GeekWire GeekWire AI summit takeaways: Token budgets, watermelon metrics, and the $5k weekend coder
10 snips
Mar 28, 2026 A post-summit recap about the economics of AI, from token budgets shaping hiring to startups relying on subsidized credits. Stories of runaway spending, costly shutdowns, and why some metrics look healthy but hide losses. Conversations about agents, vibe coding and changing developer roles. Practical prep tips using Claude, Gemini and inbox automation.
AI Snips
Chapters
Transcript
Episode notes
Use A Longterm LLM Project For Event Prep
- Use a long-term LLM project as a single source of truth to prep and organize event tasks and speaker info.
- Todd Bishop fed invites, surveys, and conversations into a Claude project so he avoided repeating context and tailored questions quickly.
Try Multiple LLMs And Choose The Best Output
- Try multiple LLMs and pick the one that produces the clearest, usable outputs for your task.
- John Cook tested Gemini then switched to ChatGPT for tighter, cleaner dinner questions and used only a few prompts live.
Token Budgets Are Now A Hiring Expectation
- Token budgets are becoming a standard part of hiring negotiations as engineers expect access to AI compute.
- Charles Lamanna and panelists compared tokens to essentials like email and mobile phones for modern developers.