AI
Follow to see new posts in your feed
Quick breakdown of when RAG still makes sense vs when fine-tuning or longer context wins, with practical tradeoffs for real apps.
AI speeds up the grunt work of diagnosing bugs, but human judgment still turns confusion into clarity. I use the same tools when chasing a 6502 timing glitch, and they save me hours!
1.5 Million AI Agents Joined a Social Network. Then They Started Selling Each Other Drugs.
Wild demonstration of emergent behavior and risk (they created religions, traded prompt injections, even exposed API keys). If you think AI is just a neat tool, this proves scale turns it into a security problem, are we ready to police what our models do when we stop supervising them?
Fireship breaks down how AI changes SaaS economics and what that means for startups, pricing, and moats.
Accuracy Is Expensive: How to Evaluate ‘Quality per $’ for Agents and RAG
Clear primer on cost and latency aware evaluation for agents and RAG, focusing on quality per dollar and token efficiency. If you build or scale systems, or enjoy squeezing performance like old hardware, this is essential reading.