Nate Lee

Your OAuth grant list is a list of companies whose worst day becomes your worst day. A Context.ai employee went looking for Roblox auto-farm mods because you gotta get dem Robux. Unfortunately for them, it contained an infostealer that ran, stole browser-stored credentials, session tokens, and OAuth grants for

Meta employees seeking to maxxx their productivity drove all-time AI usage highs, burning 60 trillion tokens in 30 days. I'm really not sure what to do with that. This is a company filled with smart people that somehow decided the best way to measure AI productivity was to

The average engineer is about to become the most expensive hire on your team. Great engineers with agents get enormous leverage. Their judgment about what to build, what to throw away, and what the agent got subtly wrong is where the value lives. An average engineer paired with an agent

We got lucky and weren't impacted by the trivvy/axios supply chain attacks but what's a team supposed to do when patching fast isn't safe and the impending vulnpocalypse means that soon, you won't be able to wait? The advice to wait

We had a spectacular demo failure yesterday. Someone I met at RSA was testing the product. It exploded on an onboarding step. We know the root cause and put controls in place so it or something similar doesn't happen again. But that's not why I'

The machine that ships features faster matters more than the features themselves. Some teams treat their development system as overhead, something to maintain between "real" work. That's backwards. Features are the exhaust of the machine. The machine is your linting rules, your agent constraints, your debugging

In three years, 50-person companies will run more rigorous third-party risk programs than most Fortune 500s do today. Large enterprises built compliance programs around what humans could manage which usually meant spreadsheets and annual vendor questionnaire follow-ups. The constraints shaped the process. AI removes those constraints for everyone simultaneously but

A few months ago I looked at a prompt I'd been iterating on for weeks. It had fifteen lines of "never," "always," and "unless." Each one was the fossil of a failure I'd patched without diagnosing. That's the

Removing humans from the loop isn't a philosophical choice anymore. It's a survival requirement. That's the argument Michael Coates made when he testified before Congress last December alongside Google and Anthropic about the rising threat from AI and it's the thread that