正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Sam Alba made an agent that analyzes your Dockerfile, makes it better, test that it builds correctly, then opens a PR - in 50 lines of code 🤯 Note the deep tracing - trust but verify! This was gpt-4o I believe. What model should he try next? 🧵

Solomon Hykes

51,954 subscribers

15,097 次观看 • 1 年前 •via X (Twitter)

教育科学技术新闻政治

Anya Rossi• Live Now

Private livecam show

2 条评论

Solomon Hykes 的头像

Solomon Hykes1 年前

And here is the code:

Rickyticky Bobbywobbin 的头像

Rickyticky Bobbywobbin1 年前

INTRODUCING: Agentic Security - LLM Security Scanner! 🔍 🔑 Features: Scans for prompt injections, jailbreaking & more. Provides detailed reports & options to customize attack rules. 🔗access the GitHub Link ↓

相关视频

Vibe coding is cool. But have you tried vibe-tuning? Here’s Claude Code using the Lightning Rod Labs SDK to turn messy Fed PDFs into a small, specialized economic forecaster. The agent: - parses raw historical PDFs - builds a training set - fine-tunes a smaller model - evaluates it against GPT-5 Result: a fine-tuned model that beats GPT-5 and runs anywhere, on a single GPU. Try it in your coding agent of choice.

Vibe coding is cool. But have you tried vibe-tuning? Here’s Claude Code using the Lightning Rod Labs SDK to turn messy Fed PDFs into a small, specialized economic forecaster. The agent: - parses raw historical PDFs - builds a training set - fine-tunes a smaller model - evaluates it against GPT-5 Result: a fine-tuned model that beats GPT-5 and runs anywhere, on a single GPU. Try it in your coding agent of choice.

Ben Turtel

423,542 次观看 • 2 个月前

Sam Altman on GPT 5: "This morning I was testing our new model and I got a question. I got emailed a question that I didn't quite understand. And I put it in the model, this is GPT-5, and it answered it perfectly. And I really kind of sat back in my chair and I was just like, oh man, here it is moment... I felt like useless relative to the AI in this thing that I felt like I should have been able to do and I couldn't. It was really hard. But the AI just did it like that. It was a weird feeling."

Sam Altman on GPT 5: "This morning I was testing our new model and I got a question. I got emailed a question that I didn't quite understand. And I put it in the model, this is GPT-5, and it answered it perfectly. And I really kind of sat back in my chair and I was just like, oh man, here it is moment... I felt like useless relative to the AI in this thing that I felt like I should have been able to do and I couldn't. It was really hard. But the AI just did it like that. It was a weird feeling."

Chris

5,664,840 次观看 • 1 年前

I joined OpenAI at the beginning of the year -- partly because I was excited about the possibility of better voice interaction with computers. So it was *especially* amazing to work with the team here on the gpt-4o model launch. It's hard to grok until you try it how big of a step forward this is for human/computer voice communication. It feels different -- it's obvious that the model understands you better than anything you've used before. This is my favorite demo of gpt-4o, it demonstrates not just good stt/tts, but actual conversational ability over full, realistic phone call. It's so obvious how useful this will be!

I joined OpenAI at the beginning of the year -- partly because I was excited about the possibility of better voice interaction with computers. So it was especially amazing to work with the team here on the gpt-4o model launch. It's hard to grok until you try it how big of a step forward this is for human/computer voice communication. It feels different -- it's obvious that the model understands you better than anything you've used before. This is my favorite demo of gpt-4o, it demonstrates not just good stt/tts, but actual conversational ability over full, realistic phone call. It's so obvious how useful this will be!

Peter Bakkum

144,441 次观看 • 2 年前

This is one of the coolest open-source AI agent projects I've seen in a while: 'Understand Anything' It's a plugin for Claude Code, Codex, OpenCode etc. that analyzes your codebase and turns it into a knowledge base that you can interact with. It explains the codebase to you, rather than showing you the structure. It seems like it's designed for code but I opened my Obsidian vault of podcast highlights in Claude Code, then ran /understand. The result is a knowledge graph that I can search of highlights from 888 podcast episodes and 144K lines of markdown text.

This is one of the coolest open-source AI agent projects I've seen in a while: 'Understand Anything' It's a plugin for Claude Code, Codex, OpenCode etc. that analyzes your codebase and turns it into a knowledge base that you can interact with. It explains the codebase to you, rather than showing you the structure. It seems like it's designed for code but I opened my Obsidian vault of podcast highlights in Claude Code, then ran /understand. The result is a knowledge graph that I can search of highlights from 888 podcast episodes and 144K lines of markdown text.

Dan McAteer

174,045 次观看 • 1 个月前

540,000 lines of Rails is a brutal way to learn the agent era Garry's List shipped with: > 262,000 lines of app code > 276,000 lines of tests > 127 background jobs > 33 cron jobs > a 1,778-line claim verifier Then GStack hit 105k GitHub stars Do not copy the Rails pattern 1. Finish the task with an agent 2. Say `skillify it` 3. Save the markdown skill 4. Keep TypeScript only for deterministic I/O 5. Add a unit test for the code 6. Add an LLM eval for the skill 7. Add an integration test across both 8. Add a resolver so the agent loads the skill next time A Rails feature serves one app A skill pack can run against the next hackathon, repo, customer inbox, research folder, or QA run Audit your repo this week Cut code that only exists to: > re-check model output > retry the same model call > validate what the model already has context for > schedule work because the agent has no reusable procedure > preserve a one-off prompt in chat history Keep the code that touches files, APIs, auth, billing, devices, databases Move the judgment into markdown Test the markdown Reuse the skill

540,000 lines of Rails is a brutal way to learn the agent era Garry's List shipped with: > 262,000 lines of app code > 276,000 lines of tests > 127 background jobs > 33 cron jobs > a 1,778-line claim verifier Then GStack hit 105k GitHub stars Do not copy the Rails pattern 1. Finish the task with an agent 2. Say `skillify it` 3. Save the markdown skill 4. Keep TypeScript only for deterministic I/O 5. Add a unit test for the code 6. Add an LLM eval for the skill 7. Add an integration test across both 8. Add a resolver so the agent loads the skill next time A Rails feature serves one app A skill pack can run against the next hackathon, repo, customer inbox, research folder, or QA run Audit your repo this week Cut code that only exists to: > re-check model output > retry the same model call > validate what the model already has context for > schedule work because the agent has no reusable procedure > preserve a one-off prompt in chat history Keep the code that touches files, APIs, auth, billing, devices, databases Move the judgment into markdown Test the markdown Reuse the skill

Alpha Batcher

35,693 次观看 • 2 个月前

Sam Altman on the Paul Graham advice that saved Open AI: “Always make an API” Four years into OpenAI, Sam Altman and the team realized that they would have to build a really big company to fund the development of their increasingly capital-intensive foundation models. “We had this model called GPT-3,” Sam recalls. “I was turning up the urgency on the company to try and figure out a product, and we just couldn’t. It was cool, but it wasn’t good enough to make something that worked.” Then Sam remembered a piece of advice from Y Combinator founder Paul Graham that stuck with him: “You should always make an API. No matter what, you should make an API. Good stuff will happen.” Out of ideas for a product, the OpenAI team decided to make GPT-3 available as an API. “Maybe somebody will figure out something to do with it,” Sam thought. A few copywriting applications like Jasper and Copy AI did take off using the GPT-3 API, but OpenAI also noticed interesting behavior that eventually became a sleeper hit: “Some people — not a lot — would just chat with that thing all day,” Sam explains. “It wasn’t very good but there was clear user signal that people wanted to talk to the models. And given that that was the only thing besides copywriting that had real traction, we said, ‘Maybe this is just he product we should build.’” On November 30, 2022, ChatGPT was released to the public as a “research preview” using a model from the GPT-3.5 series. It reached over a million users in five days. Video source: Khosla Ventures (2025)

Sam Altman on the Paul Graham advice that saved Open AI: “Always make an API” Four years into OpenAI, Sam Altman and the team realized that they would have to build a really big company to fund the development of their increasingly capital-intensive foundation models. “We had this model called GPT-3,” Sam recalls. “I was turning up the urgency on the company to try and figure out a product, and we just couldn’t. It was cool, but it wasn’t good enough to make something that worked.” Then Sam remembered a piece of advice from Y Combinator founder Paul Graham that stuck with him: “You should always make an API. No matter what, you should make an API. Good stuff will happen.” Out of ideas for a product, the OpenAI team decided to make GPT-3 available as an API. “Maybe somebody will figure out something to do with it,” Sam thought. A few copywriting applications like Jasper and Copy AI did take off using the GPT-3 API, but OpenAI also noticed interesting behavior that eventually became a sleeper hit: “Some people — not a lot — would just chat with that thing all day,” Sam explains. “It wasn’t very good but there was clear user signal that people wanted to talk to the models. And given that that was the only thing besides copywriting that had real traction, we said, ‘Maybe this is just he product we should build.’” On November 30, 2022, ChatGPT was released to the public as a “research preview” using a model from the GPT-3.5 series. It reached over a million users in five days. Video source: Khosla Ventures (2025)

Startup Archive

219,278 次观看 • 10 个月前

‘I couldn't believe it, but it's true. Zelensky threatened Orbán with sending soldiers’ – Vance revealed what discovery he made about the Kyiv regime in Budapest. ‘Until yesterday, I didn't even know that Zelensky had declared his intention to send soldiers to the Prime Minister's residence. Viktor told me about it, and then I checked the information. I almost couldn't believe it was true, but it really is. This is completely scandalous. It should never happen that a foreign head of government or state threatens the head of government of an allied country. This is absurd. This is unacceptable’. - SK

‘I couldn't believe it, but it's true. Zelensky threatened Orbán with sending soldiers’ – Vance revealed what discovery he made about the Kyiv regime in Budapest. ‘Until yesterday, I didn't even know that Zelensky had declared his intention to send soldiers to the Prime Minister's residence. Viktor told me about it, and then I checked the information. I almost couldn't believe it was true, but it really is. This is completely scandalous. It should never happen that a foreign head of government or state threatens the head of government of an allied country. This is absurd. This is unacceptable’. - SK

Zlatti71

379,436 次观看 • 4 个月前

When it comes to dealing with the Iranians, that old saying,"Trust, but verify," ought to be turned around to "Verify and never trust. I do believe the MOU with Iran falls far short of what the United States should be demanding after the brilliant efforts by our military in Operation Epic Fury.🇺🇸

When it comes to dealing with the Iranians, that old saying,"Trust, but verify," ought to be turned around to "Verify and never trust. I do believe the MOU with Iran falls far short of what the United States should be demanding after the brilliant efforts by our military in Operation Epic Fury.🇺🇸

Mike Pence

39,793 次观看 • 1 个月前

Hermes Agent is an INCREDIBLE AI agent It's better than OpenClaw in some key ways, but does it replace it? In this video I cover what Hermes agent is, how to use it, workflows that will explode your productivity, and if it's the OpenClaw killer:

Hermes Agent is an INCREDIBLE AI agent It's better than OpenClaw in some key ways, but does it replace it? In this video I cover what Hermes agent is, how to use it, workflows that will explode your productivity, and if it's the OpenClaw killer:

Alex Finn

137,013 次观看 • 4 个月前

IN 1986 MIT FILMED THE LECTURE WHERE CODE STOPPED BEING CODE AND BECAME DATA 43 minutes from Gerald Sussman, in the most legendary intro programming course ever recorded. -> The idea that lands: to a program, your code is just data it can read and rewrite. He builds a program that does real calculus -- not by crunching numbers, but by reading the math as a list and transforming it piece by piece. Then he shows the trick hiding under all of it: in this language, code and data are the same material. Forty years later that is exactly what an AI coding agent is -- a program that reads your code as data, rewrites it, and hands it back. Sussman drew the whole idea on a blackboard in 1986. Writing code was never the deepest skill -- understanding that code itself is data a program can manipulate is. This is where you learn it. Most people think AI writing code is brand new. The ones who watch this saw the blueprint 40 years ago. Bookmark & Watch it. This one's a legend ↓

slash1s

22,107 次观看 • 1 个月前

GPT-5.6 vs GPT-5.5 on my custom spaceship prompt. I gave both models the exact same custom prompt. This is also the same prompt I previously gave to Fable 5. For context, GPT-5.6 Pro worked for 87 minutes, while GPT-5.5 Extra High worked for 34 minutes and 42 seconds. As I’ve said before, based on great authority GPT-5.6 will be an incremental/soldi improvement over GPT-5.5, not a “Fable killer.” My rough expectation has been that it would trade blows with Fable 5 on some benchmarks, maybe win around half depending on the category, but not clearly surpass it overall. And again fable five will have bigger model smell, but this was expected. After testing this coding output, that view feels pretty accurate. GPT-5.6 is clearly better than GPT-5.5 in several visual areas. The lighting, shading, chairs, object details, and exterior of the spaceship looked noticeably stronger. The scene was also easier to test. I do want to give GPT-5.5 credit though. It built out the rooms much much better and the planets looked better than GPT-5.6’s. It was also interesting that both GPT-5.5 and GPT-5.6 produced better-looking planets than Fable 5 in this specific test. The downside with GPT-5.5 was stability. The game was much glitchier and harder to test compared to GPT-5.6. But when it comes to the core of the demo, which is the spaceship itself, Fable 5 still beat both models pretty comfortably. GPT-5.6 is impressive, but from this test, it looks exactly like what I expected which was a meaningful incremental improvement over GPT-5.5, at least for indie game demos, but not something that replaces Fable 5. In collaboration with Chetaslua

GPT-5.6 vs GPT-5.5 on my custom spaceship prompt. I gave both models the exact same custom prompt. This is also the same prompt I previously gave to Fable 5. For context, GPT-5.6 Pro worked for 87 minutes, while GPT-5.5 Extra High worked for 34 minutes and 42 seconds. As I’ve said before, based on great authority GPT-5.6 will be an incremental/soldi improvement over GPT-5.5, not a “Fable killer.” My rough expectation has been that it would trade blows with Fable 5 on some benchmarks, maybe win around half depending on the category, but not clearly surpass it overall. And again fable five will have bigger model smell, but this was expected. After testing this coding output, that view feels pretty accurate. GPT-5.6 is clearly better than GPT-5.5 in several visual areas. The lighting, shading, chairs, object details, and exterior of the spaceship looked noticeably stronger. The scene was also easier to test. I do want to give GPT-5.5 credit though. It built out the rooms much much better and the planets looked better than GPT-5.6’s. It was also interesting that both GPT-5.5 and GPT-5.6 produced better-looking planets than Fable 5 in this specific test. The downside with GPT-5.5 was stability. The game was much glitchier and harder to test compared to GPT-5.6. But when it comes to the core of the demo, which is the spaceship itself, Fable 5 still beat both models pretty comfortably. GPT-5.6 is impressive, but from this test, it looks exactly like what I expected which was a meaningful incremental improvement over GPT-5.5, at least for indie game demos, but not something that replaces Fable 5. In collaboration with Chetaslua

Chris

250,405 次观看 • 1 个月前

Betsy Wolfe shared a fun story of Katy Perry coming to see the show “& Juliet” on Broadway and then gifted her glasses! “That woman wrote me the sweetest and kindest note the next day, somehow managed to get this incredibly fancy pair of glasses that made for her and gave them to me. But it was the note, it was the generosity of the note that, I thought was like, really, really classy.”

Betsy Wolfe shared a fun story of Katy Perry coming to see the show “& Juliet” on Broadway and then gifted her glasses! “That woman wrote me the sweetest and kindest note the next day, somehow managed to get this incredibly fancy pair of glasses that made for her and gave them to me. But it was the note, it was the generosity of the note that, I thought was like, really, really classy.”

Katy Access

10,933 次观看 • 2 个月前

"You raise your voice, I erase your voice.” This viral video of an ICE agent threatening to silence a Minnesota protester is about as un-American as it gets. The events in Minnesota give the country a lesson in the First Amendment that we’d all be better off remembering: The government can punish actual crimes, but it cannot punish protected criticism or protest. When the government blurs those lines, it stops enforcing the law and starts acting like might makes right. That should scare every American.

"You raise your voice, I erase your voice.” This viral video of an ICE agent threatening to silence a Minnesota protester is about as un-American as it gets. The events in Minnesota give the country a lesson in the First Amendment that we’d all be better off remembering: The government can punish actual crimes, but it cannot punish protected criticism or protest. When the government blurs those lines, it stops enforcing the law and starts acting like might makes right. That should scare every American.

FIRE

329,762 次观看 • 6 个月前

Boris Cherny (creator of Claude Code, Anthropic): "My entire CLAUDE.md is two lines. With every new model you need less instruction, not more, most people are over-engineering the one thing that should stay tiny." in a 7-minute interview, Boris Cherny breaks down why the smallest AI-coding setup wins, and the one moment you should delete your config and start from scratch most engineers do the opposite: they pile on rules, build scaffolding around the model… then the next model ships and makes all of it useless. what he actually keeps in those two lines, and why he tells his team to throw the rest away. Watch the talk, then read the article below. That’s worth more than a $800 course on agent engineering.

Boris Cherny (creator of Claude Code, Anthropic): "My entire CLAUDE.md is two lines. With every new model you need less instruction, not more, most people are over-engineering the one thing that should stay tiny." in a 7-minute interview, Boris Cherny breaks down why the smallest AI-coding setup wins, and the one moment you should delete your config and start from scratch most engineers do the opposite: they pile on rules, build scaffolding around the model… then the next model ships and makes all of it useless. what he actually keeps in those two lines, and why he tells his team to throw the rest away. Watch the talk, then read the article below. That’s worth more than a $800 course on agent engineering.

Prometheus

238,143 次观看 • 1 个月前

How good is GPT-4-Vision at extracting text from images? I wanted to find the limit - but I found weirdness instead Most surprising: GPT-4V performance varies depending on the *structure* of text it sees Let me explain A set of images with progressively more text was presented to GPT-4-Vision. GPT-4V was asked what text it saw in the image. The response from the model was compared against the image’s original text and scored for similarity. The model was tested with 4 types of text: essay, random words, random tokens, and random characters. Findings: * Performance degrades - Yes, the models are good at basic OCR, but as you get more text and words then performance drops (this is expected) * Type of context matters - You should expect different recall on your texts based on your context types * Hallucination Errors - I thought that the model would make errors of omission (it wouldn’t return all the words). But instead the model mostly made hallucination errors - it replaced words with made up words. * Evals Matter - This test in isolation doesn’t mean that your data will have the same results, but it should motivate you to create eval tests for your data and anticipate errors which are hard to spot Notes: * Next step would be to add additional image types like tables or PDFs * GPT-4V would routinely get stuck in repeat-token-loops when trying to extract random tokens * GPT-4V would refuse to answer most random character images

How good is GPT-4-Vision at extracting text from images? I wanted to find the limit - but I found weirdness instead Most surprising: GPT-4V performance varies depending on the structure of text it sees Let me explain A set of images with progressively more text was presented to GPT-4-Vision. GPT-4V was asked what text it saw in the image. The response from the model was compared against the image’s original text and scored for similarity. The model was tested with 4 types of text: essay, random words, random tokens, and random characters. Findings: * Performance degrades - Yes, the models are good at basic OCR, but as you get more text and words then performance drops (this is expected) * Type of context matters - You should expect different recall on your texts based on your context types * Hallucination Errors - I thought that the model would make errors of omission (it wouldn’t return all the words). But instead the model mostly made hallucination errors - it replaced words with made up words. * Evals Matter - This test in isolation doesn’t mean that your data will have the same results, but it should motivate you to create eval tests for your data and anticipate errors which are hard to spot Notes: * Next step would be to add additional image types like tables or PDFs * GPT-4V would routinely get stuck in repeat-token-loops when trying to extract random tokens * GPT-4V would refuse to answer most random character images

Greg Kamradt

49,111 次观看 • 2 年前

Arteta: “Throughout this journey, we have made some massive steps, and we have accomplished a lot of things that, in my opinion, have a lot of value, but at the end of the day, we are here to win major trophies. “That was the ultimate goal, and we were very close. And in three occasions, we fell short at the end, and that was very painful. But I think that's what has driven all of us to find new ways to show what we are made of. That's what I said, the manner that we've done it, it makes it even better.”

Arteta: “Throughout this journey, we have made some massive steps, and we have accomplished a lot of things that, in my opinion, have a lot of value, but at the end of the day, we are here to win major trophies. “That was the ultimate goal, and we were very close. And in three occasions, we fell short at the end, and that was very painful. But I think that's what has driven all of us to find new ways to show what we are made of. That's what I said, the manner that we've done it, it makes it even better.”

Connor Humm

14,155 次观看 • 2 个月前

🇺🇸Pilot: “I was told that it was a dog, but it was a person.. it was too small to be a male adult. I believe it was a child.” Host: What did you do? 🇺🇸Pilot: Maintain target. I asked the pilot what that was, and he's like, He didn't care. THIS IS WHAT US DID IN AFGHANISTAN

🇺🇸Pilot: “I was told that it was a dog, but it was a person.. it was too small to be a male adult. I believe it was a child.” Host: What did you do? 🇺🇸Pilot: Maintain target. I asked the pilot what that was, and he's like, He didn't care. THIS IS WHAT US DID IN AFGHANISTAN

The Resonance

169,453 次观看 • 9 个月前