AI Devices: OpenAI and Jony Ive Plan an "Always-on" Device

A new AI device from OpenAI and Apple's ex-chief designer Jony Ive is intended to be "always-on" and protect data privacy. A prototype is already circulating.

listen Print view
Jony Ive and Sam Altman

Jony Ive and Sam Altman

(Image: OpenAI)

4 min. read

The secrecy and the well-measured but very prominent public relations in the run-up are suspiciously reminiscent of the failed AI Pin: an Apple designer, artificial intelligence, and a specially tailored device for it – that was exactly the recipe, with which former Apple employees Imran Chaudhri and Bethany Bongiorno promised great things two years ago. The rest of the story is known: the startup Humane was bought by HP, and the AI Pin was discontinued.

A similar constellation is in the pipeline for the coming years, but with ChatGPT maker OpenAI and Apple's former chief designer Jony Ive, the project is in a completely different league. Monetarily too: Altman's company has already generated 6.4 billion US dollars in warrants and taken over Ive's company, "io". Some are already suspecting: If Ive and Altman don't succeed, it would possibly be the death knell for the idea of creating dedicated consumer devices for AI. While Google and Meta are also researching in this field: Google is developing AI glasses with Warby Parker for 2026, Meta acquired the startup Limitless, which builds "AI-Memory" wearables. But the prominent constellation is given a lot of importance. The focus could then definitively shift to enhancing existing devices with AI rather than creating new ones specifically for it.

But despite all skepticism, the industry is still eagerly watching what might be released in California as early as next year. Reuters now reported on a prototype that is circulating internally. And on the framework conditions necessary for such a device. Above all, the report contains something sobering: Because Ive and Altman's grand vision is not yet taking shape in 2026. Instead, there is talk of a growing device family, which will only be followed by the big breakthrough later.

Videos by heise

A basic requirement is described as the OpenAI device always being present. Unlike the smartphone, which also seems ubiquitous, it should not take breaks. Only then can it have the necessary context to be useful as an everyday companion. Which applications these exactly will be remains to be seen. It is conceivable that the device should proactively remind its owner of things because it notices when someone calls out to them or the user performs a task that entails follow-up tasks. It could also be a kind of digital memory that allows one to recall and query something heard at some point during the day.

For OpenAI, the device in its planned maximum configuration is a particular challenge, according to Reuters, as it requires highly miniaturized hardware, unlike energy-intensive cloud hardware, which is actually Apple's specialty. It should be energy-efficient and fast. And powerful, as local data processing is likely to be given great importance. Only then would people accept it in their everyday lives – sending everything to the cloud is unthinkable. OpenAI is currently investigating the development of a custom chip. And that is only one component: the company must also provide AI models that are powerful and can be operated on a small device at the same time. So far, OpenAI has focused more on the cloud.

There was already talk of an entire device family recently. Before the described AI device with a local model appears, there will first be variants tailored to specific tasks that will then rely on the cloud. It remains unclear how long this transition will take until OpenAI is actually able to bring constantly active devices to market that pay special attention to data protection.

(mki)

Don't miss any news – follow us on Facebook, LinkedIn or Mastodon.

This article was originally published in German. It was translated with technical assistance and editorially reviewed before publication.