Those tuning in for NBC’s Olympics coverage couldn’t have missed the Google ads trumpeting possible uses of its Gemini artificial intelligence bot. In one spot, a young fan uses the tool to generate a letter to her idol, American hurdler Sydney McLaughlin-Levrone.
Gross! What could be more impersonal than AI-generated fan mail? What parents would deprive themselves of the bonding experience of this writing exercise with their child?
The other ads — including those from other companies — aren’t much better, and they all point to what’s been a recurring theme lately: What exactly are regular people supposed to do with generative AI? If these ads are any clue, Google doesn’t have much idea.
Which is a concern, to say the least. And it is at least one small part of the psychology behind the 5% drop in parent company Alphabet Inc.’s stock last week. The company announced that capital spending was $13.2 billion in the second quarter, more than analysts had expected, as the company continued to keep up with the Joneses of Microsoft Corp., Apple Inc., Amazon.com Inc., and Meta Platforms Inc. Those competitors all report their earnings this week. With each passing quarter, investors’ patience with the costs of AI is being tested further.
Google’s slump showed signs of rebounding Monday after some analysts re-highlighted the company’s enviable position as owner of top properties like Google Search, Gmail, YouTube and more — all seen as great vectors for one day delivering monetizable AI features. From this we can infer investors are holding their ire. But they will look to the other tech earnings this week to help them sleep more soundly. They want a sign that ballooning capital expenditures are being offset with genuine new revenue streams. I don’t think they will find much encouragement.
Microsoft will report on Tuesday after the market closes. Consensus data from Bloomberg suggests analysts expect year-on-year growth of its Azure cloud business of just more than 30%. However, strong performance of Google’s cloud unit — up 29% on last year, smashing estimates — has fueled hopes Microsoft can outperform. It will need to knock it out of the park if capital spending increases as much as (or more than) anticipated — analysts think it could rise by 48% compared with last year’s level.
But strong cloud performance won’t fully answer the bigger-picture question: How much genuinely sustainable business is being created by the advent of AI? This period of intensely high demand, which has pushed hardware costs through the roof, is no way to measure AI’s longer-term potential. Unfortunately, investors have little else to work with given the table scraps of useful data being offered by the tech giants. None of the top companies have yet adequately spelled out the financial performance of AI adoption on their income statements.
In its Olympic TV spots, Microsoft is promoting Office 365’s CoPilot features, such as summarizing meetings into bullet points or generating PowerPoint presentations from notes. That looks marvelous, but alas, we don’t know how many of Microsoft’s customers are actually opting in for these more expensive, AI-enhanced plans (each user costs an additional $360 a year on top of existing Office 365 expenses).
We also don’t know how much of the company’s cloud growth has actually been driven by AI needs or is simply a return to more general IT spending after the post-pandemic slump, when companies pulled back on their cloud plans. Last time out, in Microsoft’s fiscal third-quarter earnings call, Chief Executive Officer Satya Nadella threw in some morsels. “Our AI services contributed 7 points of growth” to the Azure cloud business, he said, without providing more context.
Bloomberg Intelligence analysts suggest most of that likely comes from being the sole cloud provider for ChatGPT maker OpenAI, but we don’t know for sure. Nadella also said “1.8 million paid subscribers” were using GitHub CoPilot, the company’s AI-powered coding assistant. What this equates to in revenue isn’t clear.
Amazon is even more anecdotal. When it reports earnings on Thursday, investors will hope for more detail than the “multibillion-dollar revenue run rate” it cited for its AI-related gains within its Amazon Web Services business. Similarly, last quarter, Meta CEO Mark Zuckerberg said it was too early for “any kind of hard stats” on the use of Meta’s new AI chatbot tools within its apps. I’m not sure he can get away with the same line again.
Zuckerberg has previously projected that the company will soon offer the most widely used AI assistant globally, thanks to its strategy of stuffing Meta AI into the apps that billions of people use already. But how much context we get on this will be telling: Some users of the apps have complained they’ve invoked the assistant accidentally when trying to use basic search. I have yet to see how Meta AI’s existence is of any significant benefit to Meta or its users. Perhaps the more important number to look for is the use of AI in powering content recommendations within Facebook’s News Feed. In the second quarter, it was 30%, Zuckerberg said. This is good for ads, though precisely how good is still a mystery.
The only one of the big tech companies that will get a relatively free pass this time will be Apple, which has yet to roll out its AI strategy to its billions of devices. CEO Tim Cook can credibly tell investors there is little to share. But with the coming Apple Intelligence launch expected to kick off what analysts hope will be a super-cycle of iPhone upgrades, the company will be less able than its rivals to hide behind piecemeal anecdotal data when it comes to detailing the true contribution of AI on its business.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out some of our webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
Read more articles by Dave Lee
