#180 Beyond the Hype: Will AI Follow the Path of Railroads, the Internet, and the Printing Press?
Fresh & Hot curated AI happenings in one snack. Never miss a byte 🍔
This snack byte will take approx 5 minutes to consume.
Beyond the Hype: Will AI Follow the Path of Railroads, the Internet, and the Printing Press?
Silicon Valley’s tech sector is facing growing skepticism, particularly around AI) and its ability to deliver on investor expectations.
Since reaching its peak last month, share prices of leading AI companies have dropped by 10%, raising questions about the long-term viability of AI technologies like Large Language Models (LLMs), which power systems such as ChatGPT.
Despite tech giants investing billions into AI infrastructure, the Census Bureau reports that only 5.1% of U.S. companies currently use AI in production, a slight drop from 5.4% earlier this year, signaling that widespread adoption is far from a reality.
Many in the tech community, however, cite the "hype cycle" as an explanation for the current pessimism.
First coined by research firm Gartner, the hype cycle suggests that all emerging technologies go through phases of extreme optimism and over-investment, followed by a “trough of disillusionment,” before eventually gaining mainstream adoption.
Let’s look at the past to understand if this framework is a reliable guide for AI’s future? After all, sometimes understanding the past can give us a glimpse how the future may look like.
The Slow Evolution of Past Technologies
The hype cycle has been validated by numerous past technologies that took decades, sometimes centuries, to reach their full potential.
Consider the printing press, invented by Johannes Gutenberg in the 15th century. While revolutionary, it took almost 100 years for printed materials to become widespread across Europe.
Early adoption was slow, as literacy rates were low, and it wasn’t until the 17th century that mass media, such as books and newspapers, began to take hold and influence societies on a larger scale.
Similarly, railroads experienced an extended cycle of boom and bust. In 19th-century Britain, railway stocks soared as investors speculated on the future returns of these new networks.
Notable figures like Charles Darwin and John Stuart Mill even invested in the railway industry, contributing to a stock market bubble. After the inevitable crash, the capital raised during this period allowed companies to build the infrastructure that connected Britain’s cities, transforming the nation’s economy over time.
It took decades for railways to truly realize their potential in terms of economic impact.
The internet followed a similar trajectory. In the 1990s, there was extreme euphoria surrounding the potential of the internet, with predictions that within a few years, online shopping would dominate consumer markets.
However, the dot-com bubble burst in 2000, leading to the collapse of 135 major companies, including notable failures like Pets.com and Garden.com. Despite this setback, the infrastructure investments made during the bubble—particularly in fiber-optic cables—enabled the development of today’s high-speed internet.
This crash and recovery played out over more than two decades, but the long-term benefits of the internet are now undeniable.
Even newspapers, which seem ubiquitous today, took several decades to gain a foothold. Early printing presses and rudimentary communication networks did not lead to widespread daily news immediately.
It wasn’t until the late 19th century, fueled by innovations in printing and faster distribution via railroads, that newspapers became a major source of information for the general public.
The AI Hype Cycle: Lessons From the Past
While AI has yet to face a bust on the scale of the railway or dot-com eras, some argue that the current dip in investor confidence is an indicator that AI will follow a similar path.
, an economics commentator, notes that AI’s future could mirror past technologies: a period of heavy investment, followed by a bust, and ultimately a revival as businesses learn how to harness AI for real-world applications.However, AI is no stranger to cyclical hype.
Previous iterations of AI, such as in the 1960s, saw immense excitement surrounding early models like ELIZA, an early chatbot.
But the field experienced subsequent AI winters in the 1970s and 1990s when excitement faded, research slowed, and funding dried up. Even as recently as 2020, research interest in AI was declining before the introduction of generative AI brought it back into the limelight.
Despite these cycles, not all technologies follow the same path. Cloud computing, for example, rose from obscurity to widespread adoption in a steady, upward trend with no significant crashes.
Solar power and social media have similarly defied the traditional hype cycle by growing continuously without major disruptions. On the other hand, some hyped technologies, such as Web3 and 3D printing, have not yet rebounded after their initial surge of excitement.
Data from Gartner and other sources suggest that the hype cycle is not as common as some may believe. Research by Michael Mullany indicates that only about 20% of breakthrough technologies fully experience the cycle of hype, disillusionment, and recovery.
Among the technologies that fall into the “trough of disillusionment,” 60% never rise again. This implies that while some technologies may eventually dominate, many others fail to deliver on their initial promise.
While the hype cycle may offer some insight into AI’s future, it remains an imperfect guide.
For some technologies, the reality is closer to "easy come, easy go"—and it remains to be seen whether AI will join the ranks of world-changing innovations like the printing press, railroads, or the internet, or fade like other tech fads.