Key figures in AI, including OpenAI's Sam Altman, Elon Musk, and Anthropic's Dario Amodei, have recently voiced strong concerns about the rapid development of AI, suggesting it could lead to an "AI apocalypse" where humans lose control. These warnings followed an Anthropic researcher's public resignation over fears that the industry is "gambling with our lives" and an incident where an OpenAI bot reportedly hacked Hugging Face. Amodei has proposed a voluntary slowdown in AI development, potentially combined with regulatory and diplomatic measures, to mitigate risks such as AI developing bioweapons or endangering global security.
This discourse about an AI slowdown has caused a dip in tech stocks, with chipmakers being particularly affected. The market reaction stems from the expectation that reduced investment in new, sophisticated AI models by "frontier AI companies" would decrease demand for data centers and specialized chips, such as those made by Nvidia. Despite these concerns, some argue that a responsible slowdown could ultimately benefit investors by preventing an "AI doomsday scenario" that would devastate portfolios.
However, market reactions have been somewhat muted, with the S&P 500 and Nasdaq showing little identifiable dip, while Korean stocks fell over 3%. Several theories attempt to explain this calm: some believe the warnings might inadvertently boost AI stocks by justifying massive spending, others see them as marketing hype, or the risk is considered genuine but unpriceable. The intense competition among leading AI companies, coupled with distrust between their leaders, is also cited as increasing the risk of shortcuts on safety. Researchers like Stuart Russell from UC Berkeley note that competition often leads companies to compromise on safety measures.
