You have a point, the problem is that AI might be better in some instances but in other instances for example insurance, banking, medical, in more critical cases, having an opaque system with wrong and rigid rules is worst than a nightmare. But everyone will jump on the bandwagon, not only the less problematic cases (I thing we agree that not buying a book might be considered less important than not receiving medical care) but everyone. In somes instances if AI does a worst job the consequences on human life might be terrible
> ... the problem is that AI might be better in some instances ...
It's not good for anything. Every time you spot how useless it is for a specific task, the AI fans say you are using it wrong and for the wrong purpose, and please give it more data and time to get better. It's garbage and the VC money is running out. It'll become a forgotten acronym soon.
