Investing in AI: Lessons from the Evolution of Autonomous Driving

By Alexandria Investment Research on October 14, 2025

October 2025

The transformational potential of Artificial intelligence (AI) is fueling an investment cycle reminiscent of the railroad boom in the mid-19th century.  Trillions of dollars have already been invested or committed for the development of AI infrastructure over the next several years⁽¹⁾.  As AI-related companies significantly outperform the broader market, pundits and investors are debating whether this is an historic opportunity, a bubble, or somewhere in between.  A good starting point for answering this question may be examining the trajectory of another recent technological phenomenon, autonomous vehicles (AVs).

The Gartner hype cycle (figure 1) visualizes the adoption and maturity of a new technology by mapping the technology’s hype against its potential commercial use.  Personal computers, sequencing the human genome, and the Internet were all greeted by a burst of excitement as they captured the public’s imagination, which in turn fueled inflated investor optimism. Inevitably, those initial expectations proved unachievable over such a compressed timeframe, giving way to a period of disillusionment and a quieter phase of steady, less publicized progress toward commercialization.

Breakthroughs in the early 2000s reframed autonomous driving from a Sci-Fi fantasy to an attainable reality. Google founded its Self-Driving Car Project (later renamed to Waymo) in 2009 during a time when the promise of fully autonomous driving began to enter the public consciousness. As the hype began to build, a rush of investment from technology companies, traditional auto manufacturers, and venture capital (VC) firms poured money into dozens of new companies. Prominent figures fed the hype, including Tesla CEO Elon Musk’s recurring assertion that full autonomy would be achieved within one year (the initial prediction was made in 2014).  By 2018, Morgan Stanley analysts estimated Waymo’s value at $175 billion, two years before it carried its first paying customers. Between 2018 and 2019, VCs invested more than $15 billion into AV technology companies⁽²⁾.

As expectations rose to unachievable levels delays mounted and aggressive predictions failed to materialize – souring investor and public perception. Amid the post-pandemic technology sell-off funding dried up and many AV companies were condemned to bankruptcy, while others were forced to sell themselves for a fraction of their previous peak valuations. In fact, those same Morgan Stanley analysts dropped their valuation for Waymo by 40% in September of 2019.

Despite waning investor enthusiasm, AV technologies continued to advance. Waymo began offering paid, fully autonomous ride-hailing services in Phoenix in late 2020.  The company expanded to five more cities in the ensuing years and currently has plans to launch services in several more.  Waymo’s progress as measured in miles driven is notably accelerating. In January 2023, two years after launching, Waymo reached one million cumulative autonomous miles driven. By December 2024, it hit 50 million.  As of June, Waymo exceeded 96 million miles ⁽³⁾ and estimates of unit economics illuminate a clear path to profitability as it continues to scale operations.

The safety data for Waymo also looks very promising compared to human drivers. Waymo has reported dramatic declines in serious crashes and related injuries (figure 2) ⁽⁴⁾.  These statistics may even understate the full benefits. An analysis by The Substack Understanding AI ⁽⁵⁾ of crash reports compiled by the National Highway Traffic Safety Administration indicates that in a large majority of crashes involving a Waymo vehicle, the other driver was at fault (often hitting a stopped Waymo). Further, Waymo’s data shows that the company’s technology is also reducing the severity of crashes, turning potentially high-speed collisions into minor fender benders.

The potential cost savings from accident avoidance represent a massive investment opportunity. Vehicle crashes in the U.S. result in approximately 40,000 deaths and an estimated $340 billion in economic costs annually. When factoring in quality-of-life effects, the annual costs jump to over $1.4 trillion ⁽⁶⁾.

While Waymo’s cumulative miles driven are still a small fraction of the 3.2 trillion miles ⁽⁷⁾ Americans drive each year, the data thus far is pointing toward meaningful benefits. Based on historical estimates of 1.26 deaths per 100 million miles driven ⁽⁸⁾, if every driver in the U.S. drove like a Waymo annual fatalities could be reduced by more than 30,000. Even with partial adoption the impact would be significant; a 25% adoption rate of Waymo-like AVs could potentially save around 10,000 lives annually⁽⁹⁾ (figure 3).  While 25% adoption might seem ambitious, Waymo is already operating at a rate that exceeds 100 million miles per year, up over 70% in just six months.

Goldman Sachs estimates the U.S. robotaxi market will grow 90% per year to $7 billion by 2030, and some projections point to a $25-$50 billion AV market by 2035. Estimates of the global market are five times as large. This is all before factoring in any potential insurance and maintenance savings.

Just as autonomous driving has shifted from science fiction to a safe real-world service, AI will follow its own path toward lasting value.  It’s difficult to pinpoint exactly where we are in the hype cycle. However, the continued scaling of AI capabilities, combined with many Tech CEOs’ existential view of the technology, suggests the boom could last much longer than investors expect.  At the same time, every major investment boom in history has experienced a period in which it is significantly overestimated. Ultimately, sentiment will swing, winners and losers will emerge, and society is likely to reap the benefits of progress.  For investors, the lesson is simple: take a long-term view, make decisions rooted in data rather than emotion, and maintain a diversified portfolio.

Footnotes

(1) ACAP team analysis of public company capital expenditures, company commentary and public backlog as of September 30, 2025.

(2) ACAP estimates of announced deals, Brookings Institute, & Crunchbase as of September 30, 2025

(3) Waymo Safety Impact Report: https://waymo.com/safety/impact/    October 5, 2025 (Data Through June 2025)

(4 Waymo Safety Impact Report: https://waymo.com/safety/impact/    October 5, 2025 (Data Through June 2025)

(5) https://www.understandingai.org/p/very-few-of-waymos-most-serious-crashes     September 17, 2025

(6)NHTSA. The Economic and Societal Impact of Motor Vehicle Crashes, 2019 (Revised) February 2023

(7) Federal highway Administration Traffic Volume Trends July 2025

(8) NHTSA data for the year 2023, as of July 2025

(9) Analysis by Dr. John Slotkin X.com September 17, 2025