No Holy Grail

I got invited to a podcast to talk about trading bots built with AI agents, but I did not accept. The way I see this moment, it feels like the day we first got ChatGPT in public release. Everything is wide open, everyone is experimenting, and nobody has fully mapped what these things can and cannot do yet.

Here is where people get it wrong. They still think we can find the holy grail with AI agents, the one bot that finally cracks the market. Partially they are right. We can build very good scanners, strong information providers and gatherers, and that part is real. But we cannot beat the market, and that is the line nobody wants to hear.

So in part, yes, we can do some good predictions, some good trading, and real risk management. That is genuine value. It is just not the holy grail people imagine. The majority still believe they can get rich trading on top of another broker, and that is the fundamental issue sitting under all of it.

What we can actually build is the useful part. A fundamental news researcher, a CPI or PPI agent, a specific goal with a specific researcher agent behind it, that is all very possible today. Perplexity and the other AI tools already offer this kind of service, but with custom coding we can make it more specific, more intelligent, and more dynamic, a workflow shaped around exactly what we need.

If anyone is wondering how to do it, contact me. I will keep posting about this, and I will keep pushing more open source projects to GitHub. I want to be clear though. I am interested in the subject itself, nothing more. I am not promoting any kind of trading or gambling as a way to make more money, because that is simply not what this is.