So I skipped the risqué photos and went to chatGPT 4.0 and asked it to build me a stock watch list, with relevant news, performance fundamentals and technical analysis with buy hold sell advice based on my investment thesis. I spent much of the weekend deep diving into my investment thesis, locating stocks, understanding the various companies business and performance and on and on.
I discovered several mistakes, not shocking, but the funny part was I asked it to prepare reports, organize the information, and even automate a weekly update and summary for a Monday morning 15 min read.
It said it was working on it, and when I would ask for an update the response was I’m working on it, it’s taking time to build this complex tool and that I’m awesome, super smart and building something spectacular that some companies hire teams to do. It would tell me things like it’s almost done just checking for consistency blah blah blah. Then it would say it needs 6 more hours, 24 more hours, 48 hours. The bitch ghosted me.
so now I know it can be sloppy with facts, it can flat out lie and use flattery to cover it up, it will refuse to do something without telling me it is refusing to do it, what else…? Kinda human like now that you think about it.
for now I use it to gather information give easy recommendations that I then use as a base case and go vet on my own with additional study. Basically it can do what a first year grad employee might be able to do in bite sized chunks. The rest of the research and judgment is up to me.
the good news though is that I am still using my analytical skills, maybe even more now in this corner of my world than before.