Overview
According to some, AI represents a massive shift in the way business is conducted. Just as the internet created massive shifts in business models, AI presents similar risks to seemingly invincible businesses. The “trick” for risk managers and institutional investors is to anticipate such casualties and hopefully sidestep them.
Internet Casualties
Yellow Pages (via Google search), Blockbuster (via Netflix), and Kodak (via cellphone cameras) are all casualties of the technology shifts broadly described as the internet. In fact, given the struggles of many retailers (Nieman Marcus, JCPenney, Circuit City, Rite Aid, Macy’s), and the continued rise of Amazon, one could view the “creative destruction” as continuing.
AI Capabilities
To determine AI casualties, perhaps it is worthwhile determining what activities AI is particularly adept at. It is already obvious that AI is good, and soon will be excellent at gathering, analyzing, and making recommendations based on a wide dataset. However, although it is not normally viewed this way, perhaps the more powerful applications will come when AI is paired with other capabilities, such as real-world robotics or open-ended tasks accomplished with AI “agents.” However, large language models, which power AI, struggle at some complex reasoning tasks according to a recent study from Apple.
The New World
Assuming AI, robotics, advanced communications, power storage, thorium power, and a variety of other capabilities are combined, it is only a matter of time before a multitude of activities are replaced. Regarding robotics, we have addressed this in previous installments, but if a humanoid robot costs $10K, lasts 5 years, and can work 20 hours per day, the annual cost, including servicing, might only be $4K per annum, which translates into a cost of merely $3 per hour.
Capturing Value
Although it may have been evident to an observer in the early twentieth century that the automobile and airplane would revolutionize commerce, warfare, energy use, tourism, culture, city planning, etc., it was not evident which firms would benefit most from this revolution.
In 1908, there were 253 active American automobile manufacturers.¹ There was a similar plethora of American airplane manufacturers.
Today there are only four American car manufacturers (if you count Stellantis, a Dutch conglomerate that acquired Chrysler), and only one major airplane manufacturer, Boeing (you could argue there are up to six if you include Lockheed Martin, Northrop Grumman, General Dynamics, and Textron).
It’s clear that there has been significant consolidation; most early manufacturers failed. In fact, only two American car manufacturers have avoided bankruptcy or dissolution: Ford and Tesla.
Similarly, operators such as airlines have clearly not reaped most of the benefit of new transportation technology (major airlines have had a poor bankruptcy record). Rather, investors and consumers across many industries benefitted most from productivity increases due to the transportation revolution and this is likely to repeat for the AI revolution.
Thus, it would behoove investors to identify those markets that are (1) likely to have new entrants due to AI (2) likely to see productivity improvements but see few new entrants due to other entry barriers.
The Threatened
Now for the hard part. Our view is that if AI lives up to only a portion of its promise, it will massively change the business landscape. Below is a list of the potentially vulnerable industries.
Mainframe Computer Systems – with the traditional mainframe firms and the infrastructure supporting such systems are likely to falter.
Auto Manufacturers – the rapid development of electric vehicles, improvements in battery effectiveness, and over-supply of EV production capacity, is likely to place significant pressure on most manufacturers and the related infrastructure.
Trucking Firms – it is only a matter of time before self-driving vehicles replace traditional truckers, with massive, related implications.
Low-end Manufacturers – as robotics become more capable, traditional manufacturing jobs, particularly the lower end ones, are likely to be replaced.
General Service Firms – the trend of replacing manual tasks with technology has been an ongoing theme and is likely to accelerate. There was a time when meters were read by hand to provide an accurate read of water, gas, and electricity usage. However, for most users, those functions have been replaced by technology. We expect that trend to continue and accelerate.
Medical Service Firms – perhaps it is merely a matter of time before some of the traditional proceeds are increasingly done via machine. Laparoscopic surgery is now the norm in many areas and is likely to increase.
Architectural and Engineering – CAD/CAM (Computer-Aided Design/Computer-Aided Manufacturing) has and will continue to make inroads, thereby increasing productivity and effectiveness.
Construction Industry – much of construction is repetitive and, over time, can be conducted more effectively with advanced robotics.
Conclusion
The pace of change is accelerating, with broad implications for risk managers and institutional investors.
Sources
[1] https://www.history.com/articles/automobiles