What's next after the AI bubble bursts? | Hacker News Reader

What's next after the AI bubble bursts?

seangoedecke.com

21 pointsgfysfm26 comments

Personally I see there being a whiplash effect in the US workforce where after the crash (stock market, egregious wealth gap, cost of living in food/medical basics) the nation will return to a “local minded” situation where jobs will become lower efficiency, higher headcount, akin to the CCC’s “make work” type programs. The efficiency of AI in replacing humans is evident in many menial tasks…but the consumer driven economy of the US is showing real signs of pending disaster. Unemployment isn’t low, the numbers are troubling when seen in real terms.

To put it another way, cutting people who stopped looking as “not unemployed” is as ridiculous as excluding the price of milk or eggs from real inflation / COL metrics. The fudging of numbers simply cannot last forever.

So while AI will absolutely be a boon in the short term, in the long run, and how it all come apart could be very disturbing circa 1793 France, the human element will rebound. Will we see self-checkout at grocery stores taken away for humans once again? Will there be work from home operators instead of AI call routing? It’s hard to imagine just what a collapse of the high point of human productivity (and general repression of lower economic classes through insurance and wage stagnation tracing back decades) but it is a cyclical historical framework.

In short, whatever it is, it’s going to be pretty ugly and potentially break apart the US Federal structure because the rifts are beyond closing. AI is simply the gasoline on the fire. True fear and loathing my friends…

Spy telescopes that travel faster than light and watch you jerk it 20 years ago.

I’m convinced the people who think there is an AI bubble are simply not using it.

It is increasing my ability and productivity beyond belief and I see incredible potential.

The author mentions they don’t think it is a bubble necessarily but that we will have lots of GPUs, however I seriously doubt anything other than more AI will be the result of lots of GPUs.

I tried using it every once in a while and it still feels like a massive bubble.

It can get basic front-end react apps done but nothing beyond that.

But since front-end programming wasn't ever really programming, I don't mind that and I sometimes use it for that.

For everything else, it just makes a massive mess.

It's good to ask it questions about pretty common topics and get a cleaner reply than searching stuff yourself on Google and going through spam articles.

But any sort of insight or any sort of new learning, it just simply doesn't have it.

Furthermore, all of the hyping of research assistants for $20,000 and all of that is just absolute bullshit.

And no, there is no AGI coming.

Hey, AI has every chance of attaining parity with human intelligence, if it succeeds in making humans stupid enough.

> It is increasing my ability and productivity beyond belief and I see incredible potential.

How?

IMO this only applies to people who are already an expert in the domains they are using AI for. Only they can guide the AI to produce something useful. In all other cases it's the blind leading the blind.

It may be non-obvious, but bubble doesn't mean that there isn't potential or that there isn't value - which is basically what the point of this post is.

It's that the amount of investment outpaces the current value of return.

Think about the internet. Everyone rushed in and billions was spent. But we never stopped getting utility from the internet. In fact utility likely increased during the bust, which is why there was a 2nd boom in 2003, which may have continued had the GFC not occurred.

Or tulips, even after the tulip mania of the 1630s tulips are still worth something. Perhaps that's not a fair comparison, though, since they're very pretty flowers and AI is only AI.

I know (or think) you're being facetious, but the tulip market never reclaimed it's former heights. Not yet at least...after NFTs, maybe that's next!

3.0lip

I do too, but I still believe there is a bubble. Reason is that I can do most of my work with self-hosted models so there’s little need to pay for a subscription. Sure, I’m a minority for now because I’m tech savvy, but with that in mind it’s hard to justify the valuations of most AI companies. We don’t know if they’re keep accumulating users and most LLMs seem to have hit a plateau of performance.

And local models are getting more accurate and efficient. And they really don't need a dedicated GPU for fast response, just really fast ram. New APUs such as the AMD 395 are going to be game changers for local users with 256 GiB/s bandwidth.

The new Mac Studio is great for this job too. With the Max chip you get 128GBs of unified RAM for $3.5k. Memory bandwidth is at 546GB/s which is good enough for most tasks.

> I’m convinced the people who think there is an AI bubble are simply not using it.

I have used it and it IS a bubble.

Why? There are lots of free AI models that one can run on a Macbook, and even hosted AI models are racing themselves to $0.

There will be a crash and this so-called "AGI" will takeaway more jobs than it creates. [0]

It's no secret that in 2030, it will become a reality as we already have accelerated layoffs and it isn't going to stop.

It appears in this hype cycle, it is the year 1999.

[0] https://www.weforum.org/publications/the-future-of-jobs-repo...

Id have gone with "Singularity to sigmoid, whats next for AI"

There's certainly an AI valuation bubble. It's starting to look like AI isn't going to be expensive, and that everybody can do it. So the startups with a P/E ratio over 10-20 are going to get a haircut. That includes OpenAI and Tesla. Even if Tesla can get self-driving to work, that's still just part of the low-margin auto industry.

There's Trump's "Stargate" consortium.[1] Don't want to go there.

We will see more routine use of AI. But it's still not good enough to use where you can't offload the externality costs of errors onto someone else. Customer service being the classic example of where you can.

There are probably two decades of wrenching change ahead as the world adjusts to a weaker US, a stronger Russia, and a much more dominant China. That may slow down civilian tech adoption.

The next big growth area is probably low-cost killbots in high volume.

[1] https://www.cnbc.com/2025/01/22/kelly-evans-what-on-earth-is...

Don't forget heated mannequins... with killbots come decoys!

I guess, my bet would be VR then (third time's the charm!), of the matrix/neuromancer/etc "get a brain implant, plug in and forget the outside world" kind, with a focus on people living in VR for sustained periods of time.

Technology-wise, we seem to be making steps of getting there, a large part of the population seem to be "living" in virtual worlds anyway (through their phones) - and my hunch is that if the world will become progressively nasty due to wars, pandemics, climate change and massive social inequality with no hope of change, there will be growing demand to just to be able to escape all of it and try a better life "somewhere else".

The companion technology might be neuromancer-style "surrogate bodies", where instead of "plugging into" a virtual world, you remote-control a robot in the real world. Want to stay in the real world, but retain the ability to choose whatever body you like? Want to have an insurance against a sudden case of rocket attacks? Want to keep it cool even with 50°C/80% humidity outside? No problem, just transfer your mind to a robot body of your choosing!

>a stronger Russia

Russia is in its death throes. The invasion of Ukraine was inevitable since it needs to expand outward to remain demographically solvent, which it cannot in the long run even with Ukraine. [0]

[0]https://upload.wikimedia.org/wikipedia/commons/thumb/e/ec/Ru...

The entire developed world is in population decline. It's not just Russia. South Korea, Japan, and Hong Kong are far worse off.

Asia is ahead in automation investments, Russia is stuck with a persistent natural resources curse. It is hard to see them breaking out of that curse anytime soon, especially with the leadership they have.

Demographic decline isn’t going to be much of a problem if AI and robotics keep advancing at their current rates, as long as those investments are made (like as China is doing now).

South Korea, Japan and Hong Kong didn’t recently throw what little proportion of their 20 something men who didn’t flee the country into a meat grinder operated by middle aged alcoholics.

After AI, the next boom will be robotics.

Robotics does indeed sound like a field that synergizes well with AI advancements.

It's an emerging market.

If you include drones in robotics, you're certain it will be popular.

Reply on news.ycombinator.com