Machine learning in trading: theory, models, practice and algo-trading - page 3733

 
Ivan Butko #:
It doesn't matter.

You're using AI ideas because you have nothing of your own.

What ideas, verbose verbose?

You were told a year ago that AI is incapable of generating new ideas, even that hasn't reached you yet?

A lot of things that were thesesised a long time ago haven't reached you yet. So respect the hierarchy.
 
Maxim Dmitrievsky #:
AI is incapable of generating new ideas...

Yerp. I proved this by fiddling with presently existing AI.

If you work at JP Morgan, you might have access to agentic AI which allegedly can create new ideas... or maybe you were replaced by it before you got access. 😏

 
Ryan L Johnson #:

Yerp. I proved it by messing around with the existing AI.

If you work at JP Morgan, you may have access to an agency AI that can supposedly create new ideas.... Or you may have been replaced by it before you got access. 😏

What exactly has been proven and to whom? )
 
Maxim Dmitrievsky #:
What exactly has been proven and to whom? )

I uploaded 4.6 GB of M1 EURUSD OHLC historic price data to Gemini and challenged it by prompting it to give me the most profitable trading strategy.

It returned absolute nonsense, e.g., a 10 period SMA, a 20 period SMA, and advised me to use my own discretion to identify swing highs and lows, SL's, and TP's. This is the epitome of generic garbage.

Jamie Dimon claims to have agentic AI at JP Morgan, and I don't work there.