No Priors

Erik Allebest on chess thriving after machines beat it

Erik Allebest· Co-founder and CEO at Chess.com
·~46 min·English·No Priors
AI CompanyBusiness StrategyReasoning
TL;DR

Chess.com's founder bootstrapped a roughly $200M business from a domain investors dismissed as uninvestable, and argues that superhuman machines make people value human skill more, not less.

01Core Mental Model

Humans Want to Do Human Stuff

Machines have crushed humans at chess for 30 years, yet the game is more popular than ever — because people still want to do and watch human things, not outsource them.

fundamentally, humans want to do human stuff. That's, you know, humans competing against other humans. It's humans making things. It's humans solving problems.

— Erik Allebest, No Priors
Key Insight
The lesson isn't really about chess. Every time a machine masters a human domain, the confident prediction is obsolescence — and the actual result is a larger audience for the humans still doing it live.
02The AI Arc

Boring, Then So Much More Exciting

Engine perfection briefly turned elite chess into a boring grind, but neural nets like Leela Chess Zero played aggressive, alien moves that pushed the game forward and made it exciting again.

computers at first made chess a little bit more boring, but then made it so much more exciting. And we're seeing more exciting chess than we ever have.

— Erik Allebest, No Priors
Key Insight
A superhuman tool's value isn't the answer it hands you — it's the search space it opens. Stockfish's flawless play closed the game down; a self-taught neural net's weird ideas reopened it.
03Origin Story

They Told Him It Was Uninvestable

Most venture capitalists called chess an uninvestable niche, so Allebest bought the domain for $56,000 out of a bankruptcy auction and did the exact opposite of the standard startup playbook.

most of them said this is uninvestable. You should get a real job. Don't waste your time with this.

— Erik Allebest, No Priors
Key Insight
The 'uninvestable niche' was the moat. No capital meant no clock — and no clock let a passion product compound for twenty years while faster-funded rivals had to justify a return.
04How Mastery Works

Multiply Any Day by a Thousand

There is no shortcut to expertise, only reps: take any small daily habit, multiply it by a thousand days, and that compounding is what separates masters from everyone else.

you just take any given day in your life and multiply that by a thousand, that's kind of what your life looks like. So, if you do five chess puzzles a day, multiply that by a thousand, you've done 5,000 chess puzzles.

— Erik Allebest, No Priors
Key Insight
This is why information access alone never equalizes skill. The bottleneck was never knowing what to do — it's the reps. AI can compress the learning curve, but it can't do the compounding for you.
05The Cheating Problem

You Can't Put a Webcam in Every Home

Engine cheating can't be fully prevented, so chess.com treats detection as a permanent, well-funded arms race — logging every move, flagging anomalies with ML models, then banning and educating offenders.

cheating it can't be prevented because you can't, you know, you can't put a webcam in every single person's home. Um but we do do a really good job at catching it and and banning people who do that

— Erik Allebest, No Priors
Key Insight
Chess has fought superhuman impostors for decades, so its playbook is a preview for everyone else: not prevention but detection, run as a core, permanent cost of keeping the game trustworthy.
06Why Chess Endures

The Beatles of Board Games

When AI can generate 10,000 new games overnight, a fixed, luck-free, shared ruleset like chess becomes more valuable, not less — the timeless classic everyone can rally around.

when there's infinite music, sometimes people go back to like the Beatles and Led Zeppelin because it's like the classics that everyone can kind of rally around and have a shared experience.

— Erik Allebest, No Priors
Key Insight
In a world where novelty is suddenly free, scarcity moves to the timeless. A fixed, luck-free ruleset everyone shares gets more valuable precisely because AI can mint infinite alternatives.
07The Bigger Picture

Not a Technology Problem — a Cultural One

Allebest believes superhuman AI is coming, and thinks the real danger isn't the technology but the culture around it — wealth and power concentrating in too few hands.

I think we're already seeing the economic impact of just big tech and that just too much money is flowing into the hands of too few and we're not doing enough as as humans to share what we're creating.

— Erik Allebest, No Priors
Key Insight
Allebest quietly reframes the AI-risk debate: if the tool is neutral and the danger is concentrated wealth and power, then 'AI safety' is downstream of a distribution problem that technology alone can't solve.