I think they know the bubble’s going to burst, and they’re just trying to make sure they’re the survivor that’s still standing at the end of it, the next Google and not the next Lycos or Alta Vista. But in that effort I’m guessing at least one or two major players are going to destroy themselves when they might’ve otherwise survived, or at least turn themselves into the next Yahoo! or AOL, still existing but just as a shell of what they were, and leaving people surprised when they realize they’re still around.

This technology would have been lauded if they had presented it as what it was - natural language recognition for machines - and stopped there. But they decided to push this entire narrative of it being able to analyze and make decisions, replacing all knowledge workers with it, which it fundamentally is incapable of doing, because that's not how vector transformers work. There's going to be an equally diminutive reaction to those exaggerations. The fact that this race to an impossible goal is actively harming hundreds of other industries and ultimately the consumers across the board is only adding to that resentment.

actually capable of listen to sentiments that they do nothing useful at all?

They do nothing useful for skilled workers working within their domain of expertise.

For the unskilled, the magic copy machine is amazing.

Now, who am I to judge which category the average CEO falls into. lol.

I haven't seen case studies proving it's better or more efficient. You save on labour in one area but now need AI compliance, policy admin, hallucination checkers, prompters .etc. the total amount of jobs saved is close to nothing. Companies are hiring people back again because they see this.

Also there's corporate slop where you summarise a report and outsource the cognition of reading and summarising it to someone else. It looks like you're super productive but really you're using AI to obfuscate the fact that you aren't doing your job and are relying on the people you interact with to do your work (reading and understanding the slop summary. Verifying it's authentic and wasn't hallucinated)

The only use case I've seen that works is unit test generation and limited integration testing. That way you have a test to fail an AI with if it changes things and breaks shit. You get it to write its own tests so you confine it's behaviour. I personally do this when programming (I also dont let the AI make the code. It writes the tests, I make them pass)

You're right, the reasonable position is to buy up every single piece of technology on earth so we can see what might do faster.

My experiments with some of the code assist models is they can be an amazing productivity boost, but only if you are not greedy.

If I ask the bot "Please make me this data structure with this list of parameters" its probably 99% perfect. Moving up, the sweet spot is closer to "Make me this module of code that matches this ABI/API". Alternatively "Hey can you make this entire feature?" I will get a 8000 line of code blob that may or may not work. Even worse if I get really greedy I will get a 8000 file project filled with 8000 lines of code that is most definitely going to start the robot uprising.

Giving the junior and midlevel code monkeys access to AI coding tools is an amazing way to generate a mountain of slop the senior and most expensive code monkey gets trapped trying to decipher while the junior monkeys make even more shit that needs to be dealt with.

Boil it down, if we make centaurs of humans leading AI it will definitely work better than a reverse centaur that will confidently tell a gig driver to drive off a cliff.