The paper also argues that stock gains in companies like Nvidia and other S&P 500 companies, driven by AI optimism, are based on the assumption that generative AI will lead to higher productivity. However, Goldman Sachs warns that stocks often anticipate higher productivity growth before it materializes, raising the risk of overpaying. The report also suggests that AI technology is currently unreliable and may not significantly change anything, despite the hype and investment driving up stock prices.
Key takeaways:
- Goldman Sachs published a research paper questioning the economic viability of generative AI, noting that there is “little to show for” the huge amount of spending on AI infrastructure.
- The paper suggests that while AI optimism is driving large growth in stocks like Nvidia and other S&P 500 companies, the stock price gains are based on the assumption that generative AI will lead to higher productivity, an assumption that may not materialize.
- MIT professor Daron Acemoglu and Goldman Sachs’ head of global equity research, Jim Covello, expressed skepticism about the transformative potential of AI, noting that the technology is currently unreliable and expensive.
- The report follows a piece by David Cahn from Sequoia Capital, which argued that AI companies need to vastly scale their revenue to break even on what they’re spending on AI compute infrastructure, a scale that is not currently happening.