On the AI hype machine
From the late 18th century onwards, industrial capitalism fostered a rate of scientific and technological advancement unparalleled in previous human history. Beginning with the steam engine, the cotton gin and the telegraph, incredible breakthroughs in transportation, communication, medicine, infrastructure and entertainment have driven human productivity and wealth exponentially higher. Will future AI innovations have as profound an impact on our economic life?
In his seminal book, The Rise and Fall of American Growth, economist Robert Gordon argues that the kind of rapid economic growth we consider our due, and now expect will resume as a result of breakthroughs in AI, was in fact a unique period in history that’s unlikely to be repeated. Gordon describes what he calls the Special Century, the period between 1870-1970 which liberated households from the daily grind of manual labor, darkness, isolation, drudgery and premature death. The Special Century was marked by innovations that unleashed human productivity: electricity, the combustion engine, plumbing, heating, air conditioning, refrigeration, plastics and antibiotics. Today, it’s easy to forget that in 1890 water had to be carried in and out of the house, sickness and food contamination were constant threats, infant mortality was north of 20%, getting around by horse was the pinnacle of transportation, and a decent education was available only to a minority of children. A mere forty years later, a majority of the population had migrated to cities where almost all dwellings were equipped with electricity, natural gas, telephones, clean running water and sewers. The productivity benefits of these breakthroughs echoed for decades and were reflected in impressive GDP figures.
Things changed markedly around 1970. Over the following fifty years, productivity growth, as measured by GDP output per unit of labor, declined to one-third the rate of the previous century despite the boost in productivity in the late 1990s and early 2000s when offices and households embraced the marriage of computers and new communication tools.
Productivity change in the nonfarm business sector, 1947 - 2023
In the spheres of human life that don’t involve communication, information processing, and entertainment, innovation has slowed dramatically. Construction methods haven’t changed much, travel times by air, car and rail haven’t improved, and life expectancies are no longer increasing.
Artificial intelligence, while still a far cry from the bold technological innovations imagined by mid-20th century science fiction writers, admittedly looks to have more staying power and disruptive potential than crypto, the metaverse and other recent flops from the tech behemoths. It’s clear AI is poised to disrupt certain sectors — the bloated healthcare industry for one looks to be particularly vulnerable — and it’s also likely to reduce the number of mediocre lawyers, graphic designers, travel agents, writers, actors, and the like. That said, the idea that AI is going to transform human life to a degree never seen before is a self-serving narrative emanating from the Silicon Valley incumbents who are desperate for a new growth engine and to further cultivate their monopoly power.
When people talk about AI, what they’re really talking about are old statistical techniques coupled with brute force computing, made possible by NVIDIA GPUs and massive amounts of data acquired by a small set of companies whose principal business is expropriating human experience for profit. AI executives, engineers, and researchers speak as though they're giving birth to God, but what they're actually doing is recreating human intelligence to serve commercial and military interests. The sense of inevitability around AI and the surveillance model at its foundation is becoming solidified by a growing consensus among US leaders that if we don't do it someone dangerous will.
$NVDA - trailing 1-year stock performance
The recent stock market performance of the “Magnificent Seven” — Apple, Meta, Alphabet, Microsoft, Amazon, Tesla, and NVIDIA — has highlighted the power of the AI hype machine. The group now has more than a quarter of the US market’s valuation riding on them, and where the stocks go from here is the central question for US equity managers. Now with enough capital to crowd out startups, universities, municipalities, established corporations in other industries, and less wealthy countries, these tech giants surely seem to be winning at their game.