Every decade, another iteration of replacing thinking in humans with nothing comes around. In the 90s, visual languages were all the rage, don't need to know anything, just drop a box here, select another there and link them by a "make money" arrow. South Park brilliantly captured this in the "Gnomes" episode. Take "steal underpants" box, link it to "profit!" box, just name the arrow "make money" and that's all it takes! Billions and trillions of investor money follow because, why not? You see dissociated comments on the BBSes as if coming from a schizophrenic world: same developers that are supposed to be replaced report both "fantastic, 100x increase in productivity" and "this isn't doing anything for me, dumber than a 100x increase fanboy who doesn't yet know that 100x of nothing is still nothing". We've stolen underpants worth more than a trillion dollars now. Where's the profit?
Similarly Musk's brilliance is not in anything that he conjures up but that he gets people to believe what he conjures up is anything of great significance.
yep, AI does not help you. Because the most important is reasoning on backtesting, why something should work. Without answering this to yourself any backtesting and number crunching and huge optimizations in any space does not make sense, and is more just a waste.
Wrong... "AI", "machine learning", call it what you will... is the genie and she's out of the bottle. Just keep your expectations realistic. P.S. "4GL" languages aren't a panacea either but they still contribute to productivity.
For example backtesting on fundamental data or what has a fundamental meaning makes much more sense to do. But you need to know the constants, because any slight overoptimization can lead to (much) worse results out-of-sample. So you need to have a feeling how much you can optimize and what inputs of course. And that comes from understanding what you do. Without any reasoning it does not make sense. Have not seen any AI so far, which can do this well.
This is a perfect example of unrealistic expectations. No one is saying "AI" is going to become sentient and make humans obsolete. These sophisticated statistical techniques will just help us become more productive and gain deeper insights into the world. You still need to be careful that you don't shoot yourself in the foot. That's how all technology works. Keep your eyes on the stars, but your feet on the ground.
I didn't even read this but anybody's saying AI isn't the future are probably the same people that we're saying Bitcoin is going nowhere LOL. I use ChatGPT daily. As far as using it with stocks, then it's a tool like any other tool. I don't think anybody's saying it's going to be making you millions in stock picks...but great for evaluation. Like if I ever want to interpret the finances of a company I just paste the report in. You get this pushback every evolution like snowboarding, CDs, Bitcoin, VR, computers. cell phones, electric cars Etc...sometimes the technology gets ahead of itself, like with 3D tv's but if they make them without the need to wear glasses (autostereoscopy) then I wouldn't be surprised if there was a comeback.
for those who remember programming before the internet took off, it was slow, tedious, reading technical/protocol specs. Rewriting what everyone else already did; no source control. Advanced Programming in the Unix Environment by Stevens and C Programming Language by Kernighan and Ritchie were good but with AI I could create all *BSDs and Linux in no time. And everything is easier now.