If it’s anything like the thousands of pilots and implementations in my company, only like 2 actually made it to production use.

If you listen to the ai-bad crowd, screw science.

You can just skip the experimentation and go directly to conclusions. Testing things is for idiots, the real enlightened among us already have the conclusion, they just need to gather the right observations to prove themselves right.

I prefer real science, thanks. You know, the kind that gets published in scientific journals and peer-reviewed, where they actually try to provide controls for what they are measuring, rather than going “trust me bro, I tried it and it’s a total game changer”.

You’d think, as a science enjoyer, you’d know better than to conflate the process of testing and forming conclusions based on observation and academia.

On the other hand, your idea of a scientific process is par for the course for a clankerfucker