-
After the bubble pops how much would our lives be impacted?
-
Would AI vanish or still be there?
-
How exactly do you think the bubble will pop? Will AI companies simply run out of money? Or will it be because of the environmental effects?
-
When do you think the “pop” will take place?
-
After the bubble pops, in future there will be companies/people who will try the AI thing again? What will that be like?
Ai is here to stay. There will be winners
Many companies will take huge losses but the real losses will be communities with huge dead buildings and unrewarded tax incentives. Power companies with wasted investments trying to feed datacenters that no longer exist and the rate payers that have to pay for it
I think there will continue to be a free tier: Apple is a great example of using small local ai to do useful things for their customers. However the latest models will be have to be much more expensive to cover the business case. Scale matters, so there’s a good chance the winners won’t just dominate in a country but globally
Anybody thinking AI goes away is just wishfully thinking, and/or ignorant.
deleted by creator
It will if it is too expensive to generate a positive return on investment. That’s not to say that there aren’t (very) specific use cases where running an LLM makes sense, but the general use case of attempting to replace human workers with AI will likely be recognized for the pipe dream that it is. Companies are already realizing this with the AI bills coming due now, and this attitude will compound once they start having major failures and no one is still around who can fix the problems.
Are you saying that you think there is a possibility that AI technology, especially LLMs, would disappear from usage? Even as search tools? Even used locally?
Like I said above, I think there will be some very specific cases where running an LLM makes sense. Think along the lines of querying lakes of scientific data. This won’t be something of value to the average Joe or the average company. My experience with using LLMs for general search has been very hit or miss. I always end up using a general search engine, even for internal data sources.
Maybe that’s just you but millions of people are interfacing with LLMs every day as a replacement to a traditional search engine. I don’t see that changing. Either way, sounds like you are agreeing that AI will not go away entirely. That would be a silly thing to say.
It’ll also go dormant as LLMs are only one AI tech. Other AI tech will be developing while supporting infrastructure grows until we have another boom. World models for example and JEPA. That’s why Meta is pushing AI glasses. They need training data.
Those millions of people using AI for general search are currently getting it for free, which is going to end when the bubble bursts and the companies providing it go under, if not sooner when said companies try to make a profit. This is exactly what’s happening to corporate AI customers now. They are realizing there isn’t a good ROI now that OpenAI and other AI providers are being forced to charge proper costs for queries and tokens, rather than the teaser or “adoption” rates they charged before.
Interesting theory. I’d be surprised if Google started charging for their AI results. Too much free training data and they can afford to keep models running for the masses. They’re not solely an AI company. At the worst, these AI companies get swallowed up by bigger companies that can keep running even if AI is not profitable, which it is in some cases.
The industry will end up with one or two winners and a lot of losers. Very hard to predict these things (smart money in 1997 was that Apple was about to file bankruptcy). One or two companies will have success and the rest will evaporate or be absorbed. Then the AI product itself will be priced up (and dumbed down) significantly.
AI will not go away, but will be far more limited and specialized. The day of free general access to powerful chatbots will be up.
GenAI is here to stay in one form or another, with long term impacts.
I wager the bubble will pop when OpenAI finally admits they have no path to making good on their purchasing commitments. They aren’t the whole bubble and Altman has made what should be obviously the worst financial moves, however the broader market will be more bearish on anything they vaguely think to be OpenAI like.
I suspect it’ll be with a year from now, based on what I’ve read.
The companies will never stop doing the GenAI things, but they may be less utterly obnoxious about it. I suspect Google and Anthropic have the highest chances of enduring a pop, anthropic for being well regarded in the field and Google because they get to make most phones go to them automatically whether requested or not.
I think GenAI will persist in obnoxious ways, but at least more bearable without folks desperately needing it to be adopted for the sake of their wealth.
After the bubble pops how much would our lives be impacted?
A bunch of pensions will lose a lot of money. The billionaires will know when it’s the right time to short it, and so they’ll get wealthier.
Would AI vanish or still be there?
Higher end models like Mythos will still exist but be too expensive for most people to use. Lower end models will still exist. People will start developing actual tools focused on specific tasks using LLM algorithms instead of AGI. Companies will be able to buy LLM appliances they can put into their server rooms that will help with data analysis which doesn’t require sharing data which will be more useful for use with information they want to keep private.
Eventually people will have PCs, then laptops, then phones with chips optimized for LLM algorithms which don’t require sharing private information with a data center. It won’t be AGI, but it will be able to automate things on the computer and help with stuff like helping with grammar. People will be able to say something “computer, bring up the financial report from last week” and it’ll be able to do that. But it’s not going to be an AGI thing, just be able to find the correct program, files, etc. based on some simple cues.
How exactly do you think the bubble will pop? Will AI companies simply run out of money? Or will it be because of the environmental effects?
Similar to how the dot com bubble popped. Some chatter from billionaires about getting out results in them all shorting it at the same time, the share values collapse, but they make a bunch of money.
Will AI companies simply run out of money?
Yes, when they’ve gotten the last investment dollar, the billionaires will know it’s peaked which will result in a big short.
Or will it be because of the environmental effects?
They don’t care about environmental effects LOL.
When do you think the “pop” will take place?
You need to be invited to whatever is the present day equivalent of Epstein Island to know that.
After the bubble pops, in future there will be companies/people who will try the AI thing again? What will that be like?
Yup. We didn’t stop using the internet after the dot com bubble burst. It’s just the financials of it wasn’t insane anymore. LLMs can’t do everything, but it’s an algorithm that’s useful in some scenarios. It will still be useful in those scenarios, but people will stop being weirdos thinking that it’ll be able to do everything. It’s an algorithm, there’s a cost for the hardware and power to run it, in many cases it won’t work at all, in many cases it’ll work but be cost prohibitive, and then in a few cases it will be actually useful and cost effective. We’ll move to using it for only the latter case.
Over time, it will get better and be used for more things, some of the things they’re promising now may become a reality in a few decades. Kinda like the ridiculous idea to sell dog food on the internet during the dot com bubble. It was ridiculous then, but eventually the logistics caught up and more than a decade after the dot com bubble you could indeed buy dog food on the internet. I imagine it will be like that with AI. Eventually some (though not all) of the things they say you’ll be able to do in six months may actually happen… 15 years from now. But it will probably be enshittified and suck more than you imagined it would be.
The only piece I think you missed is that nation states and surveillance corps will buy up foreclosed data centers for pennies on the dollar to use them to power a surveillance/police state.
You’re assuming they don’t already have a copy. All of the AI companies have government contracts after all.
The AI bubble will pop the same way the dotcom bubble popped. A lot of money lost and the hype will die down, but the technology will remain and reach a homeostasis at the level where it’s actually a helpful tool.
When the dotcom bubble popped the internet didn’t disappear, it just became what it was supposed to be. When the AI bubble pops the technology will still be there for use cases where it’s actually appropriate and useful.
I think part of the process is going to be the enshittification of AI services. (Temporary, until the technology for sustainable infrastructure catches up.) Tokens will cost a premium, and a lot of businesses that depend on AI will turn to China’s cheaper offerings.
Hobbyists will turn more to home-grown AI systems.
In recent news, Chinese AI is catching up to Anthropic. It’ll be interesting if the Trump administration tries to block the Chinese AI offerings in order to bolster US ones. Trump and the AI tech bros are familiar bedfellows after all.
One change for me is celebrating on the anniversary of an AI company shutting down.
Big tech will keep their AI. Like others here have said, AI companies without another income stream will go belly up.
Some data centers will be sold and repurposed. Others will sell off their hardware to the consumer market.
Lower temperatures and less noise near the former data centers.
I don’t see property values going back to where they were unless a data center is dismantled and the water table recovers.
Probably:
- GenAI continues to be used for marketing, and anything else where form matters more than substance
- anything where meaning is really critical might still use it, but will require solid oversight/governance. Software, research, medicine, teaching, etc.
Maybe:
- AI bubble bursts, with ripple effects into the broader economy, particularly in the US. This will be a huge problem if big banks are heavily leveraged for their AI investments (not sure this is the case, but it seems likely). If this happens, the collapse will be much worse than the GFC, perhaps worse than the great depression.
- Political spillover… No idea what this will look like, but it’s a big trigger…
Dogs and cats living together etc.
Social media companies will fill the gaps, continuing to spread their toxicity through society unchecked. The Pied Piper of Hamelin is starting to look more like a cautionary tale.
