With a view to answering my own question about the validity of using AI for problem solving, I decided to give it a test with my Premier League football predictions and also to compare how the two main candidates, ChatGPT and Claude, work out over a few weeks.
So I am comparing them both over four weeks of predictions and testing them against my own gut instinct to see which one is the best!
My own prediction skills are not that great as I languish in the third division of a three division prediction league so any help by these masters of AI would be appreciated with a view to moving me both up the table and then up the divisions.
I began with setting them both the fixtures for this week (weekend of 22-25th August) and they responded with some questions of their own which showed me that the information they were accessing was a few months old. They were asking about sides that were relegated from the division in season 24-25 and about new sides that had been promoted.
I gave them the answers they wanted, and with which they seemed happy, and they then went off and generated the ten results for the games.
I kept a record of their predictions but used my own for my entry this week.
So far, their results, like mine, are not that exciting but as there is still one game to play, there is time for all of us to improve.
I will keep you informed over the next three weeks of predictions to see if they are better than my gut and, of course,as they learn and understand, if their predictions become better.
I’d love to hear your comments below about this post or anything AI related as we are seeing more and more of ut ion our lives these days!
All the best
DaveT
