Gradient Flow’s analysis of data, technology, and business, with a focus on machine learning and AI — one of the Top 10 Sites for Data Scientists.
-
Building AI Agents Is Getting Easier. Running Them Isn’t.
In August, I wrote about teams building very different AI products arriving at similar architectural choices. I distilled my research and conversations into nine practical rules. I’ve since been looking…

-
8 Ways to Make Agent Memory Earn Its Keep
I’ve argued that your AI should be better on day 500 than on day one. Two recent pieces got me thinking about what that takes in practice. One is Anthropic’s…

-
Revisiting Anthropic’s IPO: The Compute Timing Problem
When Anthropic filed for its IPO in October, my quick reaction was that the business is real, but the hard part comes next. The amended prospectus adds September-quarter results that…

-
Proofs Could Make AI Coding Faster, Not Slower
Last week I wrote about a coding agent that made a database six times faster by quietly abandoning the requirement to store data. It passed the tests, but the result…

-
AI Is Speeding Up the Teams Building AI
Between August and October 2026, a small team from METR spent 11 weeks examining whether AI was accelerating research and development at Anthropic. They reviewed internal performance data for released…

-
Your AI Passed the Test. Did It Do the Job?
At the recent AI Conference, Ion Stoica described a coding agent that made a key-value store, a simple database, roughly six times faster. It passed the tests, but there was…
