That's fine. Good luck with your endeavor but I still think you should contact some friends and family. If that's not possible then a helpline might also be beneficial: https://sczaction.org/helpline/ (toll-free number: 800-493-2094).

We can also continue this thread if you want but you'll have to make more coherent points about what exactly you're trying to achieve with Solomonoff induction. Most of the research on AIXI(tl) is pretty much complete. Whatever could be recovered from computable distributions and agents that learn properties of that distribution to maximize some kind of objective has already been mostly figured out by Hutter. The theory is complete so unless you have some other ideas I don't think there is anything worthwhile left to research with Solomonoff induction and computational agents based on that formalism.