Firm says Nvidia's skyrocketing AI valuation is in a 'bubble' and 'overhyped' — Elliott says AI apps are not viable

Nvidia
(Image credit: Nvidia)

Elliott Management, a hedge fund with approximately $70 billion in assets, has raised concerns about Nvidia and the AI industry, describing them as being in a 'bubble,' reports Financial Times. The firm doubts the sustainability of current AI investments and believes that many AI applications are not viable or cost-effective.

The hedge fund expressed skepticism about the ongoing purchases of Nvidia's GPUs by large technology companies, stating that these companies may not continue to buy in such large quantities. Elliott Management questioned many AI technologies' practical utility and efficiency, suggesting that they may not deliver the promised returns. Elliot Management is certainly not the only one to express concerns about the AI industry as a partner of Sequoia Capital recently calculated that the AI industry needs to make at least $600 billion per year to pay for the already made investments. The AI industry is not even close to that number.

Elliott's commentary comes amid a rally in AI-related stocks, driven by investor enthusiasm for generative AI, the FT says. Meanwhile, Nvidia has lost nearly $600 billion in market capitalization since early July, and that says something.

Recent trends have shown a general pullback in semiconductor stocks, reflecting concerns about the durability of spending in this sector. For example, Intel shares dropped 30% after announcing significant layoffs, highlighting industry volatility.

The hedge fund warns that a market correction could occur if Nvidia's financial results disappoint, potentially shaking investor confidence in the AI sector. Meanwhile, most AI sector companies are private, which almost wholly means there is no adequate view of their financial viability.

Elliott notes that, so far, AI has not delivered the significant productivity boosts promised, with most applications limited to tasks like summarizing notes, generating reports, and assisting in coding.

Even considering Elliott Management's cautious approach to its investment strategy (the firm has largely avoided what it calls 'bubble stocks'), according to FT, its stake in Nvidia is minimal, and it was worth about $4.5 million as of March.

Anton Shilov
Contributing Writer

Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.

  • Alvar "Miles" Udell
    People and experts have been saying this for some time, it's only starting to become more evident to people who don't really follow financial news as companies banking on AI, like Microsoft and Google, are continuing to show dismal AI profits that don't justify the hundreds of millions in investment in it.

    Just better hope it deflates in spurts, like last week, and not all at once.
    Reply
  • yahrightthere
    So many so called experts have crystal balls & can say watch out for this that & the other, that any investor that relies solely on what others say without doing their own due diligence get what they deserve.
    I believe that diversification is key to investing, along with knowing when to take profits & how to mitigate losses, & evaluate a companies financials.
    How will Nvidia do in the coming months remains to be seen, there are those for & those against, only time will tell.
    My own belief is that if a stock doubles in value, sell half to get the initial investment back & see where the stock goes on a longer term.
    If a stock takes a 15-20% loss in value from the initial investment you may want to sell.
    Reply
  • ThomasKinsley
    I see AI marketing as an economic football. Coming out of the pandemic companies are desperate for any reason to justify a sale, and if that means harping on AI for the next 2-5 years then so be it. They'll find another marketing strategy after that.
    Reply
  • Alvar "Miles" Udell
    ThomasKinsley said:
    I see AI marketing as an economic football. Coming out of the pandemic companies are desperate for any reason to justify a sale, and if that means harping on AI for the next 2-5 years then so be it. They'll find another marketing strategy after that.

    "AI" is great, but right now it's like an electric car in 1839: Not practical. Cost to develop is too high, and the products it deliver 99.9% of people aren't going to pay for.
    Reply
  • ThomasKinsley
    Alvar Miles Udell said:
    "AI" is great, but right now it's like an electric car in 1839: Not practical. Cost to develop is too high, and the products it deliver 99.9% of people aren't going to pay for.
    Agreed. I've been experimenting with LMStudio. I've been thoroughly impressed with its free, offline models on my older hardware that I don't think I'll use ChatGPT any longer. They're going to need to significantly increase the intelligence if they want to monetize it.
    Reply
  • MacZ24
    AI is extensively used in a lot of fields. If you used a website that provides recommandations, it most probably uses machine learning algorithms to create those recommandations for example.

    But chatbots, though impressive at first glance, are unreliable and cost a fortune to setup and train. And there are a lot of legal hurdles that have not been overcome yet and that leads to uncertainty.

    AI makes sense as dedicated tools for professionnals.
    Reply
  • Alvar "Miles" Udell
    ThomasKinsley said:
    Agreed. I've been experimenting with LMStudio. I've been thoroughly impressed with its free, offline models on my older hardware that I don't think I'll use ChatGPT any longer. They're going to need to significantly increase the intelligence if they want to monetize it.

    Even say AI was mature, efficient enough to run locally at a good speed, and did everything you wanted. Would you pay $20 a month/$240 a year for it like Google and Microsoft want now? It's a little different for professional applications, like law and medicine, where it's a small price for a huge benefit, but for the average consumer it's going to have to be included with an existing Microsoft 365/Google One/Apple One subscription for no or minimal additional cost for them to pay for it.
    Reply
  • ThomasKinsley
    Alvar Miles Udell said:
    Even say AI was mature, efficient enough to run locally at a good speed, and did everything you wanted. Would you pay $20 a month/$240 a year for it like Google and Microsoft want now? It's a little different for professional applications, like law and medicine, where it's a small price for a huge benefit, but for the average consumer it's going to have to be included with an existing Microsoft 365/Google One/Apple One subscription for no or minimal additional cost for them to pay for it.
    Considering I'm turning down ChatGPT which is free, I have very high demands. 😄

    I agree with your assessment. I would consider an extremely mature, offline AI for that price, but it would have to make a massive improvement in my workload to justify the cost.
    Reply
  • Notton
    I'm not an expert, but this AI stuff doesn't pass the sniff test.
    It doesn't help this is hot on the heels of the NFT boom.
    Reply
  • thisisaname
    Yet another story that is looking to reduce Nvidia share price..
    Reply