The issue with these AI engines is that they will confidently give you the wrong answer. It depends on the topic of course. Some topics it does a very good job with. But with wood identification, Ive found that its absurdly incorrect. It'll rattle off a bunch of stats about a species (janka hardness, density, pore details, color, etc) and provide references for the data, but if you go follow the links provided you don't find the information AI provided anywhere. It will seemingly just make up a janka hardness for species that have never even been evaluated for Janka hardness. That's my big issue with it - it looks good on the surface, but its all nonsense fabricated numbers. My experience with is trying to identify based on photos has been equally bad. Its just like all the people in the Wood ID groups on facebook who think any dark wood is walnut, wenge, or palm.