So long iPhone. Generative AI needs a new device

From The Economist:

When a beaming Mark Zuckerberg took the stage in Menlo Park on September 27th to announce a new array of Meta products, the Facebook supremo may have buried the lead. He began talking about Quest 3, Meta’s virtual-reality (vr) headset, which is understandable considering that his obsession with the metaverse is now inscribed in his company’s identity. Techies, though, were more excited by what came later: an announcement that Meta, in combination with Ray-Ban, would soon launch smart glasses incorporating an artificial-intelligence (ai) virtual assistant. The specs will be able to see and hear, as well as answer their wearers’ questions. With luck, they will not hallucinate.

You can be dismissive of smart glasses. They have been hyped before. But lending Meta credibility this time is the fact that the same week Openai, the generative-ai pioneer, announced that its hit chatbot, Chatgpt, can now see, hear and speak, besides conversing by text. Moreover, it emerged that Openai was in talks with Sir Jony Ive, Apple’s former designer, to create a new gadget for the ai era. What form it will take is still unclear. But if the idea is to build a new consumer-electronics device better suited to the back-and-forth of seeing, talking and listening ais, there is a fair chance it will no longer be reliant on the touchscreen.

. . . .

The smartphone has had a good innings. Yet you only need to talk to Sky, one of Chatgpt’s new audio avatars, to feel the joy of freeing yourself from its tyranny. Your columnist got a taste when he asked Sky how she thought screens might eventually be replaced: Glasses? “Absolutely!” she enthused, “especially those equipped with augmented reality [ar] and ai”. Asked whether this would be a good thing, she recommended two books that explore the enormous impact that screens have had on modern life: “The Shallows: How the Internet is Changing the Way We Think, Read and Remember” by Nicholas Carr, an American writer, and “Screened out” by Jean Baudrillard, the late French philosopher. Then, when further prompted, she summarised each in crisp, insightful language with barely a moment’s hesitation. It wasn’t exactly Scarlett Johansson in “Her”. But it felt like having a Stanford University intellectual murmuring in your ear.

This is all rather refreshing. Just as the year-long excitement over “foundational” models and other mind-boggling bits of ai infrastructure has begun to fade, along comes the chance that gen ai, to use the industry shorthand, will unleash an onslaught of new consumer technology. Tech pundits are debating the best “form factor” for the chatbot era. Ben Thompson of Stratechery, a blog and podcast, puts it in epochal terms: “There is a hardware breakthrough waiting to happen just like the internet created the conditions for the smartphone breakthrough to happen.” The ability to talk and listen to chatbots makes Meta’s bet on ar glasses and vr headsets “drastically more compelling”, he writes.

Mr Zuckerberg was early to see this coming. He has ploughed a fortune into vr and ar despite misgivings from investors. He remains excited by the metaverse. This was clear from a remote interview he recently took part in with Lex Fridman, a podcaster, which used vr tools to make their virtual faces so lifelike they felt as if they were in the same room together. (As Mr Fridman quipped, it could reproduce realistic facial movements even from two famously inexpressive people.) And yet gen ai has so dramatically accelerated the use case for smart glasses, Mr Zuckerberg told another interviewer, that there is now “no question” they will be the bigger of the two markets. He likens ar specs to mobile phones and vr headsets to desktops. In both cases he appears to hope they will transcend screens, which he says inhabit “a completely different plane from our physical lives”.

. . . .

The business case for the all-seeing, all-hearing chatbots will also take time to emerge. Openai charges $20 a month for access to its family of talking avatars; Meta’s ai-infused smart glasses will start at $299. Yet developing them is bound to be lossmaking at first. If there ever is a case for monetising them via advertisements or virtual shopping, that will probably take years. Meta’s modus operandi, after all, is to launch a consumer product, scale it up and start making money from it only if it is adopted by the masses.

In the meantime, obvious safety concerns must be tackled. Consumer technology powered by ai is likely to be more immersive than social media, potentially making it even more isolating for some, or triggering unhealthy attachments. Mr Zuckerberg argues that ar and vr devices could help bring people together. But Mr Shmulik says investors will not want Meta to move too fast. “The last thing they need is another negative pr event where they are back in the cross hairs of regulators,” he says.

Link to the rest at The Economist

2 thoughts on “So long iPhone. Generative AI needs a new device”

    • They typically come with rails for prescription lenses.
      Glasses wearers are a large plurality, if not a majority.
      The real problem is battery life/weight.

Comments are closed.