The artificial intelligence (AI) party is still in full swing, with tens of billions being invested globally into infrastructure, startups, and attracting the best talent.
Among this year’s key announcements: ChatGPT’s parent company Open AI, SoftBank and Oracle pledged to invest $500 billion (€433 billion) in AI supercomputers, Open AI and chip giant Nvidia committed a $100 billion fund to maintain the United States’ dominance in advanced chips, while Chinese tech giants Alibaba and Tencent pledged to accelerate China’s ambition to lead AI by 2030. Investments announced to help.
According to Bloomberg Intelligence, since ChatGPT’s debut in November 2022, AI-related stocks have added an estimated $17.5 trillion in market value, driving nearly 75% of the S&P 500’s gains and driving companies like Nvidia and Microsoft to record-breaking valuations.
Corporations are hesitant to adopt AI
But the signs of a hangover are becoming harder to ignore. The use of AI by corporations is declining, spending is falling, and the promotion of machine learning has largely overtaken profits.
Many economists believe that usage concerns, given that AI has taken barely three years to become mainstream, defy the prevailing narrative that AI will revolutionize the way businesses operate by streamlining repetitive tasks and improving forecasting.
“The big bet on AI infrastructure assumes growing use, yet several US surveys show that adoption has actually declined since the summer,” Carl-Benedict Frey, an AI professor who works at the University of Oxford in the UK, told DW. “Unless new, sustainable use cases emerge rapidly, something will be found – and the bubble may burst.”
The US Census Bureau, which surveys 1.2 million US companies every fortnight, found that AI-tool use among firms with more than 250 employees dropped from nearly 14% in June to less than 12% in August.
AI’s biggest challenge remains its tendency to hallucinate – generating plausible but false information. Other weaknesses are inconsistent reliability and poor performance of autonomous agents, which successfully complete tasks only about a third of the time.
“Unlike an apprentice who learns on the job, today’s pre-trained [AI] Systems do not improve with experience. We need continuous learning and models that adapt to changing circumstances,” Frey said.
burning of unsustainable capital
As the gap between lofty expectations and business reality widens, investor enthusiasm for AI is beginning to wane.
Market intelligence firm CB Insights wrote last month that in the third quarter of the year, venture-capital deals with private AI firms declined 22% quarter over quarter to 1,295, although funding levels remained above $45 billion for the fourth consecutive quarter.
“The scale of money being invested compared to the amount of revenue coming from AI worries me,” economist Stuart Mills, a senior fellow at the London School of Economics, told DW.
Market leader OpenAI, which is backed by Microsoft, generated revenue of $3.7 billion last year, while total operating expenses were $8-9 billion. The company says it is on track to earn $13 billion this year, but still expects to earn $129 billion before 2029, news site The Information calculated in September.
Mills believes that generic AI companies like Elon Musk’s Grok and ChatGPT are “charging much less than they need to make a profit” and should raise subscription prices.
Few people have pinpointed the AI bubble more strongly than Julian Garran, partner at UK-based research firm MacroStrategy Partnership. He argues that the sheer volume of capital flowing into AI – despite little evidence of sustainable returns – dwarfs previous speculative frenzies.
“We estimate that capital has been misallocated equivalent to 65% of US GDP – four times larger than the housing boom before the 2008/9 financial crisis and 17 times larger than the dot-com meltdown,” Garan told DW.
Investors are becoming increasingly cautious
Big Tech’s recent earnings have sparked cautious optimism, but also fresh skepticism about the desired power of AI. Data analytics and intelligence platform Palantir’s Q3 revenue rose 63% year-over-year, but its stock price fell 7% following the news. AMD and Meta also saw their strong AI-related earnings impacted by market concerns about sustainability.
The gap between rising valuations and shaky fundamentals really worries Mills, who sees a wide gap between the promises of AI and what it actually delivers to businesses.
“The figures show that AI is not penetrating the value chain enough. A lot of people are using it, but it is not being used for tasks that directly contribute to value production,” he told DW.
Nvidia’s upcoming earnings on November 19 could prove to be an important test of whether the AI boom still has legs. In the second quarter, Nvidia’s data center sales alone made up 88% of total revenue, reaching a record $46.7 billion. For Q3, the company guided for $54 billion, which is projected to be 54% year-over-year growth, which would bring the full-year total to more than $200 billion.
When will the bubble burst?
“With the exception of Nvidia, which is selling shovels in a gold rush, most generative AI companies are wildly overvalued and wildly overpriced,” Gary Marcus, professor emeritus of psychology and neuroscience at New York University, told DW. ,My guess is that this will all probably fall apart soon. The fundamentals, technical and economic, don’t make sense.”
Meanwhile, Garran believes that the era of rapid progress in large language models (LLMs) is coming to an end not because of technical limitations, but because the economics no longer stack up.
“They [AI platforms] “are already in trouble,” Garan said, adding that the cost of training new models “is skyrocketing, and the improvements aren’t much better.”
Speaking in a more positive tone, Sara Hoffman, director of AI thought leadership at New York-based market intelligence firm AlphaSense, predicted a “market correction” in AI rather than a “catastrophic ‘bubble burst’.”,
After an extended period of extraordinary hype, enterprise investment in AI will become far more sensible, Hoffmann told DW in an emailed statement, adding that the focus is “shifting from big promises to clear proof of impact.”
“More companies will start formally tracking AI ROI [return on investment] To ensure that the projects provide measurable returns,” he further said.
Edited by: Uwe Hessler