Chipotlemonger said:
@NFLPlayerProps said:
That's correct if you don't really understand what you're doing. If that's the case I would invest some time into getting past that point very soon.
Come on, you can't just make these last 2 posts and not provide any directions or points. Otherwise your argument stands on nothing.
Sorry, very busy day at work. My current setup uses Claude Max and Gemma inside an Obsidian vault. Once I realized I could feed it academic research about options markets and it very quickly knew more about it than me I got very aggressive about giving it more info. So far I've had it ingest 500+ books on investing/trading and a countless number of academic research papers.
So far I've backtested thousands of strategies, most die at that stage. The ones that don't get forward tested via paper trading on IBKR. Leading candidate at the moment is an ORB strategy on QQQ and SPY. Others include mean-reversion, SPY-GLD ratio rotation, global equities momentum, session flow for FX/futures, seasonal rotation of commodities, event drift etc. I also built a Polymarket copy trading bot that was incredible, but its not legal in the US and they're pretty good at detecting VPNs and then taking your money so I opted not to fund it.
I'm on the $200/month Claude plan and my goal is to hit the limit every week. Its challenging to give it that much work since I discovered the concept of model tiering but I usually get there, especially if I'm using Fable a lot. Especially when I have it create a workflow that fans out to multiple agents and just works autonomously for hours. Usually I'll create one of those before I leave to go to the office and when I get back its done exactly what I asked by the time I get home (but not always, once I created one that was so big it ran for 11 hours). And I always have the workflow invoke Fable to plan and then review/adjudicate results, that burns a ton of tokens.
I also use it as a financial advisor and accountant, it has ingested the entire US tax code and enough study materials and practice exams to pass the Series 7. And I'm also buidling a subscription-based website that will help high school athletes get recruited and get scholarships, about to start beta testing that with families I know from my kids playing club sports.
Its honestly incredible, life-changing is not an exaggeration. And I didn't even touch on how it has improved my work, I just had the best year of my entire career by a wide margin and even won an award that less than 0.5% of people in my company win annually.
Here's a graphical view of everything I've got going on in the vault currently. Each node is an idea or a strategy or a project or a PDF of a book or research paper. I know exactly what its doing and why and so yes I will trust it to trade actual dollars once ORB gets to 200 trades (assuming the success rate stays where its at so far).
