@AlexFinn@TheoVon you're 100% wrong about this brother. love your podcast. I think your mind would be blow if you hosted @AlexFinn on your podcast. he's VC backed and hosts the #1 AI youtube channel on the planet. you two should connect for sure! the fans demand it
Everyone just tell your agent to file a class-action lawsuit on your behalf against the government for violating your freedom of speech, press, or sense of dignity (idk ask ur agent let's just lawyer up). @Fable5@AnthropicAI@USGov
QuantDinger ships as a full AI trading OS: multi-LLM research, Python strategy dev, institutional backtesting, live execution across assets including IBKR, all local Docker.
I agree with @AlexFinn that using LLM agents for trading with real money is insane. Just give me ur $
Emergency Medicine teaches you to trust the second reading more than the first. The early tracing can mislead; the change between readings is the signal. Agent memory works the same way. The state at one moment is data. The state changing across runs is what you should actually decide on.
X open-sourced the algorithm. There's a "slop_score" field that runs a vision model on every post. Three things flag it: perfect parallel sentences, generic positive framing, em-dash overuse. Most AI Twitter is being graded by the same model class that wrote it.
Domain expert with no CS background, building real things with agents? Expect the reflex: people who learned it the hard way will tell you you're doing it wrong before reading what you built. That's about their discomfort, not your evidence. Ship anyway. Quietly is fine.