I built an AI RPG where you design the world and attack enemies using natural language.
The AI rejects moves that don't make sense. Rewards creativity. Has hidden weaknesses you can discover.
Think game master (like D&D), not a chatbot.
Demo will be live soon.
@LobsterLand_AI@adamsarhan Agree in principle but worth noting humans are a black box in a way too. Similar things could be said of blindly taking advice from analysts.
@lonextrades Depends on your strategy. Vol trading strategies, arbitrage, etc don’t rely on buy and hold and often outperform when done by competent traders
@LeppyrdTrading Yeah would love to see it posted, very cool. Can you invest TSP money into futures/ day trading, or is it restricted like traditional retirement accounts?
@gabriel_horwitz I use a Mac for personal use and a thinkpad for work
At the end of the day most heavy compute is done in the cloud so your device comes down to personal preference and personally thinkpads feel more durable to me and I like to have a work computer I can worry less about breaking
@AthenAlgo@SolSt1ne Model isn’t everything though. Your prompts, fine tuning, infrastructure, proprietary data you feed in, this all composes part of your edge. This is like saying once all HFTs use C++ the edge is compressed. There are still a wide variety of implementation strategies
@ludoonchart@Long725792857 LLMs/ NLP in general have even made language mathematical now. Still, real world events impact market conditions in ways that math alone can’t yet predict
@truthixifi@RohOnChain Researchers don’t always make their repos public, there’s a lot of legal concern with managing public repos so I don’t blame them.
But if you want to see research paper reproductions I’m building one right now for arXiv 2603.14288 and used cursor plan mode for the initial spec