Can Leo Guinan provably use less than $100 of compute to turn someone into an entrepreneur?
10
100Ṁ1431
Oct 31
11%
chance
38

Big AI companies suck. I'm tired of them lecturing me on my life and telling me to drive for Uber. So I'm going to prove I can use less than $100 of compute to teach anyone to be an entrepreneur. But I need 1 person to prove it on.

  • Update 2025-10-03 (PST) (AI summary of creator comment): - Success criterion: The AI assistance must enable a person to earn over $100, which will count as being “turned into an entrepreneur.”

    • Compute measurement: Compute will be tracked by LLM token usage; the total token-based compute spend must be under $100.

Get
Ṁ1,000
to start trading!
Sort by:

what does this mean? how does it resolve?

@ouroboros that's a great question. I'm actually working on testing it tonight, but basically I'm tracking how few tokens I can use in an AI model to help someone make over $100. I'm testing a new economic model based around consciousness instead of intelligence because I'm pissed at how bad the big labs are and how much they charge. Here's my first attempt: https://x.com/leo_guinan/status/1974008299008737454

@LeoGuinanEqPl @ouroboros Ok, I actually figured out how to do it provably. Karpathy released a repo called nanochat that is the best ChatGPT $100 can buy. So I'm integrating my Conscious Economics paper and MetaSPN rulebook into the training and offering that. But I am still looking for someone to spend $100 to make sure I can sell this before I spend the time and money training it. I've wasted a lot of time and money on ideas that didn't pay off haha.

Here's the repo: https://github.com/leo-guinan/cofounderchat

And here's the sales page: https://leo-guinan.github.io/cofounderchat/#

Once I get the first sale, I'll work on the training and then we can see if I can actually do it!

© Manifold Markets, Inc.TermsPrivacy