AI Practicum · CEO cohort · Sep 8, 2026
September 8, 2026
Sep 8, 2026 CEO session — frontier model choice, agents vs chat, Flash economics at scale.
Each deck was authored by the model it showcases.
Agenda
- Block 1 — Frontier model comparison (Astra · Fable · Flash)
- Block 2 — Agents, not chats (GrokBot) → handoff to Mario Montag
- Block 3 — Gemini 3.8 Flash with thinking — economics at scale
Deck 1 · Frontier model comparison
When to Use Which: Astra · Fable · Flash
Built with GPT-6 Astra (xhigh) via Codex
Fair comparison of GPT-6 Astra, Claude Fable 5.1, and Gemini 3.8 Flash for CEO decision-making.
Open deck →
Deck 2 · GrokBot / agent demo
Agents, Not Chats: Fable 5.1 Max in the Enterprise
Built with Claude Fable 5.1 Max via Claude Code
What an agent can do that a chat can't — GrokBot, Aegis secrets gatekeeper, CRM + Notion, Mario Montag (Asana + Slack + GrokBot).
Open deck →
Deck 3 of 3 · Flash economics
Gemini 3.8 Flash with Thinking
Built with Gemini 3.8 Flash + thinking
Gemini 3.8 Flash + thinking as the economic workhorse for org-wide scale.
Open deck →