Google has introduced AI-generated search overviews, which could hit the company’s ad revenue if users choose not to click through on search results.Francis Mascarenhas/Reuters
If there is one force that can overcome U.S. President Donald Trump’s harebrained trade wars and drive the stock market higher in months to come, it is excitement over artificial intelligence.
Over the past three months, the share prices of Alphabet Inc. GOOGL-Q, Amazon.com Inc. AMZN-Q, Meta Platforms Inc. META-Q and Microsoft Corp. MSFT-Q have surged anywhere from 18 per cent (Alphabet) to 35 per cent (Meta Platforms).
The reason? Growing evidence that each of these companies is benefiting from AI-related spending. They have all reported strong earnings in recent days and all of them have been vocal in telling investors how they are positioning themselves for an AI-powered future.
But with success comes questions. One of the most intriguing is whether AI will ultimately prove to be the reliable servant of the tech superpowers or a destructive obsession that gobbles cash and eats them from within.
The tension is clearest in the case of Google’s parent, Alphabet. For years, it has earned much of its revenue from selling ads that appear on third-party websites. This has been a highly profitable arrangement. Google has helped direct its users through a sprawling labyrinth of third-party sites. In exchange, Google has made money by populating many of the sites they visit with ads.
But now Google has introduced AI-generated overviews, perched on top of its regular search links. That changes the nature of internet search. What if many people decide they don’t have to enter the online labyrinth at all? A big chunk of Google’s ad revenue could vanish. Users might read the AI-generated answer to their question and be satisfied. They might never click any further, meaning they will never visit the third-party websites where ads appear.
A billion-dollar bet on artificial intelligence is about to hit reality
The risk to Google’s existing ad business seems substantial, yet markets are being remarkably complacent about it. “Google is destroying its own business – and still winning,” writes tech-industry analyst Rich Holmes.
In fact, to aid its creative self-destruction, Google is planning to boost its capital expenditures from an already massive US$75-billion this year to US$85-billion. Like its fellow tech giants, it is pouring money into data centres, networking gear and engineering talent to make sure it’s not left behind in the great AI arms race.
Google’s bet is that it can transition “from being a mere index to an oracle,” declares economist Brad DeLong. He says the company wants to use its AI smarts to keep people in its own domain rather than sending them elsewhere.
To be sure, this strategy would reduce the volume of visits that Google users make to third-party sites. However, the value of each outbound click to advertisers could rise, Prof. DeLong argues, because any user who chooses to go beyond Google’s quick AI-generated answer will presumably be more highly motivated and more likely to buy or engage than a casual searcher.
Maybe so. At least so far, everything seems to be working out splendidly. Alphabet reported a 14-per-cent jump in second-quarter revenue.
Still, risks remain. Microsoft or Meta could start targeting the quick-answer business with their own chatbots, as could privately owned AI upstarts such as OpenAI and Anthropic. As chatbots grow more reliable, and people become more comfortable prompting them, the demand for Google’s standard search functions may wither.
AI gains and strong earnings support Wall Street even as tariff woes linger
Google could counterattack, of course. It might shapeshift and use its expanding AI capabilities to challenge Meta’s social media empire or Microsoft’s software kingdom.
We simply don’t know. In a future world where AI chatbots and AI agents answer our questions, fulfill our wishes and become our constant companions, the boundaries between tech giants are likely to fade. Social media, search, software and retailing will all begin to blur together.
It’s difficult to know which companies will come out on top or which business models will work best in this brave new world. The uncertainty makes the current market mood a bit of a head scratcher. Investors are treating AI as an undoubted positive for all the big tech giants. Yet it’s not clear it will be, given the staggering amounts that are being thrown into AI development.
In a recent note, Morgan Stanley analysts estimated that tech companies will be pouring nearly US$3-trillion – yes, trillion with a “t” – into building new data centres between now and 2028. Companies will need to invest even more to build new generating facilities to power those fancy new data centres. And then, of course, there is the cost of developing AI itself.
The torrent of spending on AI is beginning to sound a lot like the massive fibre-optic splurges of the dot-com era of the late 1990s. That did not turn out well for many of the companies involved. Don’t be surprised if there are also some losers this time round.