[wronglebowski]

At what point will the massive investments into AI show a respectable return? With the literal Trillion dollars OpenAI is constantly trying to raise what type of revenue would make that type of investment make sense? Even if you're incredibly bullish I don't know how you make that math work anymore.

[jt2190]

I think it’s hard for individuals to think at the scale of very large institutional investors. They have lakes of money [1][2] that they have to invest in a balanced way, including investing a small percentage into “it probably won’t work but if it does we’ll make a fortune”-type bets. Given the size of these funds even a small percentage is a very large number.

There are also a finite number of opportunities to invest in, so companies that have “buzz” can create a bidding war among potential investors that drives up valuations.

So that’s one possible reason but in the end we can’t know why another investor invests the way they do. We assume that the investor in making a rational decision based on their portfolio. It’s fun to speculate about though, which is why there’s so much press attention.

[1] https://en.wikipedia.org/wiki/List_of_largest_pension_scheme...

[2] https://en.wikipedia.org/wiki/List_of_sovereign_wealth_funds...

[Forgeties79]

The problem is we now have municipal and state governments taking on infrastructural investments (usually via subsidies) and energy companies racing to meet the load demands. There are all kinds of institutions both private and public dumping obscene amounts money into this speculative investment that can’t be a winner for everybody.

What happens to the ones that built for projects that end up failing? Seems to me the only way the story ends is with taxpayers on the hook once again.

[jt2190]

Yes, many of us are investing in this (even indirectly) and may not realize it! The same rules still apply for municipal and state treasuries though: Only a small percentage of the overall portfolio should be allocated to high-risk investments.

Power generation, power grids are more generally useful today and less speculative than trying to win the AI race, so the risk for those types of things is somewhat lower, but there IS risk even in those.

[Forgeties79]

I think the concern is something like the huge Entergy investment going on in Louisiana. Facebook basically cut a deal with them to build out all kinds of electrical load just for them. We also saw how committed Facebook was to the metaverse - they basically spent the GDP of a small nation, nothing came of it, fired a bunch of people, and moved on.

Entergy is not just going to sit around and take the L if the project doesn’t ultimately turn out to be a good long-term investment. They’re simply going to pass the cost on to their customers in the region (more so than they already plan to in the event of success). Meanwhile Louisiana taxpayers are footing the bill for all the subsidies going through these projects.

So yeah I agree it’s not quite as high risk because at least there’s some in infrastructural investment, but that’s not the kind of investment that is really needed in the region right now and having that extra capacity is not a good thing unfortunately.

To be clear I’m not really disagreeing with you. I’m just kind of bickering over the nuances lol

[cmrdporcupine]

I think the market & political/economic actors as a whole are justifying these investments on the basis that the benefit is distributed across the labour market generally.

That is, it doesn't matter so much if OpenAI and individual investors get fleeced, if there's a 20-50% labour cost reduction generally for capitalism as a whole (especially cost reduction in our own tech professions, which have been very well paid for a generation) -- Institutional investors and political actors will benefit regardless, by increasing the productivity or rate of exploitation of intellectual / information workers.

[rramadass]

Why Fears of a Trillion-Dollar AI Bubble Are Growing - https://www.bloomberg.com/news/articles/2025-10-04/why-ai-bu...

What Would an AI Crash Look Like? - https://www.bloomberg.com/news/newsletters/2025-10-12/what-h...

[tim333]

The bull thesis is you get human level intelligence and then it can do jobs like us.

[rossdavidh]

As others have said before me: "the hype IS the product".

[shiandow]

That's just a roundabout way of saying they don't expect their money back they just hope to sell before the bubble bursts.

[lazide]

Modern American investing 101.

[IAmGraydon]

Which is the same thing as a scam.

[coliveira]

It's gonna be very fun to watch SA being tried for fraud and deceiving investors about the "future profits" of his startup.

[cmrdporcupine]

Au contraire him and associated people (Musk, among others) will be or are being received as heroes in his class for helping to seemingly break the bargaining power of software engineers and middle/upper-middle class information workers and creatives generally.

[lazide]

The folks who would have to press charges are the folks who would be far too embarrassed to admit how transparent the fraud was they fell for - and nuke any remaining asset value.

It’s why Musk is also safe from similar problems.

[otabdeveloper4]

Akshually, watch the hands. He never talks about profits at all, only about "disruptions". No refunds.

[lesuorac]

What fraud?

You should view your contributions as a donation. What donation has a ROI?