It’s a smart donut. That’s the only conclusion I can draw about OpenAI’s first, groundbreaking piece of AI hardware after reading Bloomberg’s revealing but unconfirmed report.
Here’s the TLDR summary of these rumors: This donut-shaped, hockey puck-sized AI companion, designed with Jony Ive’s LoveFrom studio, can sit on a table, be carried (or maybe worn). It’ll be festooned with sensors, cameras, microphones, and speakers. Meanwhile, it’ll be a vessel to bring ChatGPT closer to you and your life. It’ll even shape-shift a bit to indicate a response (can a donut shrug?).
OpenAI might, Bloomberg claims, price it between $300 and $400 (or around £220-£300 / AU$425-AU$570). I know. That’s almost instantly a hard no for an AI companion that has no defined purpose besides offering always-with-you AI. Even the woebegone Rabbit R1, which also features a camera, speakers, some sensors, and even a rotating camera, costs just $199 (or about £150 / AU$280).
Latest Videos FromTechRadar
After reading through this, I still wasn’t feeling it. Why do we need this gadget? What problem does it solve? Really, why do we need any AI on or near our bodies?
If this product arrived three years ago, would people, including me, feel differently? After all, in 2023, we were all excited about the potential of generative AI. These days, it’s fair to say half of us hate it or at least deeply distrust it and hate what it’s doing to our resources (water and energy) and landscapes (data centers).
On the other hand, maybe I was judging OpenAI’s upcoming device too harshly, especially without having even seen it. That sparked an idea.
Bloomberg’s report included enough detail to create an image in my mind’s eye, and I wondered if the leading AI platforms, Gemini and, yes, ChatGPT, could use a prompt to generate simulacra of the devices, in situ.
Gemini and ChatGPT take a run at it
I started with this prompt:
“I need an image of a consumer electronics gadget that is shaped like a donut and the size of a hockey puck. It should include tiny microphone holes and grills for a pair of speakers. It should have moving parts, maybe pieces that shift against each other without massively deforming the donut shape. Let’s make a pair of them in white and gray on a table. Let’s also include an image of someone wearing one like a pendant.”
And in full disclosure, after getting my first couple of images, I added this prompt because I realized I left out a pair of key features:
“Without changing the design, can you add a camera and a couple of small black sensors? I want the image to otherwise remain unchanged.”
These AI image generation systems are now good enough that they can maintain consistency between image generations (which they did here), so I’ll only show you the finished products from both platforms.
Here’s the final render from Gemini Pro:
At a glance, they’re kind of cute, looking like giant, digital Lifesaver candies. The speaker grills are kind of huge, but I like the subtle placement of the camera and sensors. I can infer the possibility of movement from the multi-part frames.
Gemini is, I can see, a bit unsure if this is a wired or wireless device, but it has made at least one intriguing leap. It assumes, for instance, a connection to a phone-based platform called Aether (the gadgets may even be called “Aether”). In medieval times, according to Gemini, Aether was considered the fifth element. “It was believed to be the pure, heavenly material that filled the region of the universe beyond the terrestrial sphere,” wrote Gemini. Yes, that feels a bit like the proliferation of AI.
The fictional device does look as silly as a pendant as I expected, but overall, I’m more intrigued by this device than I was before seeing the Gemini render.
Now let’s take a look at the ChatGPT (Free) render:
This is a simpler, even plainer device and might fit in a bit better with Jony Ive’s previous design oeuvre. I like the low-key speaker grill design, and don’t mind the smaller activity lights.
I appreciate that ChatGPT helpfully includes a schematic that breaks down all the key features and even offers a look under the hood. It also manages to look a little less obtrusive as a pendant.
Let’s get real
Obviously, this is all guesswork by Gemini and ChatGPT, and it’s based on prompts built on rumors. Put another way, OpenAI’s upcoming AI wearable could end up being significantly different than either of these generative images.
Somehow, I don’t think so. I have a feeling that we will see something small, round, lightweight, and very, very intelligent. The launch, when it happens, will be exciting and buzzworthy. Ive’s voice will likely drive the brand narrative as the donut-shaped device floats in an all-white background before settling on a desk or, better yet, in the palm of someone’s hand.
That moment, though, won’t change the trajectory of OpenAI’s AI gadget. When it arrives, it will face an uphill battle to attract consumers who are already suspicious of AI’s growing influence on the world. They’re tired of AI slop, angry about resources and jobs, and the last thing many want is an AI companion reminding them daily about their frustration.
I don’t even know what to say about a donut… er… gadget that squirms in your hand. That sounds positively creepy.
If there’s any bright side to these rumors, it’s that they probably put to bed the idea that OpenAI is allegedly stealing Apple product secrets. The iPhone maker would never build something like this.
I could be wrong about OpenAI’s chances in this space. This might be the first AI gadget to break through. I mean, look at those designs. Aren’t they cool? Sure, they are, but they’re not reality, and we all know that this gadget will soon face a very harsh reality.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
https://cdn.mos.cms.futurecdn.net/G9rMYTLsQMQd7Fc3kZuq4P-2000-80.jpg
Source link
lance.ulanoff@futurenet.com (Lance Ulanoff)
