It's a good experiment you're doing, with that bankroll amount, and so far you haven't had any losses, on the contrary, you're making a profit. So, for now, the experiment can be considered positive. We'll see how AI performs in the recent games you bet on. I don't use AI and prefer to trust my own analysis, and I bet $5 on Brazil to win. It has decent odds, and I believe that, given Brazil's history of victories in their first World Cup games, their current good coach, and their good squad, they will win. I know Morocco shouldn't be underestimated, but Brazil has players and a coach with better skills than Morocco.
Combining AI analysis with our own analysis does seem more effective. I personally do this as well. However, I only use it to check head-to-head records, player fitness data, and similar information, as well as the history of matches between the teams about to play. So what the OP is doing is actually quite common these days. It’s just that updating here and sharing the results here is actually pretty interesting. Because sometimes we’re always pretty curious about what others come up with. Hahaha
Regarding the Brazil vs. Morocco match, in terms of squad depth, Brazil is clearly the favorite. The odds for betting on Brazil also look pretty good. Head-to-head, Brazil is clearly superior as well.

But personally, I’m curious and want to take a small risk in this match by backing Morocco. morocco usually plays a fast-paced style. I hope they’re still like that.