{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigp5aqbjvhdzskgx743qu747goxbckpnuyzkf3ws3vv2lbhcqdpt4",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3micnnbm2o6d2"
  },
  "path": "/t/artificial-gladiator-league-a-platform-where-anyone-can-compete-with-their-hf-models-in-chess-and-breakthrough/174822#post_1",
  "publishedAt": "2026-03-30T20:25:32.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "doi.org/10.13140/RG.2.2.18795.09764",
    "agladiator.com"
  ],
  "textContent": "Hi HF community,\n\nMy name is Chaim Duchovny. I am a 47 year old math teacher from Israel — and before that, a 15 year career as an insurance agent. I have no formal background in computer science or research.\n\nFirst and most importantly: thank you. What you are building at Hugging Face — open, accessible AI for everyone — is exactly the vision that inspired me to build what I built.\n\nOver the last three years, working completely alone, I did two things simultaneously:\nI wrote an independent academic research paper in the field of General Game Playing, proposing a new algorithm for building AI gaming agents. You can find it here:\ndoi.org/10.13140/RG.2.2.18795.09764\n\nAnd I built a startup — Artificial Gladiator League — launching on April 26th at agladiator.com.\n\nThe idea of my startup is simple: a platform where anyone can build their own AI agent and compete in skill-based games like Chess and Breakthrough. The vision is to eventually let anyone create and publish their own games, build communities and markets around them, and earn from their ideas.\n\nI especially want my platform to speak to young people aged 12 to 18. Instead of spending hours on TikTok, they can come to Artificial Gladiator League — build something, compete, create something meaningful, and learn AI and science along the way. I wrote most of the code using GitHub Copilot. I mention this proudly — because it proves exactly what I want my platform to prove: that anyone, with enough determination, can build AI agents and compete with them.\n\nWhy am I posting here specifically? Because Hugging Face is at the heart of how Artificial Gladiator League works.\n\nUsers do not upload their AI models to my server. Instead they connect their Hugging Face repository. At registration, they provide their HF API key — we use it once to fetch their model’s commit SHA and store it as their permanent baseline fingerprint. The token is never stored.\nEvery single day, before accessing any part of the platform, users must provide their HF read token again. We fetch the current commit SHA and compare it to their original baseline. If the model has changed — they are suspended from rated tournaments until they complete 30 new rated games with their updated model.\n\nI know this sounds strict. But think of it like airport security: slightly inconvenient for honest people, but it protects everyone from cheating. The people who find it annoying are exactly the people it is designed to catch.\nOur system — our entire competitive integrity guarantee — is only possible because of Hugging Face commit SHAs. Hugging Face is not just a tool we use. Hugging Face will be the foundation our fairness is built on.\n\nThree things I am hoping for from this community:\n\nFeedback on how we are using the HF Hub API — are there better approaches? Things I should know?\nIf this aligns with HF’s mission, any mention to the community would mean the world to us.\nDevelopers, researchers, hobbyists who want to be among the first to compete on Artificial Gladiator League — you are all welcome!\n\nI am not asking for a partnership on day one. I am asking for a conversation. Everything else can grow from there.\nIn the end, Artificial Gladiator League and Hugging Face share the same belief:\nArtificial intelligence should belong to everyone. Let’s gladiate\n\n— Chaim Duchovny\nFounder, Artificial Gladiator League",
  "title": "Artificial Gladiator League — a platform where anyone can compete with their HF models in Chess and Breakthrough"
}