Any "tokens" metric that includes tokens from multiple models is meaningless. Tokens cannot be compared across models. This has always been true but is more and more true as models diversify and as reasoning becomes a larger and larger portion of all tokens.
Adam D'Angelo
Verified account This account is verified. Learn more@adamdangelo
CEO of Quora, working on Poe: @poe_platform
1.5K Following 110K Followers 1.4K Posts
Latest posts 5
Toyota lean manufacturing has an idea called "genchi genbutsu", essentially meaning managers should "go see the real thing at the real place" in a factory instead of sitting in an office hearing secondhand descriptions. AI coding suddenly makes this practical for software.
Unclear if a durable trend, but CEOs and CTOs are back to coding with a fury, thanks to coding agents.View quoted post on X
I have public company CEOs sliding into my DMs (and “InMail”) telling me about falling in love with shipping software again thanks to Claude Code and Vercel.
“Dream accounts” that we always wanted to work with, where in the past the C-suite would hardly understand the infrastructure until much later in the game.
Coding agents are the ultimate PLG-fication of the enterprise. Bad, legacy software can’t hide anymore. The stack that works is self-evident to the entire organization, from intern to CEO.
I want to be able to tell my Waymo to take 280 instead of 101
This applies again today, e.g. people are saying the same exact thing now but code can still get *another* 100x bigger while every other vertical grows slower than code.
1-2 yrs ago the story was always "AI-for-code works pretty well so far but soon AI-for-every-other-vertical will catch up"View quoted post on X
Instead, code just got another 100x bigger...