A claim that, once the AI bubble bursts, AI experts will exit the industry and join the scientific community to start fresh, innovative research has been featured in the global academic publication *Nature*. Although the collapse of the AI bubble may lead to many job losses within the sector, it is proposed that these professionals might eventually play a key role in generating significant academic research.
◇ Ongoing Conversations About the AI Bubble
Economists consider the present surge in AI to be a historic milestone. The current funding allocated to the AI industry equals 17 times the investments made in internet firms right before the dot-com crash in the early 2000s. Nvidia, an artificial intelligence company, has a market value of 4.6 trillion dollars, surpassing the economic scale of all nations except the United States, China, and Germany.
However, numerous companies feel that AI has not yet resulted in measurable profit gains. A McKinsey study revealed that around 80% of firms that have implemented AI reported that it has had “no meaningful effect on profits.” Additionally, Sam Altman, CEO of OpenAI, recently mentioned that certain areas of the industry are “quite overhyped.” This is why some experts predict that the AI sector might face a downturn soon.
◇ Investigation Persists Following the Collapse of the Bubble
On the other side, there are predictions that new scientific advancements might only start once the AI hype subsides. John Turner, an economist from Queen’s University Belfast, stated, “Even following the collapse of the dot-com bubble, the number of papers in electronic engineering, computer engineering, and science disciplines actually rose, and the growth of internet and mobile technologies did not stop.” In essence, research persisted despite industrial disruptions.
Experts think a comparable scenario might arise even if the AI bubble bursts. Brent Goldfarb, an economist from the University of Maryland, remarked, “Should the AI bubble burst, widespread job cuts are expected in AI startups or emerging firms, with only big companies such as Google, Nvidia, and OpenAI likely to endure.” Nonetheless, he noted that over time, AI expertise might shift to different areas and drive fresh technological advancements.
Professor Turner also mentioned, “Following the collapse of the British bicycle market in 1896, bicycle mechanics who had lost their jobs contributed to the development of the motorcycle, car, and aviation industries. In a similar way, after the dot-com bubble burst, the internet expanded even more rapidly.”
◇ Experts from Companies to Showcase ‘More Advantageous Technologies’
David Kirsch, a technology historian at the University of Maryland, expressed a comparable viewpoint. He stated, “If individuals who lose their positions in the industry because of an AI downturn go back to academic settings, they might develop technologies that offer significant advantages to society.” For instance, this could result in innovations such as DeepMind’s AI program ‘AlphaFold,’ which played a key role in addressing the 50-year-old challenge of predicting protein structures.
*Nature* also noted, “These kinds of initiatives have already started.” For example, this year, leading AI researchers from OpenAI, Meta, and Google departed their organizations to establish the materials science company ‘Periodic Labs.’ The firm’s motto is, “Our objective is to build an AI scientist.” It seeks to create AI systems for experiments and simulations that can be directly applied by university laboratories, national research institutions, and hospitals.
Yann LeCun, who previously served as Meta’s lead AI scientist, recently stated, “I will depart the company later this year to launch a startup dedicated to exploring a novel type of advanced AI research.” LeCun is said to have made this choice following Meta’s CEO, Mark Zuckerberg, emphasizing the need for “AI superintelligence.” He is now concentrating on developing “world models” that offer advantages to science, robotics, and society, rather than focusing on commercial large language models (LLMs). The aim is to build an AI brain capable of thinking and acting like a human, assisting in long-term decision-making.
