Trying to predict price of gold with ML – problems with variables

Hey guys, as the title suggests, I have an idea for a model that could predict price of gold. I know this stuff is hard and likely not to generate alpha but that's not what I'm interested for now. I'm just a beginner who wants to get his hands dirty.

However, I need some help: In my model I'd like price time series for each metal. What variables should I use? I've looked it up and there are a bunch of tickers for each metal so I don't know which to use. Should I just use all of them and then do dimensionality reduction? Should I randomly chose 1 per metal since they're probably very correlated (some tickers are like gold & silver so I wouldn't know what to do)?

Any help is appreciated.