AI has gone fully hyperbolic. Not the stocks or investment profits – though they surely leapt in recent years. The chatter, hype, extreme claims. Headlines herald an impending “tsunami,” “adapt-or-die” moments and looming “apocalypses” for jobs, industries … even humanity.
Reality check: understanding AI’s investment implications means muting hype and fathoming the Four “n’s”: nuance, necessity, newness and nuisance.
First, I know not just how AI’s evolution plays out. No one does. Thinking otherwise is hubris. That hasn’t stopped pundits and “experts” from offering myriad far-flung (often dystopian) theories since I last addressed AI in July 2023. They foretell fast, overwhelming, usually negative change.
Most repeat a centuries-old error: thinking innovation will only destroy and not create concurrently. Robots have been “coming for workers’ jobs” for more than a century.
In 1930, a New York Times editorial bemoaned the “thousands of columns of print about the Age of the Robot that is so rapidly displacing the Age of Man”. In a 1925 Times column, writer M.B. Levick lamented “the flivver and its little cousins” supplanting the “faithful donkey”. A century later man endures. Donkeys, too. Both with lives vastly improved.
Ignore hyperbole – consider instead nuance. Techno-optimists and those engaged in AI usually overrate the speed at which big changes come. It is an echo chamber. A prime example: the internet – and how we access it. It brought enormous change but not all at once. The early to mid-1990s featured wonky dial-up access on boxy desktop computers. That gave way to wireless internet and affordable laptops, followed by smartphones and the explosion of social media.
More transformative use followed – ubiquitous video conferencing, ride-sharing, mobile payments, delivery services and more. The evolution took decades. It continues.
Industry impact will be nuanced, too. Some will change greatly. Some won’t. Can AI improve cheesecake? Or duct tape? Can it frame a home? It may help streamline logistics for transporting and storing food and duct tape. But beyond that?
People do not pay for knowledge-based services solely for the accurate expertise. They also want responsibility. Tax preparation stocks have been crushed on fears AI replaces them. Will people ditch those services for AI-rendered bot prep? Some will. But good luck getting an AI creator to pay damages if their “intelligence” turns out to be wrong. Many taxpayers want a duty-bound professional between them and the government.
More nuance: big technological changes rarely mean either/or scenarios. Smartphone-based food delivery services exploded while grocery and restaurant sales kept growing. Big-box shops changed retailing. Then online retailers did. Yet many small shops still thrive. Look at Europe. There is room for all! No either/or. Sometimes big players help the small prosper.
Much-publicised capacity constraints are brakes on AI’s adoption. Each iteration needs more computational power. Power, data-centre land and chip shortages all limit the pace of introduction. See Nvidia on the chip-shortage dilemma. Then consider potential laws or government interventions – always innovatively helpful (not).
That doesn’t means AI will not eventually overhaul our world. Its efficiencies surely make it a necessity. Example: legal work remains way too costly from using brains to do brainless things. I doubt this means fewer first or second-chair trial lawyers. Few beyond new law school graduates with mountainous student loans will mourn the low end of the lawyer pool shrinking. It may free them to do other, more productive things.
Transport and logistics firms should benefit. Eventually automated vehicles should abound, solving longstanding driver shortages.
AI will augment financial services. But replace them? How many people nearing retirement will entrust their life savings to a liability-free bot?
Health care needs AI efficiencies. Many demographic doom-mongers fret about an ageing population overwhelming services. AI is not a healthcare job killer, but a solution.
Few of those uses are truly new. No great company has changed the world through superior efficiency. All ascended by executing never-yet-done things. Great firms see a problem, develop previously unfathomed solutions, create a product or service and market it, changing the world.
Can AI do that? Maybe someday, but not soon. Doom-mongers see AI doing all that and more quickly, than fret about endless doom loops of job cuts, plummeting consumer spending and AI reliance, foreseeing a human wasteland. If so, who buys what AI creates? Other algorithms? The whole scenario lacks logic.
Current fears and overhyped hopes are little more than nuisance. Crucial issues must be remedied before AI becomes the force many foresee. Trust is necessary … and lacking. Hallucinations and logic errors abound. Systems struggle to understand context.
They aren’t deterministic, instead giving wildly varying answers. Hype triggers incessant emotion, stoking occasional volatility. Sober thinking is critical amidst the noise.
Big change is coming but not so fast. To your fears, just say: N, N, N, N, No!
