By Steven Lewis
Undoubtedly, the long-term development of artificial intelligence (AI) and recent release of several new tools to the public have generated a high degree of excitement in the news and in financial markets. On the positive side, AI appears to offer strong potential for productivity and timesaving enhancements for personal and business uses, for which current tools may have only scratched the surface. There are a variety of research benefits as well. In industry, tools that sift through immense amounts of data, math formulas, or scientific relationships offer obvious time savings versus manual or trial-and-error work.
AI has also raised fears, as rapid technological changes have tended to do historically. On the negative side, concerns that it will take over the world, exterminate humanity (after all, aren’t we the cause of all the world’s problems?), or simply make most or all jobs redundant over the next few years seem overdone. However, there are likely real consequences that could be somewhat offsetting and may need to be ironed out from a policy and practical standpoint. For instance, having AI author student term papers is just one problem the education system will have to contend with. With so much information immediately at one’s fingertips, critical thinking and creativity run the risk of becoming scarcer, and thus, more valuable than ever. A recent comic captured this conundrum well—with an employee asking AI to craft an especially long and detailed e-mail which then proceeds to go unread by the receiver due to its ridiculous size and overly bland content. (Although no doubt this already happens without the help of AI.)
Despite the recent hype, various forms of AI have been around for quite a while, although the impact has been more subtle. For instance, predictive text on a smartphone, online map directions, or voice commands all use simpler forms. More recent tools have sharply expanded the universe from which information is pulled; a database of ‘everything ever published’ is almost too large to get one’s head around. For instance, instead of asking one’s phone simple questions and being provided weather forecasts, sports scores, or website links, newer tools can craft a more accurate bullet-point summary, or a longer essay (or poem!) to answer a question. These AI tools are built on a probabilistic platform, looking at the wisdom of the crowd, so to speak, using math to craft the best ‘average’ response. So, these aren’t 100% accurate at this stage, and answers can appear generic (since they are using averages after all, as opposed to extremes), but could be effective for day-to-day purposes.
Will AI ruin the labor market? Over the centuries, when new technology has been introduced that could threaten jobs, worker responses have run the gamut from amazement to mild worry, to outright revolt. Depending on the case, the concern has often been valid, particularly as machines were specifically designed to improve output with the same or less input, not to mention bypass the downsides of human labor (sick days, fickle morale, etc.). While this came to the point of workers destroying new inventions multiple times, progress couldn’t be unwound once it was unleashed. Eventually, as opposed to human labor as a whole evaporating, workers shifted to different types of jobs—in fact, overall jobs tended to keep expanding along with the growing global population. Many times, these were jobs that hadn’t even been imagined prior to the technological changes. However, there has been some transition period of pain as societal adjustment and/or re-training tends to be required, something that’s been brought up by U.S. policymakers in recent years but hasn’t been overly effective in implementation. Sociologists have attributed modern job-skill deficiencies to being a key cause of weaker labor force participation, even leading to drug use and crime among a portion of the working-age population, particularly as jobs were lost due to automation and/or offshoring over the last few decades in certain regions of the country. Well-known in the emerging world, large groups of the unemployed and frustrated have been catalysts for populist movements on many occasions. In milder forms, pushback against change can morph into anti-business policies, as too much of a focus on capital and not enough on labor and working conditions eventually turns negative. A stretch of the latter 19th and early 20th centuries in the U.S. saw a pushback against the ‘robber barons,’ leading to progressive labor reforms, unionization, anti-trust legislation, and generally more government oversight over business in general.
Some skeptics of AI translating into an immediate cascade of profits ask: if it becomes available to everyone, is it as valuable? This is in keeping with any change in technology. If the same tool is available to all, while useful, it doesn’t provide a relative competitive advantage to one party over another. Economics professor Robert Gordon at Northwestern University, who has described himself as ‘the prophet of pessimism,’ has been famously skeptical of the benefits modern technological change has brought. He’s pointed out that productivity gains at a marginal level today are far less impactful on society than those created during what he calls the ‘special century’ (roughly 1870-1970). Those included the exponential human experience improvements of indoor plumbing, electricity, telegraph/telephone communications, the internal combustion engine, and air travel, for a few examples. Amusingly, his thesis advisor, Nobel Prize winner Robert Solow, noted in the late 1980s that the computer age ‘could be seen everywhere except in the productivity statistics’ (although that seemed to improve in the following decade).
Nevertheless, according to some economists, AI has the potential of increasing economic growth rates, largely due to anticipated gains in productivity. While this is difficult to measure at this stage, some estimates point to a possible 1-2% per year rise in productivity growth over the next ten years, with perhaps up to a percent per year improvement in GDP growth. The improved productivity would offset mixed expectations for labor force growth in coming years, largely determined by demographics driven by family formation and immigration policies. At the same time, many expected productivity enhancements from new technologies, such as smartphones and social media, haven’t been fully realized in business output (although they have enhanced social connectivity and entertainment value).
Another fear of widespread AI adoption is its possible misuse, as in the manufacture of misinformation, making it difficult to judge the accuracy or source of output. In fact, some political strategists fear that an important threat of AI is not ‘AI vs. humanity,’ but rather potential effects of using misinformation to pit people and groups against each other through misinformation campaigns. Clearly, political polarization is already high, and such techniques have been used historically in lower tech ways to generate instability. Determining what content is ‘real’ and what isn’t will be a rising challenge. From a geopolitical standpoint, both the U.S. and China are undoubtedly seeking their own advantages in AI for commercial and defense purposes. Whether this continues down a single path or further diverges into two separate technology platforms (as with 5G wireless communications) remains to be determined.
For investment markets directly, new technology matters in terms of how it can generate revenue and ultimately translate into profits. The implications of newer AI are less clear in this regard. The popularity around technology stocks recently is at least partially due to excitement around AI, as the mega-cap firms are those with the deepest pockets to explore such technology. Some parallels about market exuberance have been drawn to the excitement over the rollout of the internet in the late 1990s. There is certainly potential for a variety of firms in this environment, while it could be the ‘picks and shovels’ equipment providers in hardware reaping many of the early benefits (explaining much of the positive sentiment around Nvidia). As the technology develops and depending on whether ‘tiers’ are formed in terms of AI services, it’s possible software service companies could find niches here as well, potentially as subscription products. Of course, this carries over to a host of other sectors that could see immense benefits from AI, including military applications/defense systems, law, logistics/transportation, pharmaceuticals, medical devices, and even financial services, among many others. All that said, AI remains a long-term and evolving project with much more unknown than known at this point.
(Author’s Note: The preceding was not written by AI.)
The information above has been obtained from sources considered reliable, but no representation is made as to its completeness, accuracy, or timeliness. All information and opinions expressed are subject to change without notice. The information provided in this report is not intended to be, and should not be construed as, investment, legal or tax advice; and does not constitute an offer, or a solicitation of any offer, to buy or sell any security, investment, or other product. Lewis and Associates Capital Advisors, LLC is a registered investment advisor.
