in a sane world. in this world, where the system is working as intended, tech bros will reap rewards at our expense. hopefully the working class response will be of a scale commensurate with the scale of the injustice.

Sorry. The techbros will get bailouts. Its the pensioners that will starve.

Remembet that they don't have the money to do it. It's also investors to blame.

That's why the techbros made sure to buy the politicians.

This will be like 208-2012 but this time around it's the techbros businesses which will be deemed "Too big to fail" and get saved with public money and the central bank printing money to buy their bonds, whilst everybody else pays for it with Austerity.

*Techbros

you will have to dish out those consequences otherwise it won't happen.

The lack of progress is one thing.

The value for money is another: AI companies are still burning through vast sums of capital, while users are indeed using the products more and more... but still nowhere near at a price that would even cover the companies’ costs, let alone be profitable. If each token costs a few dollars, demand is likely to shrink significantly again... because as long as the systems operate stochastically and, logically, produce erroneous results, it may take a very large number of tokens to obtain a result that’s actually usable (this is especially true when using AI to replace workers whose tasks and skills you actually have no clue about).

So even if the bet materializes, we lose. If it doesn't, we also lose because the economy implodes. Fun times.

Best thing about this whole AI bubble is that the mask is off and everyone can see. The rich will dump us in a instant.

AGI will never happen with LLMs, but if you think in terms of a human brain out can be effective language and output and image recognition module. But there's no there there. Something else entirely is needed for true thought and sentience. But llm ai is already sufficient to translate machine thought into language sentences. You just need the spark of life that can actually think for itself.

They know climate change will take down capitalism. They are hoping for AGI to save them.

They lose money on those $200 a month subscriptions, but e.g Anthropic won't let you use the sub outside of Claude Code, so for anything that would use their API, $200 gets you less usage than one 5-hour session of that subscription.

Which doesn't stop them from losing a bunch of money to training though

unexpectedly good churro sales?

Propane and propane accessories?

I'm convinced at this point the goal is to just not use money. Why do we have to be tied to this shitty fiat currency that isn't directly tied to physical worth?

Find out by tuning into the next season of The Economy!

Wouldn’t matter. The cost to execute the tokens is greater than the subscription. It’s all VC funds that they’re burning every day. It’s inevitable to fail unless they can convince every day users to pay $2k a month

API costs start to approach the operating costs of these systems. The ultra premium subscription tiers still lose money.

Tech has high market value because there’s the potential for growth, and the cost to grow is minimal after the initial startup costs. Adding a subscriber to a newspaper costs, say $0.20 for each subscriber; adding a Facebook user has a cost approaching zero as the user count grows. The problem with GenAI is that the inference cost grows linearly with usage, and the constant creation of new models means the training costs are also not recouped.

The debt the datacenter builders are taking on are being traded as junk with high interest rates already, and they’re going higher. They’ll never generate the income that the rates demand much less recoup the cost of building.

QUD

Live and drink, friend!

They've been comprised. Your mission, should you choose to accept it, is to infiltrate the IMF and remove compromised agents by any means necessary. If you are captured or killed the Secretary will disavow any knowledge of your actions. Good luck. This message will self-destruct in 5....

The problem is not the "losing of AI", it's the loss of human and physical resources that went into AI and will ultimatelly yield little or no Economic value when they could have been gone into productive Economic activities.

The human part is similar to COVID - people's time that could have been used in Economically productive activities was lost and will never come back - however the AI bubble differs in that actual physical resources have been wasted (mainly building and equiping datacenters) whilst during COVID fewer physical resources were used (mainly in vaccines and PPE).

Also the AI bubble is mostly concentrated in the US, whilst COVID was global, which means that the damage of it blowing up will probably be worse in the US, whilst that of COVID was much more diffused.

But the really big things is not even the AI bubble blowing up per-se - it's that there are other bubbles which are ready to blow (such as realestate Worldwide and auto-loans in the US) plus a worldwide Oil crisis caused by America's war against Iran, so the Economic shock of the AI bubble blowing will be amplified by the latter and will likely make the former blow up, almost certainly adding up to even more damage that the 2000 and 2008 Crashes (possibly as much as both put together).

COVID was a slowdown of Economic productivity, this shit will probably be the Find Out stages all on top of each other of years of Fuck Around misallocating resources in multiple Economic domains.

/s mandatory

Live in the woods?! Are you crazy, you need to work so you can pay taxes that will become their bailout!

You can go starve live in the woods when AI can do your job after stealing your data.

2 exceptions to what you said.

1. Whatever money raised through IPOs and other public investment is mostly the general public's saving accounts. So far it's just $80bn in the SpaceX IPO and $40bn from the bond issue they did shortly after.
2. Just like the mortgage crisis it's the banks holding the debt that's been issued. Now since 2008 they were meant to firewall between their retail banking arms and their investment arms so that if they fucked up again the retail banks were safe. However this legislation got repealed in the US, so we could be looking at a bank bailout again to make sure the banking system doesn't collapse with everybody's money in it.

The money loaned still exists, you're right. Trouble is it's not in the AI companies anymore. It's been spent. It's gone through them out to other companies (like Nvidia, MS Azure, AWS, Google Cloud, etc).

You think they aren’t harvesting your data for advertising?

And even if you are somehow anonymized, any usage to juice their numbers is what they want.

If you want to stick it to ChatGPT, use their competition. GLM, Xiaomi, Deepseek, Mistral, Nemotron, whatever;their absolute deepest fear is the public pivoting to actually free models. And you can use them with whatever level of privacy you wish, since you aren’t locked to vendors.

It's not taking the global economy away in that sense.

It'll blow up the global economy in that a lot of people will be unemployed and everyone's retirement funds will crash (which is fair, because AI is what made them go up so fast in the first place). A few rich people will also lose a bunch of money, but the clever ones will actually profit off it.

Yea, good. Billions of people should starve so a couple of wealthy people lose a bit of money but not enough to need to sell their yachts. Let's blow up the global economy!