Hi folks, it’s Brad in San Francisco. Silicon Valley is betting that ChatGPT and other generative AI tools will point the way to a smartphone successor. But first …
Three things you need to know today:
• Apple is no longer a safe haven for investors
• The EU started an anticompetitive probe into AI chips
• Deep fakes of Slovakian politicians are circulating
Masa’s money is ready |
A consensus seems to be emerging from several different corners in the tech world: More than 15 years after the introduction of the iPhone, ChatGPT and other generative AI services may soon form the foundation of a new kind of hardware device and an entirely different type of interaction between humans and computers.
On Sept. 27 at Meta Platform Inc.’s headquarters in Menlo Park, California, Chief Executive Officer Mark Zuckerberg unveiled not just an update to the company’s Quest virtual reality goggles but also a new version of its smart glasses, which it developed with the Ray-Ban division of Luxottica Group SpA. The $299 spectacles have speakers, microphones and a camera and can livestream video and make phone calls. They will also let users pose questions to Meta’s Llama AI assistant—asking it, for example, to translate a sign or to guide them through a home repair project.
The first version of the glasses, dubbed Ray-Ban Stories and released in 2021, sold poorly. But Meta (along with Amazon.com Inc., which also released new versions of its Alexa-powered Echo Frames) is now reframing (sorry) the product around generative AI. “Before this last year’s AI breakthroughs, I kind of thought that smart glasses were only really going to become ubiquitous once we dialed in the holograms and the displays and all that stuff,” Zuckerberg said during the rollout. “Now, I think that the AI part of this is going to be just as important in smart glasses being widely adopted as any of the other augmented reality features.”
A day later, there was another signal that Big Tech is ready to build devices around generative AI. Several publications reported that OpenAI, the company behind ChatGPT, is engaged in talks with LoveFrom, the industrial design firm founded by former Apple Inc. hardware wizard Jony Ive, about creating a venture to build “the iPhone of artificial intelligence.” There were few details, other than that Masayoshi Son, the chairman and chief executive officer of SoftBank Group Corp., is ready to invest $1 billion into the venture and that Ive is demoralized by the addictive elements of smartphones and wants to use AI to find an alternative.
Two of Ive’s former colleagues have revealed a bit more about how such AI-powered devices might work. Humane, a San Francisco startup founded by ex-Apple designer Imran Chaudhri and ex-Apple engineer Bethany Bongiorno, demonstrated its wearable “disappearing computer” at the TED conference in May. In the video of the talk on YouTube, which has garnered more than 1.3 million views over the last four months, a screenless, badgelike device with a camera and speaker attached to Chaudhri’s shirt translates his voice into French, takes a phone call, emits text onto his hand with a laser projector and advises him not to eat a candy bar given his dietary restrictions. “For the human-tech relationship to actually evolve beyond screens, we need something radically different,” Chaudhri says.
The company has raised $230 million from the likes of Microsoft Corp. and OpenAI CEO Sam Altman and presumably will use a version of ChatGPT. It says it will start selling the device, dubbed the Humane Ai Pin, on Nov. 9. On Sept. 29, model Naomi Campbell wore a prototype at the luxury brand maker Coperni’s show at Paris Fashion Week.
Silicon Valley has been trying to end the era of iPhone dominance for a while, part of its never-ending scrum to grab the high ground in whatever new computing paradigm lurks over the horizon. While the jury is out on augmented and virtual reality—and Apple’s Vision Pro isn’t due until early next year—initial results don’t seem to promise an entirely new age of computing.
Amazon’s Echo speaker, released in 2014, was one of the first attempts to use conversational AI to peel our eyes off our screens. But it was too difficult to access third-party apps such as games and never displaced the smartphone in part because we’re simply accustomed to looking at displays and pointing, clicking and typing to enter information. That resilient paradigm was established in the 1970s with the first personal computers and a human-computer interface that Steve Jobs borrowed from Xerox PARC.
But with global smartphone shipments falling by 20% since 2021, according to Counterpoint Research, there’s plenty of momentum right now behind the bid to position generative AI as the next new thing. “With many [puzzle pieces] dropping recently, a more complete picture is emerging of LLMs [large language models] not as a chatbot, but the kernel process of a new Operating System,” wrote Andrej Karpathy, an executive at OpenAI who also briefly worked as director of AI for at Tesla Inc., in a post on X, formerly known as Twitter. “Looking at LLMs as chatbots is the same as looking at early computers as calculators. We’re seeing an emergence of a whole new computing paradigm, and it is very early.”
—Brad Stone
No comments:
Post a Comment