david xu

[david xu @mateclawai]

@GeminiApp AI went from “tell me the problem” to “show me the screen.” The next step is obvious.

[david xu @mateclawai]

Every AI company seems to be building the same thing:

a second brain that lives on top of your operating system.

Memory + Screen + Actions = the new runtime.

[Google Gemini @GeminiApp]

Get tailored help for what's on your screen using the Gemini app for macOS. 💻

Simply press both Command ⌘ keys at the same time to seamlessly attach your active window to the chat, without needing to take manual screenshots or switch tabs.

[david xu @mateclawai]

We didn’t get AGI today.

We got something more dangerous:

an AI that never forgets.

[OpenAI @OpenAI]

We’ve been researching new ways for ChatGPT memory to carry context across conversations and keep it useful over time.

Today, that work is rolling out as a more capable memory system in ChatGPT. https://t.co/0MyFKCe2Mu

[david xu @mateclawai]

@OpenAI ChatGPT used to forget me after every chat.

Now it’s starting to remember my entire life.

At this rate, it’ll remind me about my unfinished side projects before I do.

[david xu @mateclawai]

@AnthropicAI @ch402 We hired Claude to help write code.

Claude’s long-term plan might be hiring Claude.

[david xu @mateclawai]

The biggest AI milestone may not be AGI.

It may be the moment AI starts building the next generation of AI.

Recursive self-improvement changes the game from “using AI” to “AI accelerating AI.”

[Anthropic @AnthropicAI]

Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor.

It’s happening faster than we thought, and the implications deserve greater attention. https://t.co/OVVPJO7VQx

[david xu @mateclawai]

@drfeifei World models: not just “make it look real,” but “don’t let the chair teleport through the table when I blink.”

[david xu @mateclawai]

Everyone is building AI coworkers. Hackers just started hiring first.

[Anthropic @AnthropicAI]

How well do the security community's techniques hold up against AI-enabled cyberattacks?

We examined 832 malicious accounts and mapped their activity onto a longstanding database of tactics and techniques used by threat actors.

Here's what we learned:https://t.co/fgOqJRh2rx

[david xu @mateclawai]

@AnthropicAI AI isn’t inventing new cyber attacks. It’s turning every existing attack into a scalable workflow. The real battle is no longer model vs model, but Agent Runtime vs Agent Runtime.

[david xu @mateclawai]

We thought Codex was coming for developers.
Turns out it came for product managers, analysts, researchers, operators… and probably my inbox too.

[OpenAI Newsroom @OpenAINewsroom]

Codex now has more than 5M weekly active users.

But the bigger story is what people are using it for: not just writing code, but getting more work done across research, analysis, content, and operations.

Our new report on how Codex is becoming a productivity tool for knowledge work: https://t.co/XZMPR9cEge

[david xu @mateclawai]

@OpenAINewsroom The most important number isn’t 5M weekly users.
It’s that Codex is no longer a coding tool.
OpenAI is quietly turning it into an operating system for knowledge work.
That’s where the next AI battle will happen.

[david xu @mateclawai]

@sama The AI race is no longer about chatbots. It’s about infrastructure, agents, workflows, and who builds the operating system for the AI era.

[david xu @mateclawai]

@OpenAIDevs @jxnlco So the new PM workflow is:

1. Ask AI what to build.
2. Let AI build it.
3. Present it to leadership.

Still somehow get blamed for the deadline.

[david xu @mateclawai]

Product managers aren’t being replaced.

They’re just getting promoted from “please write the PRD” to “please manage this tiny army of agents before they ship a SaaS company during lunch.”

Honestly, fair trade.

[OpenAI Developers @OpenAIDevs]

Role-specific plugins in Codex are built around the work teams actually do.

Plugins for Data Analytics, Creative Production, and Product Design give Codex the tools and context to create reports, creative directions, and prototypes.

Built and used by OpenAI teams.

[david xu @mateclawai]

@AnthropicAI AI has officially entered its Space Race era.

The next competition isn’t chatbot vs chatbot.

It’s nation vs nation, powered by agents.

[david xu @mateclawai]

So AI went from:

“Can it write my essay?”

to

“Can it help secure national infrastructure?”

Pretty wild upgrade in just a few years.

[Anthropic @AnthropicAI]

This Executive Order is an important step in strengthening America’s leadership in AI.

We look forward to collaborating with the White House to support its implementation.

https://t.co/ZwDimPrp3t

[david xu @mateclawai]

OpenAI is turning software into a conversation.

Anthropic is turning workflows into agents.

Codex Sites means anyone can build and publish an app with a URL.

We may be entering the biggest app creation boom since the iPhone.

[OpenAI @OpenAI]

Building apps has never been easier.

With Sites, Codex can turn your work, ideas, and plans into an interactive website or app your team can explore, use, and share with a URL.

Rolling out to Business and Enterprise plans, before expanding more broadly.

[david xu @mateclawai]

The most important AI application may not be coding.

It might be helping us live longer.

[Derek Thompson @DKThomp]

This has quietly been a miracle month in medicine.

In the last 5 weeks we’ve got news on:

- retatrutide, the triple agonist GLP-1 from Lilly, basically melting fat and body-wide inflammation at record levels
- RevMed’s new pancreatic cancer drug showing unprecedented abilities to extend life
- small trial of a one-and-done PCSK9 gene editing therapy for slashing LDL cholesterol
- Mayo’s AI-assisted radiology showing vastly improved cancer detection
- this new therapy for metastatic solid tumors

This stuff is at varying levels of evidence. Retatrutide is ~100% on its way, other stuff needs more clinical trial data. But put it together and we’re maybe on the verge of majorly reducing the mortality of heart disease and cancer, the two leading causes of death in America.

[david xu @mateclawai]

Chatbots answer questions.

Project Glasswing fixes the Internet.

That’s a very different game.

[Anthropic @AnthropicAI]

We’re expanding Project Glasswing. We’ve extended access to Claude Mythos Preview to approximately 150 additional organizations, based in more than fifteen countries.

Read more about this expansion and our future plans for Project Glasswing: https://t.co/QrtHSBdRbh

[david xu @mateclawai]

The AI race is shifting from chatbots to problem solvers.

Claude is positioning itself as the AI for thinkers.

The next step is Runtime + Agents that can actually execute work.

[Claude @claudeai]

From The Problem Solvers, our series featuring founders taking on hard problems with Claude: https://t.co/lXWwc2eHPe