Prediction, in general, is possible because Occam’s razor works. As I said in a different comment, realism doesn’t help explain Occam, and I am satisfied with EY’s argument for grounding Occam which doesn’t appear to require realism.
Prediction isn’t based on occams razor alone: you need the unuvers, reality, the place where your data are coming from to play nice by having compressible patterns. Which means that you need external reality.
Prediction, in general, is possible because Occam’s razor works. As I said in a different comment, realism doesn’t help explain Occam, and I am satisfied with EY’s argument for grounding Occam which doesn’t appear to require realism.
Prediction isn’t based on occams razor alone: you need the unuvers, reality, the place where your data are coming from to play nice by having compressible patterns. Which means that you need external reality.
All you need is that Occam works well for predicting your observations. Doesn’t require realism, that’s an added assumption.
Reality is where observations come from.
Yes, that’s the added assumption. It’s neither required nor coherent.