frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Powell – unlike the dotcom boom, AI spending isn't a bubble

https://fortune.com/2025/10/29/powell-says-ai-is-not-a-bubble-unlike-dot-com-federal-reserve-interest-rates/
29•madaxe_again•15h ago

Comments

embedding-shape•15h ago
I tried submitting this but landed here. I think I managed to cobble together a better title though: "US Federal Reserve Chair says that, unlike the dotcom boom, AI isn't a bubble"
wslh•14h ago
https://archive.is/Imalt
wslh•14h ago
I don't buy Powell view from his comment: you can have revenue but not profits, and also the dotcom [bubble] at the end is what makes AI possible.
rmah•14h ago
While I agree with your general sentiment, in fairness, Powell didn't say "revenue", he said "earnings", which is near equivalent to profits in financial reporting lingo.
wmf•14h ago
But the AI companies do not have profits. Even Google/MS/Meta are probably subsidizing AI with the profits from their non-AI businesses.
wslh•14h ago
I assume that in this context is basically about NVIDIA, and some other chip companies.
dragonwriter•12h ago
But NVidia's AI related profits are driven by the AI-related speculative spending in the uncertain hope of future profits by firms in the AI gold rush, they are just the leading seller of shovels.
emp17344•12h ago
Powell is so influential as the Fed Chair that he’s constrained in what he can even say. If he outright says there’s a tech bubble, that statement alone could cause the bubble to pop. I’m not convinced these are his actual views - he’s saying what needs to be said to protect the economy.
FrustratedMonky•14h ago
The problem is, that in past bubbles, there were also people saying 'This is not a bubble".
NoOn3•13h ago
Alan Grenspan also once said that he did not believe in the existence of a national housing bubble like the stock market bubble.
jgalt212•14h ago
says the man who printed $5T, $2T of which was printed after the economic effects of the pandemic were well in hand.

https://fred.stlouisfed.org/series/WALCL

ares623•14h ago
Please buy OpenAI IPO or we’re screwed.
Marshferm•10h ago
Inst funds are asking: what do we own when we invest, and the answers aren’t clear nor are they clarifying week by week.
ares623•9h ago
You can brag at the next dinner party you’re an AI investor
pac0•14h ago
This is the confirmation that AI is a bubble...
sindriava•14h ago
ITT: Keyboard warriors with no background in economics deciding if multinational corporate deals are a bubble or not
Esophagus4•14h ago
...without taking a position and putting their own skin in the game.

For most people in the comments section of "AI bubble" articles, it seems like "big numbers == bubble"

sindriava•14h ago
People also think that two companies doing business with each other is somehow a "circlular dependency".
stackskipton•14h ago
ITT: people believe that those in jobs with big educational backgrounds might have incentives to say something not 100% backed by evidence because it would impact their job.
emp17344•12h ago
A careless comment from the Fed Chair could quite literally destroy the economy. He’s constrained in what he can say - you almost never get to hear what Powell is really thinking.
myhf•14h ago
smh, keyboard warriors with no background in tailoring deciding whether the emperor has clothes or not
smj-edison•14h ago
> Powell’s framing echoes that view: The AI race, while frothy at times, is being financed mainly through corporate cash flow rather than speculative debt.

I think this is the most important line. Even if these companies are overvalued, I don't think there will be a nasty pop (this could change if banks start getting more involved, but so far it's been mainly company reserves and VC funding).

dathinab•14h ago
> corporate cash flow rather than speculative debt.

except that this doesn't quite add up IMHO

1. we have the current _speculation_ on an explosive need of both more data centers and energy if AI has a break through. This investments are mostly not from the big tech but VC/higher risk investment soured.

2. VC isn't always just private equity, I have seen enough rent founds, banks and similar acting as VC investors. Sure mostly by proxy, but they are doing that anyway.

3. 1st point also involves a lot of loans handled and resold by banks...

4. even with out all this a lot of founds are based on S&P 100 and similar. If the AI bubble goes pop before it's filled with enough magnetizable value a lot of the companies in there will lose a non negligible amount of valuation, which will have shock waves across much more then just "big companies".

sseagull•14h ago
That’s starting to change though

https://www.bloomberg.com/news/articles/2025-10-31/meta-xai-...

(Archive link: https://archive.ph/NBNEu)

smj-edison•14h ago
Hmm, interesting. I hadn't realized how much companies were starting to borrow.
jameslk•13h ago
Borrowing to fund AI datacenter spending: https://x.com/MikeZaccardi/status/1984036161426485358
ares623•2h ago
I feel proud that my 50 PRs are a part of that. Truly a worthy investment. Here's to the next 50 PRs.
JKCalhoun•13h ago
First paragraph for those unsure whether they want to prove that they are human (and for non-humans of course):

"Just this month, Meta Platforms Inc. has secured about $60 billion in capital to build data centers, part of its spending to get ahead in the artificial intelligence race. Half of that won’t show up on the social media giant’s balance sheet as debt.…"

techblueberry•14h ago
I thought I’ve been reading that more and more of the money financing the AI boom is debt. The numbers are big enough I’d expect Powell to know:

https://techstartups.com/2025/10/31/the-hidden-debt-behind-t...

https://www.bloomberg.com/news/newsletters/2025-10-06/big-de...

I think there’s any way we don’t see contagion. There’s too much money involved.

bostik•14h ago
I'd say the lines just before that are more illustrative of _why_ this sentiment is even possible to express: Goldman team wrote. “AI investment as a share of U.S. GDP is smaller today (<1%) than in prior large technology cycles (2%–5%).” In other words, there’s still plenty of room to run.

However, someone has now uttered the four words we should all recognise. "This time it's different."

blibble•14h ago
technically correct, but misleading

because they're doing deals to keep the debt off the balance sheet

the fact they are doing this says rather a lot

stego-tech•13h ago
The pop will be nasty for everyone that’s heavily exposed to the public players involved, which is a sizable amount of the public given their outsized impact on indices.

That said, it’s the sort of bubble pop that’s far more constrained than, say, the dotcom bubble (with tons of IPOs) or subprime mortgage bubble (when a variety of financial instruments and securities obfuscated risks to investors). It’ll hurt for the highly-paid folks working for the companies who suddenly have to downsize, and it’ll hurt for investors who are all-in on AI without diversification, but for everyone else? It’ll be a gloomy year, but nothing as spectacular as prior collapses.

dathinab•14h ago
> but they actually have earnings’

which doesn't mean it's not a bubble

a bubble is a over valuation of a industry sector

it isn't a "it has no use at all"

it's a "the use as big as assumed", "the use isn't worth the money invested in it" etc.

and so far that's the trend with AI, it has use, you can make money with it, but not _enough_

But there is just that much money you can get from people for the services OpenAI is currently usable/used for. Some (e.g. search, generating shopping lists, grammar correction) of which people expect to be cheap or even fully free and some (coding assistance) which involve other companies in the value chain.

Add circular money deals to that.

And it's very clearly a bubble.

But if you can drag out a bubble long enough you might be able to largely fill it up before it pops.

Through with how confidence in AI is moving + the general world economical situation I wouldn't bet on it.

skywhopper•14h ago
Remember when Alan Greenspan said the housing market needed to encourage more adjustable rate mortgages?
Ekaros•14h ago
Wouldn't clear overvaluation fit to term of bubble? And well if these companies are simply burning chunk of their profits and nothing turns out, wouldn't that affect their valuations? Like wouldn't any traditional company that just wasted good chunk of profits it made be valued at lower market price at somepoint?

And then you get to Nvidia. Which to me looks like clear case where price must correct if their profit drop due to demand dropping.

So even if there is no "pop", there will likely be a deflation and at these valuations even that will have impacts.

jameslk•14h ago
When your Federal Reserve Chairman starts giving you stock tips, the market top is near
resoluteteeth•13h ago
This. When he has to specifically tell people "it's not a bubble" that is not at all a good sign in my opinion.
JKCalhoun•13h ago
"I won’t go into particular names, but they actually have earnings."

Why not go into particular names? Without doing so the second half of your sentence is worthless.

Meta? Of course they have earnings. Not AI earnings.

Alphabet? Apple?

We kind of need particular names.

stego-tech•14h ago
I think I get what Powell’s laying down, here, from an economics perspective. His data shows that a non-zero number of major players in the AI space are posting earnings, not just revenue, and that this is a sign of healthy economic activity. He’s also saying that while CAPEX investment today is coming primarily from corporate books (and not major banks), he’s also stressing that these investments have a long, uncertain tail - and that at present, between lofty promises from AI companies and executives eager to hoard capital, that the job market is seeing significant negative impacts that have zeroed out job growth while capital growth remains on a rocket ship to the stratosphere.

As someone who follows the man’s speeches and comments, I’ve always found that the real message is between the lines of what he says, especially for things he cannot say explicitly. So reading his comments here, my takeaways are:

* Zero job growth is a serious concern, especially as companies and business elites continue to demand a return to ZIRP

* Inflation also remains a concern in that same context, as it continues to skyrocket for essential goods

* Even if this AI investment is highly speculative and collapses, its “blowback” is broadly contained vis a vis the AI “sphere” being a handful of companies in circular investment schemes, not mass public buy-in like prior bubbles

* His immediate concern remains trying to figure out how to address rising inflation, stagnant wages, and negative/zero job growth in the face of a red hot securities market and continued trade uncertainties. The only levers left to pull will disproportionately hurt the working class even further, and he seems to continue suggesting that this isn’t an issue the Fed can solve absent Congressional intervention (which, let’s face it, ain’t happening)

That’s my takeaway from his comments this year, and it fits my own perspectives on the AI Bubble (in that yes it’s a bubble, but it’s a highly concentrated bubble that’s painting over broader harms and misleading public officials about the state of the economy). Being head of the Fed, he can’t come right out on either side of the fence and call AI a bubble or a real boost, hence why he’s couching comments deeply within economic terms alone.

Just my two cents, anyway.

PorterBHall•13h ago
I think your take makes sense. Thanks for laying it out. I worry about the containment aspects of this. Sure, the capex for this is coming off corporate balance sheets, but a bad event could wipe away a lot of stock market value, which would trigger a deep recession given that AI spending is really the only thing keeping our economy afloat at the moment. It feels like we’re in a doomed if we do, doomed if we don’t moment. I see breadlines in our future whether or not we achieve AGI. The only question is whether or not we’ll have to deal with some added existential risk.
sema4hacker•13h ago
"Here's How the AI Crash Happens:" https://finance.yahoo.com/news/ai-crash-happens-215009627.ht...
JKCalhoun•13h ago
"The amount of energy and money being poured into AI is breathtaking. Global spending on the technology is projected to hit $375 billion by the end of the year and half a trillion dollars in 2026."

When you think what we could have done with that money instead…

I'm not as cynical about the state of the world as others but from time to time I have to shake my head when I consider how much better we could have it. That to me is the real tragedy.

yummypaint•13h ago
Vaguely projecting forward the relationship between total cumulative funding and experimentally achieved q value, it's not unreasonable to think that this money could have brought us to commercially viable fusion power if spent judiciously over a longer period of time. It's pretty typical that we collectively spent the money on an orgy of waste instead.
latchkey•13h ago
DDTL is the new subprime.
yahoozoo•13h ago
What would actually cause the bubble to pop?
wmf•13h ago
The bubble will pop if/when credit markets stop providing new money. They would do that if they just run out of money to lend or they decide that AGI is impossible.
amypetrik8•7h ago
investors dry up, prompted by a scandal (or not)

or someone figures out how to do AI at 1/100000th the CPU/memory on the edge

a company or two collapses

all the other companies now back way off

now the bubble is popped

AI returns 5-15 year later

naveen99•13h ago
What timeframe ? You get bubbles and pops on the minute candle chart daily.
dragonwriter•12h ago
What a bizarre statement if taken at face value as a description and not just as talking up the economy, even if the explicit premise is true; there were plenty of firms in the dotcom bubble that actually had real earnings and sustainable business models (they are, in some cases, still tech giants today); it was still a bubble; the contrast drawn requires false assumptions about what happened in the dotcom boom.
keernan•12h ago
I would be shocked if the Fed Chairman intended his comments as advice regarding the relative valuation of the price of the SP500.

In 1999 people correctly understood that this thing called the internet was going to transform life like nothing before it had ever done. And they were correct. But that has very little to do about investing in stocks. Just because you have an insight into a fantastic technology doesn't mean the price being asked to purchase stock will earn a profit. And there is no way Powell would opine on whether now is a good time to buy, sell, or hold stocks.

My reading of Powell's comments is they were directed upon the stratospheric dollar amounts being spent creating AI data centers and what that means for the economy. There is just no way he is providing advice about the price of stocks.

Marshferm•10h ago
Powell’s Greenspan moment.

Backpropagation is a leaky abstraction (2016)

https://karpathy.medium.com/yes-you-should-understand-backprop-e2f06eab496b
123•swatson741•4h ago•44 comments

Why do AI models use so many em-dashes?

https://www.seangoedecke.com/em-dashes/
39•ahamez•3h ago•34 comments

Notes by djb on using Fil-C (2025)

https://cr.yp.to/2025/fil-c.html
69•transpute•4h ago•13 comments

Visopsys: OS maintained by a single developer since 1997

https://visopsys.org/
328•kome•11h ago•63 comments

How I use every Claude Code feature

https://blog.sshh.io/p/how-i-use-every-claude-code-feature
255•sshh12•9h ago•72 comments

CLI to manage your SQL database schemas and migrations

https://github.com/gh-PonyM/shed
21•PonyM•3h ago•9 comments

Claude Code can debug low-level cryptography

https://words.filippo.io/claude-debugging/
316•Bogdanp•15h ago•152 comments

Crossfire: High-performance lockless spsc/mpsc/mpmc channels for Rust

https://github.com/frostyplanet/crossfire-rs
64•0x1997•6h ago•4 comments

Updated practice for review articles and position papers in ArXiv CS category

https://blog.arxiv.org/2025/10/31/attention-authors-updated-practice-for-review-articles-and-posi...
448•dw64•19h ago•206 comments

Pomelli

https://blog.google/technology/google-labs/pomelli/
177•birriel•10h ago•53 comments

FlightAware Map Design

https://andywoodruff.com/posts/2024/flightaware-maps/
20•marklit•5d ago•8 comments

LM8560, the eternal chip from the 1980 years

https://www.tycospages.com/other-themes/lm8560-the-eternal-chip-from-the-1980-years/
35•userbinator•5h ago•11 comments

Hyperbolic Non-Euclidean World (2007)

http://web1.kcn.jp/hp28ah77/
12•ubavic•6d ago•1 comments

The Naked Man Problem and the Secret to Never Forgetting Numbers

https://ninjasandrobots.com/the-naked-man-problem-and-the-secret-to-never-forgetting-numbers
22•nate•4d ago•37 comments

GHC now runs in the browser

https://discourse.haskell.org/t/ghc-now-runs-in-your-browser/13169
307•kaycebasques•17h ago•95 comments

When O3 is 2x slower than O2

https://cat-solstice.github.io/test-pqueue/
4•keyle•4d ago•0 comments

Show HN: Why write code if the LLM can just do the thing? (web app experiment)

https://github.com/samrolken/nokode
315•samrolken•16h ago•224 comments

Anonymous credentials: rate-limit bots and agents without compromising privacy

https://blog.cloudflare.com/private-rate-limiting/
67•eleye•9h ago•29 comments

SQLite concurrency and why you should care about it

https://jellyfin.org/posts/SQLite-locking/
304•HunOL•21h ago•138 comments

Automatically Translating C to Rust

https://cacm.acm.org/research/automatically-translating-c-to-rust/
53•FromTheArchives•1w ago•12 comments

3M Diskette Reference Manual (1983) [pdf]

https://retrocmp.de/fdd/diskette/3M_Diskette_Reference_Manual_May83.pdf
77•susam•5d ago•16 comments

Beginner-friendly, unofficial documentation for Helix text editor

https://helix-editor.vercel.app/start-here/basics/
132•Curiositry•14h ago•39 comments

From 400 Mbps to 1.7 Gbps: A WiFi 7 Debugging Journey

https://blog.tymscar.com/posts/wifi7speedhunt/
104•tymscar•14h ago•78 comments

The Smol Training Playbook: The Secrets to Building World-Class LLMs

https://huggingface.co/spaces/HuggingFaceTB/smol-training-playbook
188•kashifr•2d ago•12 comments

You Don't Need Anubis

https://fxgn.dev/blog/anubis/
106•flexagoon•6h ago•83 comments

Chip Hall of Fame: Intel 8088 Microprocessor

https://spectrum.ieee.org/chip-hall-of-fame-intel-8088-microprocessor
22•stmw•6d ago•0 comments

A Few Words About Async

https://yoric.github.io/post/quite-a-few-words-about-async/
46•vinhnx•8h ago•15 comments

SailfishOS: A Linux-based European alternative to dominant mobile OSes

https://sailfishos.org/info/
270•ForHackernews•11h ago•112 comments

How to Build a Solar Powered Electric Oven

https://solar.lowtechmagazine.com/2025/10/how-to-build-a-solar-powered-electric-oven/
51•surprisetalk•1w ago•24 comments

Dating: A mysterious constellation of facts

https://dynomight.net/dating/
94•tobr•2d ago•87 comments