> We asked ChatGPT to create a portfolio of stocks from high-quality businesses with criteria taken from the leading 10 funds it is competing against. These include things like low levels of debt, sustained growth in the past and assets that generate an advantage of competitors.

This sounds like a bad idea given the knowledge cutoff date, unless you also feed it the relevant data. But would that fit in the prompt?

It's easy to feed current relevant data to ChatGPT.

They asked basically "pick me a bunch of stonks based on 10 criteria". They ended up with 38 stocks. If it's to do any selection at all we need to assume the potential possibilties were at least 100. So now you want text about 10 criteria x 100 stocks into one prompt.

You could split it into multiple prompts but that's not what they claimed to have done.

Zero information about methodology or prompting means there is nothing useful here, nor is there any way to verify veracity.

The hedge fund and investment management industry ready lost the game to machines post 2008. Just look at the successes of quant firms and HFT firms relative to fundamental hedge funds. Or just look at the success of quantitative teams vs qualitative analyst teams.

Can you give us some links so we can look too?

Sorry, I don't have any links to share, since if I get n links proving my point, there are m links disproving my point, both of which are correct. But I'm an industry insider, having worked in PE before, and even my old firm acquired a small data-science and Ml-oriented firm, inspite of being very traditional and white shoe. Also look at the rise of quant divisions in banks and the shut down and consolidation of traditional Sales and Trading desks over the past 15 years.

> if I get n links proving my point, there are m links disproving my point, both of which are correct

Yes, that does sound very convincing.

It's 4.93% absolute in 8 weeks, not annualized.

which bank account if you don't me asking