E22: Forgetful Geniuses and Billion-Dollar Costs: What You Must Know About AI Agents

Key Insights

🔹 Break tasks into 15-min chunks

🔹 Use full-context windows (200k tokens!)

🔹 Verify after every step

🔹 Equip agents with read/write tools

🔹 Use KV caching to avoid cost bombs

A must for every CEO and AI leader in 2025.

For access to the full transcript, join our community!

E22: Vergessliche Genies und Milliardenkosten: Was du über AI-Agenten wissen musst

Wichtige Erkenntnisse

🔹 Chunk it or Junk it – Tasks < 15 Min

🔹 Volle Kontextfenster statt RAG

🔹 Schrittweise Verifizierung = Erfolg

🔹 Tools = echte Arbeitsfähigkeit

🔹 KV-Cache = 90% Kosteneinsparung

CEOs, CTOs, Teamleads – das ist eure Roadmap.

Listen On: