AI Hardware Is in Its ‘Put Up or Shut Up’ Era

Whether any of them will implement chatbots and agents well or in new and exciting ways is much harder to say. While the addition of AI may have been enough to garner the investment needed to build a device, it may not be enough to get people to actually buy the thing. Chatbots and AI agents don’t yet provide enough of a use case to justify people pinning them to their shirts en masse. We’re also at a point of AI saturation where the tech is in everything. So then, what makes your AI earbuds special?

“That’s the problem a lot of these startups have; if AI is their differentiator, then what happens when everybody has it?” Sag says. “It’s now table stakes.”

Wearables and devices built specifically to provide some AI-powered services may have seemed like the logical next step in the AI evolution, but so far the utility we’re getting from them doesn’t push any boundaries.

“The reality is we don’t need dedicated hardware for the kind of features or use cases that they’re showing off,” Ubrani says. “Your phone can do most of those things.”

In the space of a year, AI has gone from being a selling point on its own to something akin to a slightly more potent form of vanilla.

Making a Dent

There are AI hardware success stories, of course, such as the Ray-Ban Meta smart glasses, which have done well by incorporating AI as one of many features in a device that offers use cases—taking pictures, listening to music—well beyond what AI can do on its own. (This will certainly be a year filled with smart glasses, and CES is bound to brimming with them too.)

Meta, of course, is one of those giant companies with resources to put into incorporating AI into its services. Smaller manufacturers may not have the financial stamina to compete, but they’re feeling the pressure to get in on the game all the same.

“It’s going to be difficult to see how those smaller startups survive,” Sag says.

Sag says there are ways to stand out from the big devices and glut of other AI gadgets in the mix. Privacy, for instance. Meta may have the most successful smart glasses right now, but the company’s platform is a data vacuum that sucks up almost every bit of information about its users that it can. Sag points toward competitors like Even Realities or Looktech.AI, which make smart glasses that allow broad user controls over privacy settings and don’t necessarily just send every bit of data back to the mothership. He says startups like those can use the more secure approach to differentiate their products, offering users an alternative to the big, data mining platforms.

No matter how safe and secure the tech is, people are still going to want something that fundamentally does something beneficial for them.

“The next kind of wave of this is like, well, what is AI doing for me right now other than telling me that I have AI?” Sag says. “A lot of AI isn’t necessarily driving sales, because it’s not really changing people’s lives.”