Silicon Valley’s tech brethren are having a tough few weeks. A growing number of investors fear that artificial intelligence (AI) will not deliver the huge profits they seek. Since peaking last month, the share prices of the Western companies driving the AI revolution have fallen by 15%. A growing number of observers are now questioning the limitations of large language models, which power services such as ChatGPT. Big tech companies have spent tens of billions of dollars on AI models, with even more extravagant promises of future payouts. Yet according to the latest Census Bureau data, only 4.8% of US companies use AI to produce goods and services, down from a peak of 5.4% earlier this year. Roughly the same proportion intend to do so in the coming year.
If you put these questions gently to a technologist, they will look at you with a mix of disappointment and pity. Haven’t you heard of the “hype cycle”? It’s a term popularized by Gartner, a research firm, and is common knowledge in Silicon Valley. After an initial period of irrational euphoria and overinvestment, hot new technologies enter the “trough of disillusionment,” the argument goes, where sentiment deteriorates. Everyone starts to worry that adoption of the technology is happening too slowly, while profits are hard to come by. However, as night follows day, the technology re-emerges. The investment that had accompanied the wave of euphoria enables massive infrastructure build-out, which in turn pushes the technology toward widespread adoption. Is the hype cycle a useful guide to the future of AI in the world?
It is certainly useful in explaining the evolution of some older technologies. Trains are a classic example. Railway fever gripped 19th-century Britain. Hoping for good returns, everyone from Charles Darwin to John Stuart Mill poured money into railway stocks, creating a stock market bubble. Then there was a crash. Then the railway companies, using the capital they had raised during the mania, built the track, connecting Britain from top to bottom and transforming the economy. The hype cycle was complete. More recently, the internet followed a similar evolution. There was euphoria around the technology in the 1990s, with futurologists predicting that within a couple of years everyone would do all their shopping online. In 2000, the market crashed, leading to the bankruptcy of 135 major dotcom companies, from garden.com to pets.com. The most important result, however, was that by then telecommunications companies had invested billions in fiber optic cables, which would become the infrastructure of today’s Internet.
While AI has not experienced a crisis on the scale of the railroads or dotcoms, the current anxiety is, some say, evidence of its impending global domination. “The future of AI is going to be like any other technology: there will be a huge, expensive infrastructure buildout, followed by a huge crisis when people realize they don’t really know how to use AI productively, followed by a slow recovery until they figure it out,” says Noah Smith, an economic analyst.
Is this correct? Maybe not. For a start, versions of AI itself have for decades experienced periods of hype and despair, with academic commitment and investment waxing and waning but never reaching the final stage of the hype cycle. There was a lot of enthusiasm for AI in the 1960s, including Eliza, one of the first chatbots. This was followed by AI winters in the 1970s and 1990s. By 2020, research interest in AI was in decline, before rising again when generative AI came along.
It’s also easy to think of many other influential technologies that have outperformed the cycle of expectations. Cloud computing went from zero to hero in a fairly straight line, with neither euphoria nor failure. Solar energy Social media seems to be behaving the same way too. Some companies, like Myspace, fell by the wayside and there were initial doubts about whether it would make money, but consumer adoption increased monotonously. On the other hand, there are many technologies that caused an atmosphere of euphoria and panic, but have not made a significant comeback (or at least not yet). Remember Web3? For a while, people speculated that everyone would have a 3D printer at home. Carbon nanotubes were also a big hit.
Anecdotes only go so far. Unfortunately, it’s not easy to test whether a hype cycle is an empirical regularity. “Because this is data based on vibrations, it’s hard to say much about it definitively,” says Ethan Mollick of the University of Pennsylvania. But we’ve tried to say something definitive, expanding on work done by Michael Mullany, an investor, in 2016. The Economist collected data from Gartner, which for decades has placed dozens of trendy technologies where it thinks they belong in the hype cycle. We then supplemented it with our own numerical analysis.
Over the hill
In short, we found that this cycle is a rarity. If you follow the trail of disruptive technologies over time, only a small proportion (perhaps a fifth) go from innovation to enthusiasm to dismay to widespread adoption. Many technologies become widely used without having to go through such a roller coaster. Others go from boom to bust but do not return to normal. We estimate that of all the forms of technology that fall into the abyss of disillusionment, six out of ten do not re-emerge. Our conclusions are similar to Mullany’s: “An alarming number of technological trends are ephemeral.”
AI could still revolutionize the world. One of the big tech companies could make a breakthrough. Businesses could realize the benefits the technology offers them. But for now the challenge for big tech is to prove that AI has something to offer the real economy. There is no guarantee of success. If technology history is to be looked to for a sense of AI’s future, the hype cycle is an imperfect guide. A better one is “easy come, easy go.”
© 2024, The Economist Newspaper Limited. All rights reserved. From The Economist, published under license. The original content can be found at www.economist.com
Disclaimer:
The information contained in this post is for general information purposes only. We make no representations or warranties of any kind, express or implied, about the completeness, accuracy, reliability, suitability or availability with respect to the website or the information, products, services, or related graphics contained on the post for any purpose.
We respect the intellectual property rights of content creators. If you are the owner of any material featured on our website and have concerns about its use, please contact us. We are committed to addressing any copyright issues promptly and will remove any material within 2 days of receiving a request from the rightful owner.