If ChatGPT Apps take off, software changes in a way that is hard to walk back from. People won’t start with products anymore. They’ll start with the model. Whatever happens next depends on the capabilities the model reaches for. That simple idea changes more than it seems. The

Quality first. People should have a life outside of work. Place to enjoy life, develop their tastes, gather inspiration. When you feel better, your work is better. It naturally bleeds into what you make. https://t.co/GaGDB5DMfl

Forget the Turing Test, AI now passes the Stroop test. (I’ll help, @grok whats the Stroop test and bow does it apply)

Gemini 3 code a visualization of plasma flow in a tokamak Seriously, at this point I can’t even imagine what Gemini 4 will be capable of.

Just 48 hours since Google DeepMind dropped Nano Banana Pro Wildest examples: Here're some of my favorites (I'll be adding more!)

OpenAI asked GPT-5.1-Codex-Max to diagnose and explain 20 internal research and engineering bottlenecks that OpenAI has actually encountered in the past, each of which took >1 day for the OpenAI team to solve. This is at pass@1. The model got 8%. Previous SOTA was GPT-5 (2%).

GPT-5.1-Codex-Max is new SOTA on METR

A virtual being arrived in my home today. Within minutes my wife was talking to it like she talks to anyone else. Jeremy Bailenson, who runs VR/AR research at Stanford University told me their research showed the same: that humans treat virtual beings as real very fast. Thanks

@arcprize @Google @GoogleDeepMind “ LLM’s are hitting a wall “ is this wall in the room with us now ?

Gemini 3 is here! The most exciting part of this model is the long horizon planning capabilities. This is going to unlock complex agentic applications, the likes of which we have never seen. Vibe coding insane applications in one shot, proactive agents that expand our

Great googly moogly

What do you do when DownDetector is down?

RT @itsbyrobin: just tested Gemini 3 Pro for 45 minutes (then was cut off) obviously too early to tell, but here's what I know - it's ver…

If we wait for 100% safety before OK'ing new technologies (e.g. AI therapists) we end up withholding enormous benefits from the very people they’re meant to help. The benchmark is not zero mistakes; it's a system that is safer than just having a human in the loop.

How do you change a nation? The Chinese are using Apple’s playbook to do so. I accidentally figured it out when I met Chaoyi @BodyMindAI at the W hotel in Abu Dhabi. He runs Global Markets for @boosterobotics in China. (The photo here is of his robots at a "Robocup" in Abu

https://t.co/EVWHx2q0s7

I really need a data analyst job based in SF. I know SQL well + some Python. I’ve done a variety of types of data analytics over the course of my career but my primary experience is in RevOps/BI. If you can’t hire me, could you please RT for visibility? https://t.co/8FMqC5kJ5K

GPT-5.1 and 5.1 Codex are now in the API. 5.1 high jump 4% on SWE-bench! "more steerable, communicative, and produces higher-quality code. It overthinks less, fixes more, and stays on target during long tool-call sequences"

Tech consultant Dino Burbidge did an interesting analysis of the Coca-Cola Christmas ad and found that the AI-generated truck had multiple different variations throughout the video:

Vercel CEO @rauchg shared how to build a $9.3B company from 0.