For the last few years, the word “AI” has been inescapable. Ever since the successful launch of ChatGPT by OpenAI in late 2022, it seems like every tech company is advertising its use of AI or creating its own AI models. The way that we consume anything has changed due to the introduction of AI-generated images, videos, and music. I find myself being way more skeptical of everything I see on the internet now. However, the spectacle of AI and its fast dominance over the tech industry has hidden the truth: it’s extremely expensive and unsustainable. Thus, discussion has begun about AI being a “bubble,” an economic term for a period of time where businesses are massively overvalued and the economy appears to be booming, before the tone shifts back to reality and it all collapses. I think that we are headed here, and that a massive crash might come sooner than expected.
AI, which stands for artificial intelligence, has been around for decades. The term generally refers to any agent (something capable of making actions by itself) that can react to new situations in a safe manner. However, the “AI” that the modern tech industry talks about is “generative AI,” software trained on an extremely large set of data that builds answers based on complex prediction algorithms. For Silicon Valley and the tech industry in general, AI is one multibillion dollar bet. Billions of dollars have been invested in AI model training, data centers, and research just this year. The end goal is artificial general intelligence, or AGI for short. AGI will be able to match human intelligence, and may even be able to surpass it. It has been a major debate whether AI should be made to match or surpass human intelligence, given the possible threat to humanity it would cause. Tech companies such as Meta are willing to take the risk and try to create “super-intelligent” AGI.
And yet, as stocks of major players in the AI hype such as Nvidia (worth around $5 trillion) and OpenAI (worth around $500 billion) continue to rise and AI is promoted as the greatest innovation in history, AI is nowhere near profitable and companies are hemmoraging billions. Between data centers and model training, AI is extremely expensive and there is no clear plan to profitability in the future. That’s not mentioning how much energy AI uses up, which is a growing crisis in itself. According to an infographic from Bloomberg, the cash flow of AI companies is a circular cycle, creating the illusion of growth and profit while no money is being made from AI. Even the smaller companies are not successful. According to MIT, 95% of companies investing in AI are not making a profit.
We can also look to the history of economic bubbles to find similarities. The dot-com bubble of 2000, which was a large overvaluement of “.com” businesses in the middle of the internet craze, is a common example. Like AI, dot-com businesses were almost all unprofitable, despite being worth millions, and their massive presence and big promises made them appear more successful than they actually were. Even Sam Altman, the founder of OpenAI, says that AI is a bubble, however, he seems to believe that OpenAI will be safe from a crash, which I doubt considering how much money OpenAI loses every year.
Big tech does a really good job at making generative AI feel mindblowing: think of the extremely realistic videos from Sora 2, the generation skills of ChatGPT and other text generators, and Suno’s AI-generated pop songs. But as someone that had the novelty of AI wear off years ago from disillusionment (and upon learning how damaging AI can be to artists, the environment, and mental health), this type of AI feels unnecessary in most ways. I think the currently obsessed public will end up feeling that way too after some time. There are great applications of generative AI, such as in the medical field and accessibility, but these feel overshadowed by the “spectacle.” Which I think is all that the industry has that has avoided a crash for all this time.
I’m not an economic expert, or really know that much about market trends. However, I know that this AI hype is not sustainable and that reality will set in at some point. There’s already fear and uncertainty growing among investors and economy experts, which is a sign that perception is shifting. Reality will win in the end, especially when large amounts of money is involved.
Senior Nicholas Carpenter is the Opinions Editor. His email is ncarpent@fandm.edu.