🟒 COHORT 01 β€” OPENS SOONTHE LINE: 12 WEEKSSHARP vs SQUAREWEEKLY ACCOUNTABILITY Β· +EVSEATS AT THE TABLE: LIMITEDBUILD HARNESSES Β· SHIP AGENTS Β· STACK YOUR EDGE 🟒 COHORT 01 β€” OPENS SOONTHE LINE: 12 WEEKSSHARP vs SQUAREWEEKLY ACCOUNTABILITY Β· +EVSEATS AT THE TABLE: LIMITEDBUILD HARNESSES Β· SHIP AGENTS Β· STACK YOUR EDGE
πŸ“ˆ COHORT 01

The accountability community for AI engineers

Stop betting
against yourself.

AI is rewriting what an engineer is worth. The spread between the sharps and the squares is widening every week. AI Bets is a 3-month cohort where you place one bet β€” on yourself β€” and back it with weekly accountability until it pays.

12 weeks 1 session / week πŸ”’ 1 bet: you
BET SLIPCOHORT 01
WagerThe next 12 weeks
PayoutAn unfair edge
To enterYour email

No card. No spam. Unsubscribe anytime.

12
Week line
1Γ—
Weekly table
∞
Reps logged
+EV
Every session

A note from the table

The market just repriced you.
Are you a sharp or a square?

Here's the uncomfortable line nobody's posting: the gap between engineers who command AI and engineers who merely use it is now the single biggest spread in tech. One group ships harnesses and agents that do the work while they sleep. The other copies snippets into a chat box and hopes.

Squares bet on tools. Sharps bet on systems β€” and on showing up every single week.

You already feel it. The tooling moves faster than you can read the changelog. You half-built an agent once, it kind of worked, and then the week swallowed you. The problem was never the model. The problem is that nobody held the line with you.

The Square

  • Bookmarks tutorials, ships nothing
  • Re-types the same code by hand
  • "I'll set up my harness… someday"
  • Bets alone, folds quietly
VS

The Sharp

  • Runs agents that write, test & PR
  • Has a harness for the boring loops
  • Ships a rep every week
  • Bets with a room that calls them up

What AI Bets actually is

Not a course you abandon at lesson three. AI Bets is a 3-month accountability cohort β€” a standing weekly table for engineers who are done watching from the rail. Every week you put a rep on the board, the room holds you to it, and you walk out sharper than you walked in.

In 12 weeks you either have an edge β€” or an excuse. Pick one and bet it.

Who's invited to the table

Engineers who want AI to do more of the work β€” whether you're ready to ship autonomous agents, or you still type most of your code by hand and want to level up your whole setup. Both seats are open. We'll figure out your exact line on the next page.

Who's calling the table

Rich Steinmetz Β· your host

From the felt to the terminal

Before I ever shipped software for a living, I played poker professionally (2011–2014). Poker beats one lesson into you that no bootcamp can: you don't control the cards β€” only the quality of your bets. Think in bets. Respect variance. Protect your bankroll. Press a real edge the moment you find one.

I've been building software since 2016 β€” I dug ML and terminals back before AI was cool. Today I'm an engineer at an AI-first company, where I rotate Claude Code, Codex and Copilot and build my own harnesses and agents to do the work for me. Different table, same game I learned at the felt.

The engineers pulling ahead aren't the ones dealt the best cards. They're the ones making +EV bets β€” every single week β€” and actually sitting down at the table to do it.

AI Bets is the room I wish I'd had: poker discipline pointed straight at AI engineering. We don't gamble β€” we find the edge, size the bet, and grind the reps together for 12 weeks. I'll be at the table with you.

LAST CALLCOHORT 01

β€” The AI Bets table