By default my first question to any startup is "What's your growth rate?" That's the patient's pulse. Which is why we push startups to launch. Till then you have no pulse; till then you have no idea if you're doing well or badly.
I'm a manager at @OpenAI, but with GPT-5.5 I'm a more effective IC than I've ever been. I can now write CUDA kernels like a pro. I can rely on it to run my research experiments. And we know how to make it much more powerful from here.
This chart puts the datacenter demands into perspective very clearly. Amazon has done more capex in the last 3 years than its entire history.
Right now most AI adoption is on chat tools that are relatively token efficient. Comparatively, coding agents, use orders of magnitude more tokens, but are only used by a small population today.
These same kind of consumption patterns are about to come for the rest of knowledge work as well. The demand to process tokens for agents is probably 100s of times greater than we’re realizing right now.
Expect these charts to keep going vertical.