335
Productivity & Workflow354
Automation & Workflow227
Software Development244
Marketing & Growth193
AI Infrastructure & MLOps173
Writing & Content Creation201
Photography & Imaging153
Design & Creative169
Data & Analytics138
Customer Support134
Education & Learning131
Sales & Outreach121
Voice & Speech132
Operations & Admin87
Garry Tan told a podcast he spends $100,000 to $200,000 a year on AI “tokens” and uses AI agents for work and personal tasks.
In short: Y Combinator CEO Garry Tan says he spends about $100,000 to $200,000 a year on AI “tokens” to run AI agents for work and daily life.
Garry Tan, the CEO of startup accelerator Y Combinator, said he spends “maybe $100,000 to $200,000 a year on tokens” for AI agents. He shared this on an episode of Walter Isaacson’s podcast, The Builders, which was set to publish the following Tuesday.
Tokens are the small units AI systems use to count and charge for text and other inputs and outputs (like paying per word, but more granular). Tan said he uses AI agents for many tasks, including coding and preparing for interviews by researching and predicting questions.
Tan also described using an agent more broadly, like a constant helper that knows his preferences (he compared it to a “Jiminy Cricket on your shoulder”). He said his agent has read his email, Dropbox files, Apple Notes, and even therapy notes going back 20 years. He has also shared a set of Claude Code skills called “gstack” on GitHub, meant to show others what agents can do.
Tan said Y Combinator is using an internal AI agent as well. He said it is named “Stray Light,” it has a personality, and it has access to a “company brain,” meaning a large pool of internal context like past founder applications and emails.
Tan said personal AI agents could become common, but only if costs drop. He also raised a concern about relying on free agents run by big tech companies, arguing people may want a personal AI they control and can configure themselves.
Source: TechCrunch AI