Why Every Major Tech Company Is Betting Its Future on a Technology Most People Cannot Explain

When you ask someone on the street to describe artificial intelligence, most of them will pause, look a little uncomfortable, and provide an approximation of what it actually does—not the concept, but the mechanism, the actual process by which a machine learns to recognize a pattern or generate an output. A guess disguised as a response.

That’s alright. Flipping a light switch doesn’t require most people to understand how electricity is generated. However, the businesses that are currently investing tens of billions of dollars in AI infrastructure are not speculating. The unsettling thing is that even many of them are unable to predict with certainty when the returns will materialize, making this the biggest corporate wager in contemporary economic history.

Amazon is developing data centers at a rate that puts pressure on the construction industry’s logistics, including sourcing materials, employing specialized contractors, and organizing power supplies for buildings that will use electricity on a scale typically associated with small cities. Five years ago, Microsoft’s commitment to invest in AI infrastructure would have seemed unthinkable.

The same choice is being made by Google, Apple, Meta, and a long list of businesses that aren’t usually associated with Silicon Valley, such as regional banks, manufacturers, hospital networks, and agricultural conglomerates. They are integrating technology so thoroughly into their business processes that it is becoming increasingly challenging to distinguish between what the company does and what its software does.

Tony Uphoff, a five-time CEO who currently provides technology strategy advice to boards and senior leadership teams, has been observing this convergence for years. When he speaks with C-suite executives in a variety of industries, including manufacturing, healthcare, fashion, entertainment, and government, he frequently hears the same theme emerging from discussions that previously had nothing in common. Instead of determining whether they must move at all, they are all attempting to determine how quickly they must move. It seems that the answer to that question has already been found. Depending on your point of view, the discussion has completely shifted to pace and execution, which is either comforting or concerning.

It is difficult to dispute the historical pattern that makes this moment seem inevitable. Businesses that have viewed technology as optional have consistently had a poor track record. Not only did Kodak fail to embrace digital photography, but it also developed some of the first digital camera technology before deciding not to take it seriously in order to preserve its film business, which it thought would last.

Convinced that physical rental stores were sufficiently ingrained in consumer behavior to endure, Blockbuster passed on several opportunities to purchase Netflix at prices that would seem absurd today. Sears, which used to control American retail through its network of stores and catalogs in the same way that Amazon currently controls it through its platform, spent years thinking that scale was an adequate defense. In the same museum, all three are warning displays. The record indicates that resistance to technological change does not preserve anything; rather, it merely postpones an inevitable and typically more difficult reckoning.

The scope of the current wave sets it apart from previous technological cycles. The way businesses interacted with clients and handled orders was altered by the internet. Cloud computing altered software delivery methods and data storage locations. They were both important. However, automation, real-time data processing, and artificial intelligence are fundamentally altering what businesses do, not just how they do it. Better tools aren’t being used to accomplish the same task by a logistics company using AI-driven route optimization.

It involves performing a different task that calls for a variety of individuals with various skill sets to make various kinds of decisions. A hospital system’s current workflow isn’t being improved by using predictive algorithms to identify patients who are at risk of readmission. It is creating a completely new layer of clinical decision-making to complement the conventional one. These are not small improvements. These are long-term structural alterations.

Researchers at organizations like Berkeley’s California Management Review have been willing to be honest about the risk that lies beneath all of this excitement: most businesses won’t profit from their AI investments, at least not in the ways they currently anticipate. That conclusion is descriptive rather than defeatist. This scale of transformation is invariably uneven. Some businesses will perform well, creating real benefits in terms of cost effectiveness, speed, and customization.

Others will invest a lot of money in technology they don’t fully comprehend, implement it on top of organizational structures that aren’t designed to use it, and then question why the outcomes fall short of expectations. The factor that distinguishes these results is not the technology itself. Whether the investment is profitable depends on the ability to use it—the talent, the culture, and the willingness to reorganize around what the data actually shows.

Observing the scope of this change from the outside, it seems that the businesses making the biggest announcements today are not the ones that will most obviously look back on this time. They will be the more subdued ones, taking their time, developing internal capabilities before making external claims, and treating AI as a shift in perspective rather than a new product.

Reporting on that is more difficult. Press releases and stock price reactions are not produced by it. However, it is almost certainly where the long-term benefits will eventually be found, and it is probably closer to what actual transformation looks like.