[querez]

As an AI researcher: don't forget that we're at the peak of the hype-trend: a lot of what LLMs can do looks extremely impressive, but most of it falls apart on more detailed inspection. Unless you ask about mundane things, they will tend to get a lot of tiny (and not so tiny) details wrong. At the very best, these systems will allow us to automate some boiler plate things, but they're not good enough to do complicated stuff without lots of supervision. And what's even more important: I think a lot of people look at the current impressive steps and think "oh wow, if it continues at this pace we'll have the Singularity by autumn". But the thing is: we're not able to keep this pace of progress. What you're seeing now is as good as it gets (in terms of big breakthroughs, there will still be lots of small and medium ones). The next few months (and years?) will see a ton of incremental improvements and many, many, many people trying to apply these new technologies. But I personally (as a decently successful AI researcher who's been part of these developments for over 10 years) don't see a way forward to keep making many of the big strides we've been making. As an analogy: we're having a Bitcoin moment: imagine the first blockchain was just released for the first time: now, there'll be lots of people trying to understand the tech, come up with their own variants, make some (fundamental?) improvements. But the actual fundamental tech/idea is out now, and it's not really going to change much.

TL;DR: I think your job is safe.

[crop_rotation]

There is no comparison with bitcoin here. Bitcoin is still struggling to find a single legitimate use case, while GPT4 is already proving so useful to so many people (including me).

> What you're seeing now is as good as it gets (in terms of big breakthroughs, there will still be lots of small and medium ones).

Unless you have a strong source for this, I find it hard to believe. Also, GPT3 didn't have many big breakthroughs over GPT2, other than massive parameter size.

[querez]

My comparison was merely in terms of how the technology evolves, not it's applications. Sure, Ether and zero-trust coins seem to add a lot, but from a 10.000 feet it's pretty much more of the same. Also, agreed that GPT1->4 didn't involve any breakthroughs. In my eyes, there's not much diff between them.

> Unless you have a strong source for this, I find it hard to believe

Like I said, this is just my opinion. I do have a world-expert level understanding of the technology, but at the same time I'm a strong believer that even experts are bad at predicting the future, so make of this what you will. Also, my impression of what constitutes a breakthrough and what doesn't might vary a lot from yours.