The problem with many predictive processing researchers is that when they encounter a visual illusion or strange sensory effect they can't immediately explain, they expect the explanation to be framed in terms of predictive processing with such high precision weights… that they actually hallucinate one. --- In other news: just landed in Sweden 🇸🇪

This is actually a really significant issue. Predictive processing becomes itself a high level energy sink that traps researchers for years/decades and captures all weirdness whether it's sensible or not. A lot of effects are much better explained with holography, cross frequency coupling dynamics, or even topology. Perhaps at different Marr's levels of analysis than what you might expect. But the prototypical predictive processing researchers will instead just hallucinate FREE ENERGY PRINCIPLE everytime.