Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

After the AI boom: what might we be left with?

blog.robbowley.net

Post reply on HN
Print view View original

191–200 of 489 posts

Re: After the AI boom: what might we be left with?

#191

Post by trenchpilgrim »

pclmulqdq wrote: post #182
trenchpilgrim wrote: post #124

Earlier quoted context omitted.

Those parts take decades to wear out, excepting a manufacturing defect.

That is not correct when wires are small and you run things hot. The physics of <10 nm transistor channels is very different than it is for 100 nm+ transistors.

Your assertions are at odds with modern computers - including Nvidia datacenter GPUs - still working fine after many, many years. If not for 1. improved power efficiency on new models and 2. Nvidia'a warranty coverage expiring, datacenters could continue running those GPUs for a long time.

Re: After the AI boom: what might we be left with?

#192

Post by mike_hearn »

CuriouslyC wrote: post #84

This is where we're headed: https://sibylline.dev/articles/2025-10-12-ai-is-too-big-to-f...

Comparing to China is tricky because Chinese investment is almost by default bubbley in the sense of misallocating capital. That's what heavily state-directed economies tend to do and China's still a communist country. Especially they tend to over-invest in industrial capacity. China's massively overbuilt electricity grid and HSR network would be easily identified as a bubble if it were in the west, but when it's state officials making the investment decisions we tend to not think of it as a bubble (partly because there's often never a moment at which sanity reasserts itself).

I read an article today in which western business leaders went to China and were wowed by "dark factories" where everything is 100% automated. Lots of photos of factories full of humanoid robots too. Mentioned only further down the article: that happens because the Chinese government has started massively distorting the economy in favor of automation projects. It's widely known that one of the hardest parts of planning a factory is figuring out what to automate and what to use human labour for. Over-automating can be expensive as you lose agility and especially if you have access to cheap labour the costs and opportunity costs of automation can end up not worth it. It's a tricky balance that requires a lot of expertise and experience. But obviously if the government just flat out reimburses you 1/5th of your spending on industrial robots, suddenly it can make sense to automate stuff that maybe in reality should not have been automated.

BTW I'm not sure the Kuppy figures are correct. There's a lot of hidden assumptions about lifespan of the equipment and how valuable inferencing on smaller/older models will be over time that are difficult to know today.

Re: After the AI boom: what might we be left with?

#193

Post by pixl97 »

trollbridge wrote: post #128
archerx wrote: post #16

I believe the next step will be robotics and getting A.I. to interact with the physical world at human fidelity. Maybe we can finally have a Rosie from the Jetsons.

Robots still haven’t come close to replicating human and animal touch as a sense, and LLMs don’t do anything to help with that.

I mean, even the human brain is broken up in different parts and the parts that do touch are insanely old compared to higher thinking. The LLM parts tell the robot parts the plate should go to the kitchen.

Re: After the AI boom: what might we be left with?

#194

Post by grafmax »

CuriouslyC wrote: post #84

This is where we're headed: https://sibylline.dev/articles/2025-10-12-ai-is-too-big-to-f...

Im not convinced the geopolitical rivalry between US and China is a given. To a large degree it’s been manufactured - Friday’s antics a case in point.

The US indeed seems destined to fall behind due to decades of economic mismanagement under neoliberalism while China’s public investment has proved to be the wise choice. Yet this fact wounds the pride of many in the US, particularly its leaders, so it now lashes out in a way that hastens its decline.

The AI supremacy bet proposed is nuts. Prior to every societal transition the seeds of that transition were already present. We can see that already with AI: social media echo chambers, polarization, invading one’s own cities, oligarchy, mass surveillance.

So I think the author’s other proposed scenario is right - mass serfdom. The solution to that isn’t magical thinking but building mass solidarity. If you look at history and our present circumstances, our best bet to restore sanity to our society is mass strikes.

I think we are going to get there one way or another. Unfortunately things are probably going to have to get a lot more painful before enough people wake up to what we need to do.

Re: After the AI boom: what might we be left with?

#195

Post by pixl97 »

cyberax wrote: post #123
flyinglizard wrote: post #10

I admit to only being in this industry for three decades now, and only designing and implementing the thermal/power control algo of an AI chip family for three years in that time, but it's the first time I hear of chips "wearing under high intensity use".

Chips absolutely do wear out. Dopants electromigrate and higher temperatures make it faster. Discrete components like capacitors also tend to fail over time. Is it going to be that significant though? No idea.

Honestly it depends on a whole lot. If they are running 'very' hot, yea they burn out faster. If they have lots of cooling and heating cycles, yea, they wear out faster.

But with that said machines that run at a pretty constant thermal load within range of their capacitors can run a very long time.

Re: After the AI boom: what might we be left with?

#196

Post by pclmulqdq »

trenchpilgrim wrote: post #191
pclmulqdq wrote: post #182

Earlier quoted context omitted.

That is not correct when wires are small and you run things hot. The physics of <10 nm transistor channels is very different than it is for 100 nm+ transistors.

Your assertions are at odds with modern computers - including Nvidia datacenter GPUs - still working fine after many, many years. If not for 1. improved power efficiency on new models and 2. Nvidia'a warranty coverage expiring, datacenters could continue running those GPUs for a long time.

Which GPUs have been running for decades that you're referring to? The A100s that are 4 years old? MTBFs for GPUs are about 5-10 years, and that's not about fans. AWS and the other clouds have a 5-8 year depreciation calendar for computers. That is not "decades."

You can keep a server running for 10-15 years, but usually you do that only when the server is in a good environment and has had a light load.

Re: After the AI boom: what might we be left with?

#197

Post by mikert89 »

dcminter wrote: post #167
mikert89 wrote: post #161

Earlier quoted context omitted.

pretty much all of the classical ideas of what an ai could do, can be done with our existing capabilities. and yet, people continue to live as if the world has not changed

"AI" has been doing that since the 1950s though. The problem is that each time we define something and say "only an intelligent machine can X" we find out that X is woefully inadequate as an example of real intelligence. Like hilariously so. e.g. "play chess" - seemed perfectly reasonable at the time, but clearly 1980s budget chess computers are not "intelligent" in any very useful way regardless of how Sci Fi they w…

yeah im in agreement, ai will eventually do everything a human being can

Re: After the AI boom: what might we be left with?

#199

Post by mike_hearn »

rjh29 wrote: post #46
scellus wrote: post #17

He writes as if only datacenters and network equipment remain after the AI bubble bursts. Like there won't be any AI models anymore, nothing left after the big training runs and trillion-dollar R&D, and no inference served.

Who's going to pay to run those models? They are currently running at a huge loss.

There's a gazillion use cases for these things in business that aren't even beginning to be tapped yet. Demand for tokens should be practically unlimited for many years to come. Some of those ideas won't be financially viable but a lot will.

Consider how much software is out there that can now be translated into every (human) language continuously, opening up new customers and markets that were previously being ignored due to the logistical complexity and cost of hiring human translation teams. Inferencing that stuff is a no brainer but there's a lot of workflow and integration needed first which takes time.

Re: After the AI boom: what might we be left with?

#200

Post by squidbeak »

dcminter wrote: post #29
mikert89 wrote: post #12

Earlier quoted context omitted.

who cares about a bubble? we are on the cusp of intelligent machines. The implications will last for hundreds of years, maybe impact the trajectory of humanity

> we are on the cusp of intelligent machines That's an extremely speculative view that has been fashionable at several points in the last 50 years.

How often in the last 50 years have those machines done what these machines do?
Post reply on HN