The Crucial AI Economics Question: Are Customers Willing To Pay For It?

Demand is lagging supply, so the classic economics answer is not looking affirmative in the short term. Corporate customers are hesitant due to cost, potential lawsuits for intellectual property violation and the continuing issue of AI inconsistency.

Individual consumers are weary of inflated prices for basic goods and the surfeit of fees they are being hit with for everything from streaming to additional technology demands. The result is an absence of interest in additional cost for uncertain benefits in the short term. JL

Christopher Mims reports in the Wall Street Journal:

For years, the gap between cutting-edge devices and older versions has shrunk, prompting consumers and businesses to hold them longer. The average replacement time for phones has grown to three years, laptops to four, and desktops to six. The industry has a plan to reverse this trend: Persuade you to upgrade, or be left behind as it adds AI. (But) big enterprise customers are hesitant to invest heavily in  companywide subscriptions to OpenAI’s ChatGPT or Microsoft’s Copilot, held back by concerns about data sovereignty, employee pushback, embarrassment and liability when AIs make mistakes. (And)  consumers’ budgets are stretched after years of inflation and increased spending on services.

The tech industry has a new strategy for getting us to upgrade our aging devices: artificial intelligence.

Much of the industry’s revenue depends on us buying the newest thing as often as possible. But for years, the gap between cutting-edge devices and older versions has shrunk, prompting both consumers and businesses to hold on to them longer than ever. The average replacement time for phones has grown to three years, laptops to four, and desktops to more than six.

The industry has a plan to reverse this trend: Persuade you to upgrade in the near future, or be left behind as it adds AI to every aspect of your work and life.

And that’s just the beginning. Whether we upgrade or not, almost every name-brand tech company is betting that many of us will pile on new subscriptions for cloud-based AI services. That recurring revenue can be significant. For example, adding AI to Microsoft’s Office 365, in the form of the company’s Copilot offering, raises the cost by $30 a month per employee. That’s a 50% jump from even its most feature-rich version of the product.

 

This is the inevitable outcome of all of the hype and investment being poured into AI. After all, those new data centers and microchips aren’t going to pay for themselves. 

The approximately $300 billion pop in Apple’s market capitalization since its Monday announcement of “Apple Intelligence” AI services suggests investors are betting that this is the right move, at least for Apple. 

Run-ups in the value of Google and Microsoft—both with sweeping strategies for new AI-enabled devices and services—suggest similar investor logic is at work for those companies. Even consumer-electronics retailer Best Buy is getting in on the action. Shares in the company surged after it highlighted a potential revenue boost from consumers buying AI-powered laptops.

It isn’t yet clear whether this excitement and investment is going to pay off anytime soon, according to experts who focus on the real-world impacts and adoption rates of these technologies. There are plenty of reasons to be skeptical. For example, consumers’ budgets are stretched after years of inflation and increased spending on services for streaming video, music, home delivery, and more. Anyone hoping for hockey-stick like growth in device sales is likely to be disappointed, says Carolina Milanesi, president of tech analysis firm Creative Strategies.

In part, this is because new AI-powered devices are more expensive and won’t be universally available. For example, while Apple detailed a long list of new AI features for its devices, they’ll only be available on the company’s newer notebooks and desktops, the iPhone 15 Pro models, and only in English. 

Apple’s usual upgrade pattern for chips in the iPhone suggests these features will eventually come to all their newest devices, probably with the iPhone 16 models, which are likely to come out in September. Until that happens, and Apple’s AI models are rolled out globally in every major language Apple supports, they’re unlikely to move the needle on sales, says Milanesi. As for PCs, worldwide sales of desktops and notebooks, excluding China, are projected to grow only 2.6% in 2024, according to tech market-intelligence firm IDC. This is a reversal from the past two years of steep declines in PC sales, after their peak during the everyone-working-from-home era of 2021. Despite that expected growth, there is hardly any change in the length of time most people and businesses will be holding on to their PCs.

And then there are the big enterprise customers, who, despite enormous pressure to have some kind of AI strategy, are hesitant to invest heavily in services like companywide subscriptions to OpenAI’s ChatGPT or Microsoft’s Copilot tools, says Dan Faggella, founder of Emerj, a market-research company for AI.  

Wall Street Journal owner News Corp has a content-licensing partnership with OpenAI.

Adoption of AI by big enterprises will be slow, painful, and held back by concerns about data sovereignty, employee pushback, and the possibility of embarrassment and liability when public-facing AIs make mistakes, he adds. Concerns among executives about the costs of rolling out AI at medium and large companies all over the world have shot up over the past year, according to a recent survey from Lucidworks, which makes search tools for e-commerce companies. If AI-powered devices and services are adopted at a slow pace, only the biggest AI companies, and maybe some scrappy startups, will be the ones that survive.

This consolidation of the AI industry means that big companies like Apple, Microsoft, Meta, Google and Amazon can use their enormous pools of cash to continue to invest in data centers, new software and manufacturing capacity, says Amrita Roy, an analyst who focuses on the intersection of macroeconomic forces and the tech industry.

Apple detailed a long list of new AI features for its devices at its Worldwide Developers Conference on June 10. PHOTO: BRIAN L. FRANK FOR THE WALL STREET JOURNAL

Roy sees these companies’ investments as an attempt to capture as much market share as possible in what could prove to be a “generational opportunity” to reshape our lives and modes of work. 

Milanesi thinks that as AI features become more useful, the ever-growing demands that these systems place on our devices, from phones to desktops, could eventually reverse the trend of us holding on to them longer and longer. 

 

Her prediction is that, eventually, PCs will be replaced every two to three years, instead of every six, something the industry hasn’t seen in decades. It’s likely also to lead to higher average selling prices for those new PCs, because ones with AI capabilities cost more. That trend might also extend to smartphones, reversing what has been a continuing slump in sales worldwide.

“Remember that generative AI is not once and done,” says Milanesi. “You’re going to continue to push compute, and build models that are larger, and so you will have to continue to upgrade devices in order to keep up with the pace of generative AI.”

If she’s right, hopefully that trend will be driven by the genuine utility that businesses and consumers get from their new, AI-powered devices, and not merely the same kind of FOMO driving much of the investment in this space.