The tech industry has squandered approximately 1.5 trillion dollars in the development of artificial intelligence in a single year, which is the same total for the Iraq and Afghanistan conflicts. This is causing a stir in the investment community, where economists are left pondering the question of what people are getting in return for so much investment.
World-wide AI expenditures are predicted to reach almost $1.5 trillion in 2024, Gartner predicts. For context, direct warfare expenditures in Iraq and Afghanistan cost the United States between $1.5 trillion and $1.7 trillion from 2001 to 2014.
The comparison here, while likely painful to consider, bears an uncanny resemblance in that, similarly, the current AI revolution also appears to be fueled largely by inertia.
The number of skeptics is multiplying. Businessmen want to know more about the return on investment, especially when the technology that businessmen are counting on could be made outdated within a short period of time. But still, businessmen go on releasing press releases packed with terms that describe the latest technology of AI.
The AI Paradox, Billions in Spending, Pennies in Proof
Consider the case of Larry Feinsmith, the leader of Global Tech Strategy, Innovation & Partnerships at JPMorgan Chase, with this pearl of corporate wisdom: “In the era of AI and agents, the benefits and value will be enormous, but so is the complexity.”
The quote encapsulates everything and nothing at the same time. A great amount of value will be achieved someday, allegedly, although the timing is not that complex for the purposes of this argument despite the speaker’s background as a Wharton grad whose work entails making decisions worth billions of dollars. He basically recommends that you just purchase more AI systems and hope for the best.
The economic metrics paint an interesting picture. Torsten Slok, an economist for Apollo Global Management, pointed out in October that corporate capita expenditures are effectively stalled, except for AI investments.
Economists have also argued that AI investment is currently the only thing stopping America from entering a recession. Businesses are pouring an unprecedented amount of money into technology that has dubbed exactly one operational application: better chatter bots.
Microsoft is building data center super-clusters that span continents. Hyperscale capacity spending has gone from zero to explosion. But asked to give an example, big-player execs from companies such as Dell Infrastructure or Nvidia always return to giving roughly the same response: chatbots, with “AI agents” imminent.
Why Enterprises are Pulling the Plug on AI Spending?
Salesforce.com CEO Marc Benioff says that his AI robots are already working side by side with his customers, but according to research done by Carnegie Mellon University, 70 percent of the time, AI robots fail. This scenario can be likened to your boss wanting your favorite cousin to work alongside you. The result might be positive or negative, but either way, you still lose.
It is dawning on people. Forrester says that in order for AI to show real results to prevent its own spending from being pushed back, 25% of enterprises are already delaying AI spending until 2027. Customers are seeing that they are spending a lot on AI and that their spending is not really adding to EBITDA. EBITDA stands for earnings before interest, taxes, depreciation, and amortization.
There are warnings from the likes of Goldman Sachs that “the datacenter bubble may bust with unpleasant consequences.” Senator Bernie Sanders has called for a moratorium on data center development. These are no longer voices on the fringe who are concerned about the economic viability of where the money is being spent.
Facing the Sunk Cost of a Digital Revolution
The industry has promised a revolution in AI that matches the electricity revolution. This is a radical promise that warrants a radical delivery. As of now, what we are getting in relation to AI is chat robots that sometimes make up their so-called “facts,” automated customer support that drives people nuts, combined with lots of fluff on what they will be able to do someday.
The analogy to war spending is more than just money. Wartime tends to involve leaders committing to resource levels based on optimistic projections and sunk cost fallacies, continued campaigns beyond the point at which initial motivations have become utterly irrelevant. The question hanging over all of us in the AI world today is whether we are seeing just such a thing.
As 2024 draws to a close, the tech world is faced with a reality check that is sobering indeed. The all-you-can-eat buffet of AI expenses is perhaps nearing its final call, and it’s high time that someone picked up the tab.
