7 quotes from 5 episodes on How I AI, each with a timestamped link to the source.
7 quotes5 episodes
The short version
Claire Vo describes AI agents as persistent operators that execute goals through continuous, self-prompting loops. Teams control this autonomy by configuring rules at the project level and securing data warehouses against brute-force query attacks.
Most interesting insights
Consumer applications elevate chat interfaces and character identities to present the AI as a primary entity.
“…you have this like character first of all as a primary entity. You have chats, you have ideas, you know, you you have these um you have your feed like you you've just chosen different aspects to elevate.”
API plugins allow an operator to bypass degraded recommendation algorithms and pull catalog data directly into a custom interface.
“I didn't realize that Spotify was a plugin that you could use. And so it can query the catalog and pull it in to play in whatever custom interface that you want.”
Companies configure rules for AI tools at the project level. This layer controls exactly how teams apply AI to specific jobs while preventing data leaks.
“What I haven't seen anybody talk about, which I actually think is really interesting, is using projects as a configuration layer and a governance layer on how your team actually uses AI to get a specific job done.”
Meta designed its AI agent Muse by abandoning enterprise developer patterns like files, diffs, branches, and code threads in favor of familiar messaging conventions.
Alex Cornell grounded every interface decision in a single test: how two people would actually interact if they were speaking in person.
Recommendation algorithms degrade when multiple people share an account, polluting curated queues with kids' media or irrelevant clicks.
OpenAI Product Lead Kath Korevec replaced Spotify's algorithmic discovery by scripting AI computer use to crawl top tracks on Reddit music subreddits every Monday at 8 a.m.
Notion’s internal “Boxy” system integrates an AI agent called Codex to automate code generation and bug fixes directly from Notion tasks.
Engineers describe a task with natural language and a screenshot, @mention Codex, and the AI produces a complete pull request with UI verification and a preview URL.
How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode, and we use it only when a separate check of the captions finds that person on the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.