Silicon Valley’s tech firms are having a tough few weeks. An increasing number of investors are concerned that artificial intelligence may not provide the massive gains they expect. Since peaking last month, the share prices of Western companies spearheading the AI revolution have fallen by 10%. A rising number of observers are questioning the limitations of massive language models, which underpin systems like ChatGPT. Big tech companies have invested tens of billions of dollars on AI models, with even bigger pledges for future investments. However, according to the most recent Census Bureau data, only 5.1% of American enterprises utilize AI to generate goods and services, down from a peak of 5.4% earlier this year.
When you gently bring up these difficulties with a technologist, they will look at you with disappointment and pity. Haven’t you heard of the “hype cycle”? Gartner, a research organization, popularized this term, which is well-known in the Valley. According to the concept, hot new technologies enter the “trough of disillusionment” after an initial phase of unreasonable optimism and overinvestment, when mood sours. Everyone begins to fear that technological adoption is moving too slowly, and earnings are difficult to come by. However, when darkness follows day, the technology returns. The investment that accompanied the wave of excitement allows for a massive expansion of infrastructure, moving the technology closer to broad usage.
It surely helps to explain the evolution of some older technology. Trains are a classic example. Railway mania seized nineteenth-century Britain. Everyone from Charles Darwin to John Stuart Mill invested in railway stocks in the hope of making a profit, resulting in a stock market bubble. A crash followed. The railway firms then used the funds obtained during the frenzy to build the track, uniting Britain from top to bottom and revolutionizing the economy. The hype cycle was completed. More recently, the internet has undergone a similar transformation. In the 1990s, there was enthusiasm over technology, with futurologists claiming that everyone would be buying online within a few years. In 2000, the market fell, causing the failure of 135 major dotcom enterprises.
