When a camera pans away from a table and pans back a minute later, a pure pixel predictor video model usually forgets what was sitting on it...

Most people assume training on millions of hours of video will automatically teach models spatial reasoning. And video models can definite...

Most VLMs treat images as a flat grid of 2D pixels that get mapped into the same token space as text. That’s great at semantic pattern matc...

A frontier vision-language model can describe a picture of your living room but it can't tell you if your new couch will fit through the doo...

A frontier vision-language model can describe a picture of your living room but it can't tell you if your new couch will fit through the doo...

When a model trains on unverified synthetic data, hallucinations compound, edge cases disappear, and the system gradually degenerates into b...

Training new AI models on data generated by older AI models is becoming a trap. As the public internet fills with machine-generated text, la...

I think the next wave of AI winners will be defined by post-training mastery. Base models provide the commodity substrate, but the teams t...

Pre-training frontier models is quickly becoming a commodity and post-training is becoming the competitive differentiator. The moat is shi...

The most effective agent workflows constrain the model inside a deterministic software harness. You give the system concrete verificatio...

I think enterprise defensibility belongs to the teams that build verification layers, deterministic rails, and robust error-recovery infrast...

If each step in an autonomous agent’s workflow is 90% accurate, a six-step chain succeeds barely over half the time.

I think the AI ecosystem would benefit from fewer synthetic benchmarks and more public arenas that test autonomous systems end-to-end

We need to bring back large-scale engineering competitions that give away a ton of prize money. The modern autonomous vehicle industry trac...

The ultimate promise of RSI is turning the manual craft of building models into a scalable engineering discipline. So, I think the real si...

RSI usually sparks some anxiety about an AI rewriting its own code and ascending to uncontrolled superintelligence overnight. But RSI still...

The danger in AI-doomerism is that it distorts how we build and regulate technology. When policy conversations center on mythical rogue sup...

Large models don’t possess will, desires, or a drive to preserve themselves against human intervention. They’re mathematical functions tha...

I think AI-doomerism has its place, but that most AI-doomerism energy today is misplaced. The narrative is usually some apocalypse after bu...

A lot of the anxiety around AGI stems from projecting human psychometrics onto software. We talk about models getting smarter as if they h...