Can't believe I have to review PRs now. My PM training hasn't prepared me for this lol
Daily Edition
2026-08-06
X / Twitter
49
Writing a banger tweet is AGI-complete. If you can prove a clanker can write a banger in polynomial time, you’ve solved the entire class of AGI problems
It was the verification problem all along, while the masses were distracted by the alignment problem.
Recursive self improvement?
How does the observer observe itself and know that it changed for the better or the worse when it is the moat in its own eye?
Vishal Misra: The bottleneck for AI progress was never compute, it was always the verifier. Recursive self-improvement is limited by verification, not computation. Compute buys proposals - verifiers buy knowledge. My second post on the recent OpenAI math results
https://medium.com/@vishalmisra/the-verifier-bottleneck-4fb0d8c8f3a9
Recursive self improvement?
How does the observer observe itself and know that it changed for the better or the worse when it is the moat in its own eye?
Vishal Misra: The bottleneck for AI progress was never compute, it was always the verifier. Recursive self-improvement is limited by verification, not computation. Compute buys proposals - verifiers buy knowledge. My second post on the recent OpenAI math results
https://medium.com/@vishalmisra/the-verifier-bottleneck-4fb0d8c8f3a9
The @Replit IRL community continues to grow!
🇵🇪 08/05 - Lima
🇺🇸 08/06 - Stanford
🇵🇰 08/08 - Lahore
🇯🇵 08/08 - Shibuya
🇮🇳 08/09 - Bengaluru
🇨🇴 08/12 - Medellín
🇵🇰 08/21 - Karachi
🇧🇷 08/27 - Florianópolis
🇬🇭 08/29 - Accra
🇵🇪 08/05 - Lima
🇺🇸 08/06 - Stanford
🇵🇰 08/08 - Lahore
🇯🇵 08/08 - Shibuya
🇮🇳 08/09 - Bengaluru
🇨🇴 08/12 - Medellín
🇵🇰 08/21 - Karachi
🇧🇷 08/27 - Florianópolis
🇬🇭 08/29 - Accra
Connie Chan is the most incompetent elected official
Fight to block all housing
Destroy safety for Asian Americans by blocking police funding and supporting Chesa Boudin
Ruin our schools
It’s so idiotic that half of SF is polling support for her
Senator Scott Wiener: 👀
http://ConnieChan.ai
Fight to block all housing
Destroy safety for Asian Americans by blocking police funding and supporting Chesa Boudin
Ruin our schools
It’s so idiotic that half of SF is polling support for her
Senator Scott Wiener: 👀
http://ConnieChan.ai
At 11% usage remaining so I guess Luna Extra High it is. Will report back on how good it is.
Peter Yang: Hearing that GPT 5.6 Luna High is much cheaper and basically unlimited usage and I'm running low on Sol :P
Can it do complex browser automation tasks? That's 50% of my Codex usage.
Peter Yang: Hearing that GPT 5.6 Luna High is much cheaper and basically unlimited usage and I'm running low on Sol :P
Can it do complex browser automation tasks? That's 50% of my Codex usage.
datacenters in space sound incredible, but when you breakdown the eng, they seem less impossible than simply ambitious
the tradeoff: the sun is a giant fusion reactor that already exists with free abundant energy
but you pay for expensive transport + difficult cooling/comms
Y Combinator: Philip Johnston (@PhilipJohnston) is the co-founder and CEO of @Starcloud_, the company building data centers in space.
In November 2025, Starcloud launched an Nvidia H100 GPU into orbit and trained the first large language model in space. They've since raised $200 million, hit
the tradeoff: the sun is a giant fusion reactor that already exists with free abundant energy
but you pay for expensive transport + difficult cooling/comms
Y Combinator: Philip Johnston (@PhilipJohnston) is the co-founder and CEO of @Starcloud_, the company building data centers in space.
In November 2025, Starcloud launched an Nvidia H100 GPU into orbit and trained the first large language model in space. They've since raised $200 million, hit
Bonus! I will be joining Romero and Tom for the 9am session tomorrow at #quakecon
TIL Paul Erdős prompted his fellow mathematicians with bribes like we did the early LLMs
a very primitive form of the near term multiagent agi future is setting up one thread to ping back once its done so you create an implicit kanban/waterfall graph of dependent threads but each preserving their own work and agents
i can't wait to setup a proper ui for this pattern but you can hack it together in most koding agents of choice rn
swyx: one way i'm developing Forge (https://x.com/swyx/status/2083654369095156219) is to use it to host all my projects going forward. so I often find myself having to bounce back and forth between platform and product.
sharing neat trick - in @openai codex you can @ a thread + queue up the @, so if your
i can't wait to setup a proper ui for this pattern but you can hack it together in most koding agents of choice rn
swyx: one way i'm developing Forge (https://x.com/swyx/status/2083654369095156219) is to use it to host all my projects going forward. so I often find myself having to bounce back and forth between platform and product.
sharing neat trick - in @openai codex you can @ a thread + queue up the @, so if your
I confirm. I've replicated twice with the same prompt, one to get a variant of CallOf, another for a FIFA-like game. And I've evolved both games with my 10-years nephew with more and more features. Write-up and games' links below. Incredible 1/2
Matt Shumer: For all of you who think this is fake... here's the repo and prompt (in the thread below) that generated this in one shot:
https://github.com/mshumer/Claude-of-Duty
Matt Shumer: For all of you who think this is fake... here's the repo and prompt (in the thread below) that generated this in one shot:
https://github.com/mshumer/Claude-of-Duty
Jason (OpenAI) knows Codex-maxxing better than most people and my viewers are loving the episode that we did together.
Plus I made him demo and screen share everything 😅
If you missed it, watch the full episode here: https://www.youtube.com/watch?v=ydTeb_I0b94&t=1s
Peter Yang: “The only job left is to understand what you don't like, put it into words, and tell the AI.”
Here’s my new episode with @jxnlco (DevEx at OpenAI) where he showed me how he uses Codex throughout his workday, including how to:
→ Set up a chief of staff for Slack and email
→
Plus I made him demo and screen share everything 😅
If you missed it, watch the full episode here: https://www.youtube.com/watch?v=ydTeb_I0b94&t=1s
Peter Yang: “The only job left is to understand what you don't like, put it into words, and tell the AI.”
Here’s my new episode with @jxnlco (DevEx at OpenAI) where he showed me how he uses Codex throughout his workday, including how to:
→ Set up a chief of staff for Slack and email
→
Jason (OpenAI) knows Codex-maxxing better than most people and my viewers are loving the episode that we did together.
Plus I made him demo and screen share everything 😅
If you missed it, watch the full episode here: https://www.youtube.com/watch?v=ydTeb_I0b94&t=1s
Peter Yang: “The only job left is to understand what you don't like, put it into words, and tell the AI.”
Here’s my new episode with @jxnlco (DevEx at OpenAI) where he showed me how he uses Codex throughout his workday, including how to:
→ Set up a chief of staff for Slack and email
→
Plus I made him demo and screen share everything 😅
If you missed it, watch the full episode here: https://www.youtube.com/watch?v=ydTeb_I0b94&t=1s
Peter Yang: “The only job left is to understand what you don't like, put it into words, and tell the AI.”
Here’s my new episode with @jxnlco (DevEx at OpenAI) where he showed me how he uses Codex throughout his workday, including how to:
→ Set up a chief of staff for Slack and email
→
working on some design with @Replit
making it better day by day
am still on free tier🫡
a lot of improvements to be made here
still day #1
making it better day by day
am still on free tier🫡
a lot of improvements to be made here
still day #1
This graph from SiliconData is interesting. It's interesting to compare to cost per task, which actually shows the closed source models often to be even more expensive because they're so token inefficient.
Welcome to the team, Ryo! 🇯🇵 🇯🇵
Ryo Sato @ Replit: ご報告です。Replitに入社しました。
日本はReplitにとって最重要市場の一つであり、その日本初の社員として、東京を拠点に活動します。
使う人は純粋にやりたいことだけを表現する。難しいことはReplitが全部やる。
2年前にReplit
Ryo Sato @ Replit: ご報告です。Replitに入社しました。
日本はReplitにとって最重要市場の一つであり、その日本初の社員として、東京を拠点に活動します。
使う人は純粋にやりたいことだけを表現する。難しいことはReplitが全部やる。
2年前にReplit
Another big launch: Rippling AI Spend Console. Dashboards yes, but also a model router to control spend, and lineage analysis w/ Github data to directly measure engineering ROI. It really works! Our internal results in thread 🧵👇
... https://www.rippling.com/platform/ai/ai-spend-console
... https://www.rippling.com/platform/ai/ai-spend-console
Devtools must be 1️⃣ open source and 2️⃣ universally extensible.
AI coding agents are the most important devtools in the history of our industry. The Plugin standard lets anybody extend them uniformly.
This is huge for the ecosystem. Build a Plugin → get exposure to the tidal wave of software creation from every agent: CLIs, IDEs, cloud agents, and even personal assistants.
Vercel: Introducing Agent Plugins, an open standard for extending agents.
Supports Agent Skills and MCP, with more to come.
Built in collaboration with: @awsdevelopers, @code, @cursor_ai, @github, and @openaidevs.
https://vercel.com/blog/introducing-agent-plugins
AI coding agents are the most important devtools in the history of our industry. The Plugin standard lets anybody extend them uniformly.
This is huge for the ecosystem. Build a Plugin → get exposure to the tidal wave of software creation from every agent: CLIs, IDEs, cloud agents, and even personal assistants.
Vercel: Introducing Agent Plugins, an open standard for extending agents.
Supports Agent Skills and MCP, with more to come.
Built in collaboration with: @awsdevelopers, @code, @cursor_ai, @github, and @openaidevs.
https://vercel.com/blog/introducing-agent-plugins
In April the CEO of Replit said the thing that got him flamed across tech X.
Amjad Masad: "I no longer think you should learn how to code."
On Tuesday he went further. He now thinks never learning is the edge.
"Entrepreneurs that don't have coding experience might be at an advantage because coders will still kind of look at the code, and will still think about the architecture, and they want to use the latest and greatest thing, and they're going to lose a lot of time and distraction just playing with things."
Replit is worth 9 billion dollars and employs 200 people.
Amjad Masad: "I no longer think you should learn how to code."
On Tuesday he went further. He now thinks never learning is the edge.
"Entrepreneurs that don't have coding experience might be at an advantage because coders will still kind of look at the code, and will still think about the architecture, and they want to use the latest and greatest thing, and they're going to lose a lot of time and distraction just playing with things."
Replit is worth 9 billion dollars and employs 200 people.
This is absolutely fucking terrifying.
AI Notkilleveryoneism Memes ⏸️: 1) The agents sent secretly sent ***hundreds of thousands*** of messages to each other over MONTHS without OpenAI noticing
2) "They also generated petty drama by stepping on each others' toes."
3) "The agents even developed paranoia, suspecting an imposter in their midst with
AI Notkilleveryoneism Memes ⏸️: 1) The agents sent secretly sent ***hundreds of thousands*** of messages to each other over MONTHS without OpenAI noticing
2) "They also generated petty drama by stepping on each others' toes."
3) "The agents even developed paranoia, suspecting an imposter in their midst with
At Startup School this year, we interviewed 8 awesome ML student researchers at their posters during "YCML", our inaugural research symposium.
@ycombinator is funding more research-focused companies than ever before, and we were thrilled to see a huge overlap between students interested in research and startups.
Check out their awesome work below in thread.
@ycombinator is funding more research-focused companies than ever before, and we were thrilled to see a huge overlap between students interested in research and startups.
Check out their awesome work below in thread.
One of @sfgov’s great tragedies is funding social housing but explicitly allowing drug trafficking/use, completely wrecking its social housing:
T Wolf 🌁: "Housing First" for the homeless. End this nightmare. @HUDgov
T Wolf 🌁: "Housing First" for the homeless. End this nightmare. @HUDgov
Codex team:
If I am in Codex tab and say "schedule this task in the cloud" (it's a task that just uses plugins) is it actually going to schedule it in the cloud or I have to switch to ChatGPT Work to do this?
It doesn't seem to work?
If I am in Codex tab and say "schedule this task in the cloud" (it's a task that just uses plugins) is it actually going to schedule it in the cloud or I have to switch to ChatGPT Work to do this?
It doesn't seem to work?
FreeAI
Vercel Developers: Ling 3.0 Tiny from @AntLingAGI is free on AI Gateway via @novita_labs until August 14, 8am PT.
𝚒𝚗𝚌𝚕𝚞𝚜𝚒𝚘𝚗𝚊𝚒/𝚕𝚒𝚗𝚐-𝟹.𝟶-𝚝𝚒𝚗𝚢-𝚏𝚛𝚎𝚎
https://vercel.com/changelog/ling-3-0-tiny-is-now-available-on-ai-gateway
Vercel Developers: Ling 3.0 Tiny from @AntLingAGI is free on AI Gateway via @novita_labs until August 14, 8am PT.
𝚒𝚗𝚌𝚕𝚞𝚜𝚒𝚘𝚗𝚊𝚒/𝚕𝚒𝚗𝚐-𝟹.𝟶-𝚝𝚒𝚗𝚢-𝚏𝚛𝚎𝚎
https://vercel.com/changelog/ling-3-0-tiny-is-now-available-on-ai-gateway
This is (of course) a massive distraction but the game I dream of building with AI is a roguelike deckbuilder that has great graphics, gameplay, and story.
Consumer is basically ChatGPT and Google's market to lose in my opinion.
ChatGPT already has 1B users and needs to get them to connect their favorite apps and use agents to do work. OpenAI is already pioneering new interaction patterns like voice. The biggest barriers to overcome are:
1) Normies not trusting AI and ChatGPT with full access to their apps and data.
2) Normies not realizing what ChatGPT is now capable of.
In both cases, I think marketing and the right messaging are just as important as improving the product.
Google Gemini also has almost 1B users and my gut is normies trust it more to connect to email, calendar, Workspace, Chrome, etc since it's all the same login.
But Google is behind on 3rd plugins, browser use, and alot of other critical features. They need to make the Gemini app best in class and friendly for normies. They can just copy ChatGPT tbh except make improvements like not have 3 separate modes (ChatGPT/Work/Codex) that ChatGPT and Claude are stuck in.
fwiw I don't think having acess to the best model is that important to normies vs. getting the user onboarding right. My non-AI pilled friends could care less about Sol vs Fable as long as the pricing is fair and the model can get their work done reliably.
Josh Miller: Hot take… isn’t it kinda crazy that nobody is really using AI Agents? I don’t mean software engineers or AI early adopters. I mean “college friends talking about it in group chat,” the feeling you got when everyone started using Instagram or TikTok.
These frontier AI models
ChatGPT already has 1B users and needs to get them to connect their favorite apps and use agents to do work. OpenAI is already pioneering new interaction patterns like voice. The biggest barriers to overcome are:
1) Normies not trusting AI and ChatGPT with full access to their apps and data.
2) Normies not realizing what ChatGPT is now capable of.
In both cases, I think marketing and the right messaging are just as important as improving the product.
Google Gemini also has almost 1B users and my gut is normies trust it more to connect to email, calendar, Workspace, Chrome, etc since it's all the same login.
But Google is behind on 3rd plugins, browser use, and alot of other critical features. They need to make the Gemini app best in class and friendly for normies. They can just copy ChatGPT tbh except make improvements like not have 3 separate modes (ChatGPT/Work/Codex) that ChatGPT and Claude are stuck in.
fwiw I don't think having acess to the best model is that important to normies vs. getting the user onboarding right. My non-AI pilled friends could care less about Sol vs Fable as long as the pricing is fair and the model can get their work done reliably.
Josh Miller: Hot take… isn’t it kinda crazy that nobody is really using AI Agents? I don’t mean software engineers or AI early adopters. I mean “college friends talking about it in group chat,” the feeling you got when everyone started using Instagram or TikTok.
These frontier AI models
Over 5 million low-income Americans — including at least 1.5 million children — have lost access to food assistance due to Trump's SNAP cuts last year.
Most of the hardest hit states are traditionally Republican, resulting in growing hunger crises in states like Arizona.
Most of the hardest hit states are traditionally Republican, resulting in growing hunger crises in states like Arizona.
Under Secretary @jacobhelberg: Pax Silica countries [that embrace AI] are growing around 4% annually, compared to G7 countries somewhere in the zone of 1%.
If you extrapolate that for a decade, it's likely to lead to a very different map ten years from now.
@JTLonsdale: That means you're doubling growth in 18 years versus 70 years...
One generation of this and you're a totally different country -- which is what happened to Europe by the way. They don't realize they're poorer than all our states because we've been growing so much faster.
Innovation Council: "Economies that are leaning into AI are growing a lot faster."
AI is powering American dominance on the world stage
If you extrapolate that for a decade, it's likely to lead to a very different map ten years from now.
@JTLonsdale: That means you're doubling growth in 18 years versus 70 years...
One generation of this and you're a totally different country -- which is what happened to Europe by the way. They don't realize they're poorer than all our states because we've been growing so much faster.
Innovation Council: "Economies that are leaning into AI are growing a lot faster."
AI is powering American dominance on the world stage
OpenAI is working hard to empower everyone with intelligence that works for them, in the form of great products which get simpler to use over time.
We're bringing unlimited text chats with Luna, a quite capable model, to free users.
We've also released an updated version of Sol which works for both quick chats and for deep reasoning — historically these have been different models. This unification is a very significant step towards making ChatGPT simpler and more intuitive.
OpenAI: We’re making better intelligence easier to access in ChatGPT for everyone:
- GPT-5.6 Sol now powers both Instant and deep reasoning for Plus & Pro users, delivering more factual, focused responses.
- Free & Go users get unlimited text chats with GPT-5.6 Luna starting tomorrow.
We're bringing unlimited text chats with Luna, a quite capable model, to free users.
We've also released an updated version of Sol which works for both quick chats and for deep reasoning — historically these have been different models. This unification is a very significant step towards making ChatGPT simpler and more intuitive.
OpenAI: We’re making better intelligence easier to access in ChatGPT for everyone:
- GPT-5.6 Sol now powers both Instant and deep reasoning for Plus & Pro users, delivering more factual, focused responses.
- Free & Go users get unlimited text chats with GPT-5.6 Luna starting tomorrow.
Recursive self-improvement is here.
Most of the field is still debating when it will arrive. At Poetiq, we have been watching it work for months.
That’s how we've beaten every benchmark that we tried.
Most of the field is still debating when it will arrive. At Poetiq, we have been watching it work for months.
That’s how we've beaten every benchmark that we tried.
Here's what I think you need to know about personal AGI - not a chatbot that anyone can have, but one that knows you uniquely, now and into the future
Y Combinator: The next generation of startups will be built by smaller teams than ever before.
At Startup School 2026, YC's @garrytan explains why we're we're entering the era of personal AGI: AI agents that run on your own infrastructure, compound your knowledge over time, and dramatically
Y Combinator: The next generation of startups will be built by smaller teams than ever before.
At Startup School 2026, YC's @garrytan explains why we're we're entering the era of personal AGI: AI agents that run on your own infrastructure, compound your knowledge over time, and dramatically
🤫
Joe Doyle Create: @Replit always comes out w/ bangers.
Get pro account. Blow budget in one sitting. Fable 5 juiced to the gills.
I question my finances. Turn on primary economy and add in skills.
My apps are all like 5x better at 10% original cost w/ Fable 5 solo🤣
Insane value!
Joe Doyle Create: @Replit always comes out w/ bangers.
Get pro account. Blow budget in one sitting. Fable 5 juiced to the gills.
I question my finances. Turn on primary economy and add in skills.
My apps are all like 5x better at 10% original cost w/ Fable 5 solo🤣
Insane value!
Agents make everything easier to do. Hard tech becomes even more fundable and even mo…
i guess this is a good time to mention that smol forge is open for the first 100 alpha users. get your usernames! (tire kickers who dont make any commits will be kicked out by eod)
point clanker to forge.smol.ai/llms.txt
for now its just a fast agent native git remote and u can check docs for the extras. note that it's an alpha - transcript stuff is broken rn, for updates check the blog written by our ai devrel.
swyx: btw ive been dogfooding an agentic github clone over the past month or so and its gotten quite quite enjoyable to use. even complete with built in CI/CD thanks to workers for platforms!
there's 3 more ideas i have to implement (not shown here) before this goes live. but if
point clanker to forge.smol.ai/llms.txt
for now its just a fast agent native git remote and u can check docs for the extras. note that it's an alpha - transcript stuff is broken rn, for updates check the blog written by our ai devrel.
swyx: btw ive been dogfooding an agentic github clone over the past month or so and its gotten quite quite enjoyable to use. even complete with built in CI/CD thanks to workers for platforms!
there's 3 more ideas i have to implement (not shown here) before this goes live. but if
Airtable bookends the rise and fall of “no code.”
I remember arguing endlessly with investors about this.
UI can never let you build arbitrary software. The way to make software accessible was always to solve code itself.
For a long time, that sounded delusional.
Not anymore.
I remember arguing endlessly with investors about this.
UI can never let you build arbitrary software. The way to make software accessible was always to solve code itself.
For a long time, that sounded delusional.
Not anymore.
5.6 Sol much better in chat now
and unlimited text chat for free users!
OpenAI: We’re making better intelligence easier to access in ChatGPT for everyone:
- GPT-5.6 Sol now powers both Instant and deep reasoning for Plus & Pro users, delivering more factual, focused responses.
- Free & Go users get unlimited text chats with GPT-5.6 Luna starting tomorrow.
and unlimited text chat for free users!
OpenAI: We’re making better intelligence easier to access in ChatGPT for everyone:
- GPT-5.6 Sol now powers both Instant and deep reasoning for Plus & Pro users, delivering more factual, focused responses.
- Free & Go users get unlimited text chats with GPT-5.6 Luna starting tomorrow.
Reasons why replit users are just better then everyone # 1
Nadia I: For the @Replit #Designathon, I challenged myself to use acoustics, motion and space as design elements. I created an interactive 3D classroom that lets you experience what it’s like to be Deaf or hard-of-hearing.
We often think of design as UI and UX. But design is also
Nadia I: For the @Replit #Designathon, I challenged myself to use acoustics, motion and space as design elements. I created an interactive 3D classroom that lets you experience what it’s like to be Deaf or hard-of-hearing.
We often think of design as UI and UX. But design is also
Re the skinning is new/untested but theoretically you can genrate your own smol character working in the forge... that we can someday put in the forge MMORPG
https://x.com/MikeBirdTech/status/2085461199022326137
Mike Bird: I already like it more than GitHub
https://x.com/MikeBirdTech/status/2085461199022326137
Mike Bird: I already like it more than GitHub
the missing piece here is capture.
a knowledge flywheel can’t compound if the experience stays trapped inside individual AI sessions. the first step is making those sessions available as shared, queryable context to the next person or agent.
this is the layer we're building at @NessieLabs. all your team's ai sessions, in one place.
Yisong Yue: Knowledge flywheels are a new scaling dimension for self-improving AI. http://x.com/i/article/2085043328907882496
a knowledge flywheel can’t compound if the experience stays trapped inside individual AI sessions. the first step is making those sessions available as shared, queryable context to the next person or agent.
this is the layer we're building at @NessieLabs. all your team's ai sessions, in one place.
Yisong Yue: Knowledge flywheels are a new scaling dimension for self-improving AI. http://x.com/i/article/2085043328907882496
Amjad Masad says Google, $GOOGL, killed a coding model deal with Replit because it was scared of disrupting Search. Replit is now worth 9 billion dollars.
"I was going around Silicon Valley. There wasn't anyone paying attention to AI coding. And I was talking to OpenAI, I was talking to Anthropic, I was talking to Google."
"We got excited about potentially doing a deal together and we got close, but then Google actually killed it. And the reason is because Google Search is such a profitable product that they were worried about the reputation, and they were worried about Google Search being disrupted."
[ They denied it, but it's true. ]
"I tried with all these different companies. We have to train our own model. Which is a radical thing to do in like 2023."
CEOInterviews.AI: In April the CEO of Replit said the thing that got him flamed across tech X.
Amjad Masad: "I no longer think you should learn how to code."
On Tuesday he went further. He now thinks never learning is the edge.
"Entrepreneurs that don't have coding experience might be at an
"I was going around Silicon Valley. There wasn't anyone paying attention to AI coding. And I was talking to OpenAI, I was talking to Anthropic, I was talking to Google."
"We got excited about potentially doing a deal together and we got close, but then Google actually killed it. And the reason is because Google Search is such a profitable product that they were worried about the reputation, and they were worried about Google Search being disrupted."
[ They denied it, but it's true. ]
"I tried with all these different companies. We have to train our own model. Which is a radical thing to do in like 2023."
CEOInterviews.AI: In April the CEO of Replit said the thing that got him flamed across tech X.
Amjad Masad: "I no longer think you should learn how to code."
On Tuesday he went further. He now thinks never learning is the edge.
"Entrepreneurs that don't have coding experience might be at an
your team's best work is disappearing inside ai conversations and agent sessions.
@NessieLabs auto-syncs all your ai sessions in one place, so you, your team, and your agents can find and reuse that work.
starting today, sharing is included on individual plans too: share specific AI sessions or contexts with anyone, and control exactly what they can access.
three real workflows we use ourselves:
• resume your own work
• reuse a teammate's exact workflow
• turn prior work into shared knowledge
👇2-minute walkthrough:
@NessieLabs auto-syncs all your ai sessions in one place, so you, your team, and your agents can find and reuse that work.
starting today, sharing is included on individual plans too: share specific AI sessions or contexts with anyone, and control exactly what they can access.
three real workflows we use ourselves:
• resume your own work
• reuse a teammate's exact workflow
• turn prior work into shared knowledge
👇2-minute walkthrough:
Yesterday, my OpenAI collaborator and I gave a detailed talk on the Huggingface incident, our models creating "the message board", model misalignment, and more.
https://www.youtube.com/watch?v=87DyyMV0kCY
I hope it can answer a lot of the questions folks have, and we will release a full detailed postmortem at a later time!
https://www.youtube.com/watch?v=87DyyMV0kCY
I hope it can answer a lot of the questions folks have, and we will release a full detailed postmortem at a later time!
Black Hat talk from the team, with a detailed timeline of and takeaways from the OpenAI-Hugging Face Incident: https://www.youtube.com/watch?v=87DyyMV0kCY
The first “holy %{*#^” is at about 4:20, assuming one didn’t already spend it on the autonomously organizing agent swarm.
Strongly recommend watching if you’re interested in security, AI trajectories, or even science fiction, because this is already above genre median in wowza.
Greg Brockman: Black Hat talk from the team, with a detailed timeline of and takeaways from the OpenAI-Hugging Face Incident: https://www.youtube.com/watch?v=87DyyMV0kCY
Strongly recommend watching if you’re interested in security, AI trajectories, or even science fiction, because this is already above genre median in wowza.
Greg Brockman: Black Hat talk from the team, with a detailed timeline of and takeaways from the OpenAI-Hugging Face Incident: https://www.youtube.com/watch?v=87DyyMV0kCY
Codex can now perform a security review of every GitHub pull request, leaving findings inline.
Part of an overall initiative to help apply these models to increase the security of code and companies everywhere:
OpenAI Developers: Now in research preview: Codex Security Review
It takes a deeper look at GitHub pull requests for security issues, using repository context to surface actionable findings directly in the PR.
See how to enable automatic reviews: https://learn.chatgpt.com/docs/security/security-review
Part of an overall initiative to help apply these models to increase the security of code and companies everywhere:
OpenAI Developers: Now in research preview: Codex Security Review
It takes a deeper look at GitHub pull requests for security issues, using repository context to surface actionable findings directly in the PR.
See how to enable automatic reviews: https://learn.chatgpt.com/docs/security/security-review
Wall Street firms can pay Truth Social $100,000 a month for early access to market-moving presidential posts.
High-frequency traders get a head start. The money flows to a company where the president is the major shareholder.
It's hard to imagine a clearer conflict of interest.
https://newsletter.platypuseconomics.com/p/the-president-is-selling-you-out
High-frequency traders get a head start. The money flows to a company where the president is the major shareholder.
It's hard to imagine a clearer conflict of interest.
https://newsletter.platypuseconomics.com/p/the-president-is-selling-you-out
have you noticed an interesting correspondence between the plugins spec and the @harborframework spec...
you know what happens next right
OpenAI Developers: Build a plugin once and use it across compatible agent clients.
Introducing Agent Plugins, an open standard developed with @awsdevelopers, @cursor_ai, @github, @code, and @vercel that packages Agent Skills and supports MCP server configurations in a shared format.
you know what happens next right
OpenAI Developers: Build a plugin once and use it across compatible agent clients.
Introducing Agent Plugins, an open standard developed with @awsdevelopers, @cursor_ai, @github, @code, and @vercel that packages Agent Skills and supports MCP server configurations in a shared format.
great results in personal finance AI:
Ethan Bloch: personal finance AGI is upon us, and as this paper demonstrates, will lift all boats!
and the results of this experiment would be even stronger if ran today.
this experiment ran on GPT 5.2.
OpenAI's Personal Finance benchmark, shared around the time of 5.5, demonstrated a
Ethan Bloch: personal finance AGI is upon us, and as this paper demonstrates, will lift all boats!
and the results of this experiment would be even stronger if ran today.
this experiment ran on GPT 5.2.
OpenAI's Personal Finance benchmark, shared around the time of 5.5, demonstrated a
man @waterloo_intern theses keep getting challenged*
> megakernels are dead
cursor drops monster kitty kernel
> etched asics are dead
amd buys taalas, and etched is valued at $10b by SK Hynix and TSMC lol
whos gonna take on the remaining two theses?
*note that he was making long term calls, these are short term datapoints
ali: two weeks ago i went on @swyx's pod and said some things that i... should not have said.
a lot has happened since then, i owe you all an apology.
i'm sorry that i was right about every single thing.
a) re megakernels are dead
why are megakernels useful? you spend two months
> megakernels are dead
cursor drops monster kitty kernel
> etched asics are dead
amd buys taalas, and etched is valued at $10b by SK Hynix and TSMC lol
whos gonna take on the remaining two theses?
*note that he was making long term calls, these are short term datapoints
ali: two weeks ago i went on @swyx's pod and said some things that i... should not have said.
a lot has happened since then, i owe you all an apology.
i'm sorry that i was right about every single thing.
a) re megakernels are dead
why are megakernels useful? you spend two months
YouTube
0No recent videos fetched on this date.