Early AI adopters spent all of 2023 and 2024 trying stuff and making mistakes before finally figuring out what LLMs are good for. Now that these tools have worked their way down to the normies, I’m seeing new people make the same mistakes all over again. AI is good for a lot of stuff, but it’s no substitute for original thinking and personal communication.
Over the last few weeks, I’ve become the product leader of a small organization, directing a team of top-notch, highly-independent engineers…. I’m commanding a fleet of Claude Code agents.
2025 was the year of agentic coding for me. Perhaps the first of many. I tried out Claude Code in June and then subscribed to Cursor soon after. Since then, I’ve been using both frequently. I’ve got some thoughts.
This last week I visited Portland with some friends. I had some tremendous coffee, good food, and interesting long runs (using some new shoes!). But the main thing on my mind has been AI coding agents…
Two years of using LLMs regularly has challenged a strong belief of mine: that fundamentally, thinking is primarily just language that happens in our heads. I’m aware that this wasn’t a prevailing view in psychology, but I still want to talk about what changed my mind, since it didn’t require reading academic literature – just engaging with practical tools and reflecting on the experience.
Sometimes when you’re searching for a digital record, you don’t remember the precise wording of the original entry…. Semantic search attempts to compare the meanings of words and phrases instead of the characters.
Last week, Nvidia’s stock exploded…. The AI arms race has heated up and is even getting the attention of normal people through stock prices (among other news), so I’m sure this won’t be the last time I explain GPUs to someone without a computer science degree. I’m using this post as a chance to organize my Chips 101 talking points.