‹ BackHN Continuity

Thread

AI needs $6T in annual revenue to justify data centre boom

222 points · 335 comments · Betelbuddy

  1. 98753579909754 · · focus · HN ↗

    [dead]

    1. runarberg · · focus · HN ↗
      Because it is dumb.

      The behavior of gamblers is a poor measure of ... well, anything. People listen to experts because of their expertise, not because they gambled their money on a bet.

      1. gizajob · · focus · HN ↗
        Shorting the market is a sensible strategy if valuations are as detached from fundamentals and reality as he says they are, so shorting isn’t gambling in every case.
        1. runarberg · · focus · HN ↗
          Shorting the market is a terrible idea if you want people to take you seriously. „Of course Ed Zitron is speaking against the datacenter build-out, he has financial interests that the whole thing collapses,“ people would say.

          Aside: I also fundamentally disagree with you, I think shorting is never a good idea, nor is participating any type of market game a good idea. These market games are a rigged game where the owners of capital engage in systematic exploitation of the working classes. Everyone who makes money by gaming the market made it because hundreds of worker did not get their fair share for their work.

          1. gizajob · · focus · HN ↗
            You’re speaking in Marxist generalities which it would take too much typing to refute.

            It sounds like participation in markets isn’t right for you so don’t do it.

            1. runarberg · · focus · HN ↗
              Note I expressly labelled my “Marxist generalities” as an aside (kind of in an attempt to make my biases clear). You don‘t need to refute it if you don‘t want to. And even if you did very solidly refute them to the point of me saying “I was wrong about everything”, my actual point would still stand. Ed Zitron should in fact not short the market if he wants him self taken seriously.
              1. gizajob · · focus · HN ↗
                Ultimately though he is participating in the market by doing what counts as fundamental analysis - his take on the companies in the market and those still to IPO is that none of the prices are justified because the revenues, even after growing huge in the hardware manufacturers, are in no way justified; that the income in the part of the market where the buck has to stop is not there and can never be there, even to the point of his self-righteous and relentless bleating about all these companies and their technologies being essentially fraudulent.

                If one’s analysis of fundamentals really does point to valuations becoming completely detached and unrealistic, then it is actually sensible to short the market, and indeed, one of the functions of the market is to collectivise intelligence and wisdom about pricing, and shorting stock is a valid mechanism for doing this. People are happy even, to lend stock to shorters for a small premium, who are then free to sell and buy back when their analysis proves to have been justified.

                The issue with Ed is that his analysis wasn’t proven justified two years ago, one year ago, or today. So a large part of his analysis is simply missing, because the main thing he’s commenting on boils down to prices in the market.

                I for one would be respecting him much more if he revealed short positions and gave timeframes. If he is correct then he stands to become phenomenally wealthy. The issue right now is that he isn’t correct, and when the bubble does pop he’s going to be insufferable with his told-you-so-ing.

                Even that day, AI will still be getting used just as much as the day before and its use and capabilities will only be improving. I can’t imagine Ed will be saying “ok these valuations are now actually sensible, you should buy now” if the house of cards comes crashing down… which for me might happen around the time OpenAI and Anthropic have to publicly report earnings (or complete absence of earnings) after they IPO.

                1. runarberg · · focus · HN ↗
                  But Ed Zitron is not an investor, nor a market advisor, nor a hedgefund manager. He is a journalist, and whatever function (capitalists say) shorting stocks has, that is not the responsibility of a journalist.

                  Perhaps you are right that “one of the functions of the market is to collectivise intelligence and wisdom about pricing” however that is the primary function of journalists. And journalists do this by analyzing behaviors and trends, not by participating. In fact participating may be argued is journalistic malpractice.

                  1. gizajob · · focus · HN ↗
                    “ Generative AI is catnip for hall monitors, snitches, toadies, and any other group that hates work and loves talking down to others. Put another way, it ingratiates losers who believe that learning to do or being good at something is a waste of time, because they deserve to just do what they want without any of that messy “effort.””

                    Methinks Ed doth protest too much, as someone who loves talking down to others.

                    Saying journalists who comment on anything to do with business or markets shouldn’t participate in markets is nonsensical. But makes sense for you because it seems like you believe nobody should participate in markets.

                    Money in play or not, Ed is very much participating in the market. And also causing others to participate in the market differently to how they would have without him. At this point he is more polemicist than journalist.

                    1. runarberg · · focus · HN ↗
                      I would actually call him more of a columnist then a journalist, but regardless he does far more actual journalism then your average columnist.

                      I am gonna give you two examples of excellent journalism he has done in the past few months, I cherry-picked these two examples because they are also examples of Ed Zitron being proven right (countering the narrative that started this whole subthread).

                      1. In August this year Ed Zitron co-wrote an article in the Guardian where he and Aisha Down raised suspicion that lots of the GPUs being bought by Microsoft were not going online. A month later Bloomberg verified those suspicion and found that Microsoft had bought 12GW capacity of GPUs, but was only using 2GW of those for AI, meaning 83% of there GPU capacity was undisclosed (possibly just sitting in warehouses)

                      <a href="https:&#x2F;&#x2F;www.theguardian.com&#x2F;technology&#x2F;2026&#x2F;aug&#x2F;17&#x2F;are-microsofts-ai-plans-being-held-back-by-a-shortage-of-chips" rel="nofollow">https:&#x2F;&#x2F;www.theguardian.com&#x2F;technology&#x2F;2026&#x2F;aug&#x2F;17&#x2F;are-micro...

                      2. He has been looking into the financials of Anthropic and OpenAI where he claims that they are spending far more money than is reasonable, and are not getting nearly enough revenue to cover their expenses. The OP’s TFA is vindication of the first point, but today we also have news that Sam Altman is delaying OpenAI’s IPO effectively indefinitely: <a href="https:&#x2F;&#x2F;www.theverge.com&#x2F;ai-artificial-intelligence&#x2F;1002505&#x2F;sam-altman-openai-ipo-devday-ai-safety" rel="nofollow">https:&#x2F;&#x2F;www.theverge.com&#x2F;ai-artificial-intelligence&#x2F;1002505&#x2F;...

                      I have a hard time finding a good link here so instead I will just link to a BlueSky post from August where he made the prediction super clear:

                      <a href="https:&#x2F;&#x2F;bsky.app&#x2F;profile&#x2F;edzitron.com&#x2F;post&#x2F;3mt362mm26k2q" rel="nofollow">https:&#x2F;&#x2F;bsky.app&#x2F;profile&#x2F;edzitron.com&#x2F;post&#x2F;3mt362mm26k2q

                      I think you only consider him a polemicist because you dislike his journalistic findings. It is very common to discredit the credentials of the people you disagree with. I admit I do it all the time with economists who make the case for capitalism.

                      1. gizajob · · focus · HN ↗
                        Far from it - I think his take is essentially correct. I’ve found the lack of income in the AI companies versus the capex it takes to run the AI models to be actually disturbing. Just it was as disturbing two years ago as it is today while the market has continued to invest and rip higher. So the shout of “the market can stay irrational longer than you can stay solvent” really does apply which is why it needs interrogating to find out the deeper reasons and causes.

                        But in Ed Zitron’s world, he is unequivocally correct and literally everyone else in AI and tech and who is investing is wrong. It’s an important data point he has to forcibly overlook that millions of people building and doing more and inventing more things than he is aren’t agreeing with him.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.