News for hackers. Stuff that matters.

A comment on I spent a year and $5,700 to see if ChatGPT can beat the market (S&P 500).

Re:I spent a year and $5,700 to see if ChatGPT can beat the ... by ano-ther on Monday May 27, 2024 @03:35PM (#40493856)
Very interesting that this works. And great to see experiments like this which open their methodology.

Looking at the status page, https://www.gptinvestor.co/the-gpt-investor/

- GPT version has been better than benchmark 8 times, while the benchmark won 11 times.

- GPT return was negative in 5 experiments and 0 times for benchmark.

- GPT invests in a quite small selection of 1 or 3 stocks (mostly tech).

- The GPT outperformance comes from one 100% bet on NVIDIA that is still ongoing (+140%).

I like my blood pressure a bit more steady to start just yet with this method.

[ Parent | on HN ]
Re:I spent a year and $5,700 to see if ChatGPT can beat the ... by ishtiaqrahman on Monday May 27, 2024 @03:55PM (#40494050)
haha the thing is that I got better at prompting as I ran more and more experiments.

For example, in the beginning I was asking GPT to recommend me 3 stocks at a time, which was not a good idea. I revised the strategy to ask for only 1 stock recommendation at a time.

Also, GPT-4 was MUCH better at it than GPT-3.5 which makes sense and is also very exciting.

[ Parent | on HN ]
Re:I spent a year and $5,700 to see if ChatGPT can beat the ... by Rodeoclash on Monday May 27, 2024 @06:22PM (#40495316)
Why am I not surprised that the AI is going all in on the company that produces the hardware that it runs on.
[ Parent | on HN ]
Re:I spent a year and $5,700 to see if ChatGPT can beat the ... by ishtiaqrahman on Monday May 27, 2024 @09:02PM (#40496360)
I also noticed that. It recommended NVIDIA 5 times.
[ Parent | on HN ]