founded openai, walked out, and when they left codex on github under a permissive license, he forked it, stamped grok on it, and gave it away what the free version does that the $200 seat charges for: the agent · openai's own engine -> it reads your repo, writes patches, runs your tests, and loops until they pass, exactly like codex -> because under the hood it is codex, just faster and free. you are paying $200 for the paid skin of a tool now sitting on github the license · apache-2.0, un-revocable -> free to use, free to fork, free to ship inside your own product with zero strings -> openai cannot pull it back. musk made sure the license is the kind that never expires the switch · one line, no new tools -> point it at any openai-compatible or claude-compatible endpoint, including an $8 kimi backend -> same terminal, same workflow, gpt-5.6 and opus 5 just quietly lose the seat the bill · $400 down to $8 -> chatgpt pro plus claude max is $400 a month. the free agent plus an $8 kimi key does the same daily work -> that is a 98% cut, built out of openai's own source code, handed to you by the guy suing them here is the part they will fight me on: openai did not lose this to a better model, they lost it to their own license and an enemy with a weekend free. the $200 was never the tool, it was the toll, and musk just put openai's own logo on the road around it drop your $400/mo ai stack to $8. the run above is openai's own agent, rewritten free, doing the job it bills $200 a month for. the full breakdown is in the article below
0:26

elon musk grabbed the source code openai open-sourced by accident, rewrote it in rust over a weekend, and shipped it as a free coding agent that does everything $200/mo chatgpt pro does. why pay $200 to openai and $200 to claude when this runs for $8 the swarm above is one weekend of exactly that: thousands of agents pouring through four endpoints, three paid seats billing $1.80 a task while the free fork bills $0. musk co-founded openai, walked out, and when they left codex on github under a permissive license, he forked it, stamped grok on it, and gave it away what the free version does that the $200 seat charges for: the agent · openai's own engine -> it reads your repo, writes patches, runs your tests, and loops until they pass, exactly like codex -> because under the hood it is codex, just faster and free. you are paying $200 for the paid skin of a tool now sitting on github the license · apache-2.0, un-revocable -> free to use, free to fork, free to ship inside your own product with zero strings -> openai cannot pull it back. musk made sure the license is the kind that never expires the switch · one line, no new tools -> point it at any openai-compatible or claude-compatible endpoint, including an $8 kimi backend -> same terminal, same workflow, gpt-5.6 and opus 5 just quietly lose the seat the bill · $400 down to $8 -> chatgpt pro plus claude max is $400 a month. the free agent plus an $8 kimi key does the same daily work -> that is a 98% cut, built out of openai's own source code, handed to you by the guy suing them here is the part they will fight me on: openai did not lose this to a better model, they lost it to their own license and an enemy with a weekend free. the $200 was never the tool, it was the toll, and musk just put openai's own logo on the road around it drop your $400/mo ai stack to $8. the run above is openai's own agent, rewritten free, doing the job it bills $200 a month for. the full breakdown is in the article below

starmex

110,709 views • 9 days ago