Breaking Up Big Tech Might Have Broken the AI Revolution, Too

It was a promising project. Back in 2015, London-based Google DeepMind attempted to improve the efficiency of the UK’s National Health Service (NHS). Its initial effort was Streams, a smartphone messaging app that connected hospital blood labs directly to the relevant doctors and nurses. And it worked beautifully. A 2019 independent evaluation found that urgent cases going unnoticed fell by three-quarters, nurses spent more time with patients, and treatment costs dropped. Alas, as documented in the excellent The Infinity Machine : Demis Hassabis, DeepMind, and the Quest for Superintelligence by Sebastian Mallaby, DeepMind suffered from an original sin: It was owned by an American tech giant. That, at precisely the moment a backlash against American social media platforms—especially due to concerns about data privacy and political misinformation—was gathering force. Long story, short: NHS partners lost their nerve, and Streams was shut down. Also a result, a DeepMind retinal algorithm that could have prevented tens of thousands of cases of blindness annually went undeployed. Widespread criticism of these platforms—although it didn’t lead to federal legislation in the US—was such a critical turning point that it can be hard to recall the high level of optimism that first surrounded social media. In the first half of the 2010s, companies heavily promoted idealistic, even utopian benefits: connection, empowerment, democracy, and opportunity. All the buzzwords. Yet it turns out the most important benefit had nothing to do with any of that stuff—and it’s playing out right now. The Wall Street Journal Consider: The same internet advertising empires that frequently drew regulators’ ire in the second half of the 2010s—particularly Google’s search monopoly, Meta’s Facebook, along with the company’s Instagram and WhatsApp acquisitions—turned out to be the cash machines now helping fund the most consequential technology buildout in history. Alphabet just raised its full year 2026 capital spending guidance to as much as $190 billion, while Meta Platforms told investors the company plans to increase its capex investment to as much as $145 billion. That dough comes almost entirely from ad revenue. Break up the duopoly a decade ago, as many reformers wanted, and you would have ended up with smaller firms generating less free cash flow and facing tighter financing constraints—precisely the wrong setup for a technology wave requiring hundreds of billions in upfront infrastructure spending before meaningful returns arrive. Would their risk appetite have been the same? I don’t think that’s guaranteed.  Oh, and all that’s before considering another thing social media gave AI: billions of user-generated data points that helped train the models in the first place. And then there’s the current role played by social media in providing a pro-progress counternarrative to concerns about AI taking all the jobs, electricity, and water—which I explain at length in the essay, “How social media is saving the AI revolution .” The post Breaking Up Big Tech Might Have Broken the AI Revolution, Too appeared first on American Enterprise Institute - AEI .