Comment by ProfHewitt

For something a little more rigorous see the following:

"Robust Inference for Universal Intelligent Systems"

https://papers.ssrn.com/abstract=3603021