I’ve seen an argument for doing this along the lines of ‘Allowing for infinite, means that if someone comes up with a faster (that is, finite) way of calculating something then it may still have already been reasoned about before.’
Yeah, that is probably a good reason to do it. The first time I stumbled upon this was probably a semitechnical introductory dialogue to solomonoff induction.
Yeah, that is probably a good reason to do it. The first time I stumbled upon this was probably a semitechnical introductory dialogue to solomonoff induction.