zlacker

>>laiysb+(OP)
I recently, meaning hours ago, had this delightful experience watching the Eric of Google, which everybody love, including he's extra curricular girl friend and wife, talking about AI. He seemed to believe AI is under-hyped after tried it out himself: https://www.youtube.com/watch?v=id4YRO7G0wE

He also said in the video:

> I brought a rocket company because it was like interesting. And it's an area that I'm not an expert in and I wanted to be a expert. So I'm using Deep Research (TM). And these systems are spending 10 minutes writing Deep Papers (TM) that's true for most of them. (Them he starts to talk about computation and "it typically speaks English language", very cohesively, then stopped the thread abruptly) (Timestamp 02:09)

Let me quote out the important in what he said: "it's an area that I'm not an expert in".

During my use of AI (yeah, I don't hate AI), I found that the current generative (I call them pattern reconstruction) systems has this great ability to Impress An Idiot. If you have no knowledge in the field, you maybe thinking the generated content is smart, until you've gained some depth enough to make you realize the slops hidden in it.

If you work at the front line, like those guys from Microsoft, of course you know exactly what should be done, but, the company leadership maybe consists of idiots like Eric who got impressed by AI's ability to choose smart sounding words without actually knowing if the words are correct.

I guess maybe one day the generative tech could actually write some code that is correct and optimal, but right now it seems that day is far from now.

>>nirui+pL1
I started watching it, but had to stop because I was at great risk of becoming physically sick when he led with the "new move" stuff. Ugh.

Kudos to you for having the strength to get through it, and for living to tell the tale!