OpenAI's First Device Will Be Moveable, Screenless Speaker Built as AI Companion (bloomberg.com) 106
OpenAI is reportedly developing a screen-free, portable smart speaker meant to act as a personalized home computer and humanlike AI companion. "It will help control smart-home appliances, play media, answer questions, respond to messages and tap into the range of capabilities offered by OpenAI's ChatGPT," reports Bloomberg, citing people familiar with the matter. The device, expected to be unveiled this year and released in 2027, would mark OpenAI's first major hardware push after acquiring Jony Ive's io Products. Bloomberg reports: Apple sued OpenAI last week, accusing the company of stealing trade secrets. But OpenAI believes that the device veers significantly from anything Apple has on the market today and that it's unlikely that it violates trade secrets belonging to the iPhone maker, the people said. OpenAI's success in hardware will hinge on bringing a novel approach to the market -- something it aims to do with the smart speaker. For instance, the device's technology is meant to become increasingly personalized and proactive as it gains a deeper understanding of its owner over time, according to the people.
OpenAI envisions the device anticipating needs, surfacing information proactively and serving as an expert on its user, they said. Though the speaker is designed to stay in the home, it will be easy to move around the house. OpenAI believes the product's defining feature will be its personality and ability to connect on a humanlike level with users. The speaker incorporates mechanical elements that can move on their own, creating a sense that it is alive and not just an object responding to commands. The machine also will draw on personal information such as emails to better understand its owner. The goal is for the device to feel like a companion and become a physical manifestation of OpenAI's ChatGPT. Still, the exact plans could change as the company works through the development and legal process.
The device's communication abilities will rely on a more advanced version of the ChatGPT Voice Mode -- GPT-Live -- that OpenAI rolled out this month. The new voice mode is designed to act more like a human. It can listen and talk at the same time, adapt more naturally during conversations, and quickly process information. Though the new product resembles a speaker, OpenAI internally describes it as the first of its kind: a computer built for AI to help make busy people more productive. It includes a camera and other sensors that help it understand a user's surroundings and context, as well as advanced AI models beyond those available on conventional smart speakers. Another central difference is that the device includes a rechargeable battery, allowing it to be carried from room to room throughout the day. A user could bring it into the laundry room while doing chores, move it into the kitchen for cooking assistance, and later place it in a living room or bedroom to have it play music. It can also remain plugged into a single room if the customer chooses.
OpenAI envisions the device anticipating needs, surfacing information proactively and serving as an expert on its user, they said. Though the speaker is designed to stay in the home, it will be easy to move around the house. OpenAI believes the product's defining feature will be its personality and ability to connect on a humanlike level with users. The speaker incorporates mechanical elements that can move on their own, creating a sense that it is alive and not just an object responding to commands. The machine also will draw on personal information such as emails to better understand its owner. The goal is for the device to feel like a companion and become a physical manifestation of OpenAI's ChatGPT. Still, the exact plans could change as the company works through the development and legal process.
The device's communication abilities will rely on a more advanced version of the ChatGPT Voice Mode -- GPT-Live -- that OpenAI rolled out this month. The new voice mode is designed to act more like a human. It can listen and talk at the same time, adapt more naturally during conversations, and quickly process information. Though the new product resembles a speaker, OpenAI internally describes it as the first of its kind: a computer built for AI to help make busy people more productive. It includes a camera and other sensors that help it understand a user's surroundings and context, as well as advanced AI models beyond those available on conventional smart speakers. Another central difference is that the device includes a rechargeable battery, allowing it to be carried from room to room throughout the day. A user could bring it into the laundry room while doing chores, move it into the kitchen for cooking assistance, and later place it in a living room or bedroom to have it play music. It can also remain plugged into a single room if the customer chooses.
Re: (Score:2)
Re: (Score:3)
Just name it Wormtongue.
No (Score:2)
No fucking thank you. Screw your creep.
Re: (Score:1)
I mean, I still can't stand the voice menus when you call support...I''m constantly hitting "0" or saying Operator over and over to get to a real person asap.
Don't get me wrong, I like my AI...ChatGPT can be fun and helpful...but, I don't wanna hold a fscking conversation with it, especially not in public.
Hell, that's one of the reasons I hate the voice support calls, when in public or sitting in Cube Land where everyone has nosy ears....
what? (Score:5, Insightful)
Screenless is stupid. Screens are cheap now. By all means make it audio centric but if I just have to get out my phone to get any info that isn't convenient to get via audio, then I might as well just use my phone for the audio stuff too.
Re:what? (Score:5, Funny)
I can kinda see what they're thinking:
"we could make it a phone app but we want to sell a physical object, and if it was just a thing with a screen it might as well be an app, so we'll make it a speaker, which could also be an app, but is sorta more divorced from the phone I guess. HEY get off our backs we don't have any new ideas, we just need to keep building the possibility we may become a huge apple-like company by saying we're developing this device WHICH WILL BE REVOLUTIONARY. Also old people don't know how to use AI -or phones really- but they sure as hell know how to yell at smart speakers!"
Re: what? (Score:1)
Re: (Score:2)
don't give them ideas!
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
They are probably thinking that they want to be like the computer on the Starship Enterprise. A kind of ambient AI device that you can have a natural conversation with anywhere, anytime.
That was Google's original goal too.
Re: (Score:2)
They are probably thinking that they want to be like the computer on the Starship Enterprise.
You mean the fictional starship which has screens everywhere with a UI whose name is known to nerds everywhere?
Re: (Score:2)
It has screens everywhere, but the primary interface is voice.
From OpenAI's perspective, developing the tech to format information for that screen would delay the release. ChatGPT often screws it up.
Re: (Score:2)
Re: what? (Score:2)
We have to both be thinking of TNG and later because in TOS the primary interface is physical controls and most readouts aren't screens.
In TNG the primary interface for casual use is verbal, but there are displays everywhere and it's common for people to ask for something to be displayed on them, and the primary interface for technical activities is touch screens.
Re: (Score:3)
They are probably thinking that they want to be like the computer on the Starship Enterprise. A kind of ambient AI device that you can have a natural conversation with anywhere, anytime.
That was Google's original goal too.
More likely? They were probably thinking that if there is no screen, there will never, *EVER* be a reason anyone would want to shut off the microphone. And if the microphone is always on, it's a constant and continuous data collector for every home. And you can never supply enough data to these stalkerish creepy businesses.
Re: (Score:2)
Re: (Score:2)
My best guesses for their decision to go sans screen would be to increase
the battery life as much as possible while also shrinking the devices size.
Since I'm no electrical engineer, I can't tell you which pulls more power:
1) An always hot microphone
2) A surface that only activates the microphone after you touch it
The latter would be better privacy wise, of course.
A part of me hopes that engineers have decided too many folks walk about
with their eyes glued to their tiny little smartphone screens as it is a
Re: (Score:2)
which pulls more power:
1) An always hot microphone
A piezoelectric mike wouldn't need to draw any power when inactive (i.e., not hearing sound). A dynamic mike actually generates electricity. (So does piezo, but it's a tiny amount.) Other components would still draw some current.
2) A surface that only activates the microphone after you touch it
Depends on how it's implemented. For both privacy and battery life, the best option would be a physical switch that completely disconnects the microphone.
Re: (Score:2)
Why is it stupid? Your Amazon, Google and Apple all have screenless speakers that do what this does. You might even remember when Burger King triggered a bunch of them with an ad on purpose. (Apple processes their audio stream to avoid triggering their devices so anytime so
Not very "Innovative" (Score:5, Insightful)
Re:Not very "Innovative" (Score:5, Funny)
Great, just what we needed. ANOTHER "Alexa" type of always listening speaker device.
* NOW with extra evil!! *
Re:Not very "Innovative" (Score:5, Interesting)
If I remember correctly, surveys showed that people use smart speakers to do a very small number of things:
1. Set a timer
2. Play music
3. Turn on / off lights (if they have smart devices)
That's really it. I've had a Google Home setup with a mini speaker in most rooms of my house, and a Home compatible smart display in the kitchen. My wife uses the kitchen one to set and check timers whenever she's cooking. The minis get used to mostly turn on / off lights (all lights are controlled by smart switches). We do play music, but fairly rarely since they don't sound super great. She tried using the screen to look up recipes, but it was too small to be really useful so we set up an old computer with a 21" screen for that use instead.
The annoying thing is that since I got "upgraded" from the previous Google Assistant to Gemini, the time to complete a task went from 1-3 seconds on average to about 10-15 seconds on average. They cost Google more now, they take longer, and they still don't do anything better for me. I regret performing the upgrade and wish there was an undo button so I could go back to the fast, but dumb, assistant as that worked quite well for my needs.
Re: Not very "Innovative" (Score:2)
They neutered the old Google. It was faster, but it kept screwing up to the point that (just a few weeks ago) I was forced to update to Gemini, even after holding out for so long.
TLDR; Google screws you either way, so you didn't really lose anything ... and now if you ask it for the Rotten Tomatoes score of something, at least it answers correctly most of the time.
Re: Not very "Innovative" (Score:2)
They neutered the old Google. It was faster, but it kept screwing up to the point that (just a few weeks ago) I was forced to update to Gemini, even after holding out for awhile.
TLDR; Google screws you either way, so you didn't really lose anything ... and now if you ask it for the Rotten Tomatoes score of something, it answers correctly most of the time.
Re: Not very "Innovative" (Score:2)
Re: (Score:2)
When they switched my car's assistant (using Android Auto) over to Gemini, I absolutely hated it. I could not get it to shut up. You'd ask a question and it would go on and on with endless follow-ups. Hey! I'm driving Here! Shut up!. I said "cancel" trying to stop it and it cancelled my whole route in Maps. I had to disable Gemini on my phone.
At the very least, it needs to realize you are driving and then only give you it's best answer to your question and then shut up. You can always ask a follo
Re: (Score:2)
Coincidentally, NHTSA report [nhtsa.gov] (2005) on the (negative) effects of driving while conversing using a hands-free device: "Phone conversation was associated with increased delay in responding to lead-vehicle speed changes, which indicates significant cognitive impairment due to phone conversation."
I expect arguing with an AI which will not stop gassing on about nonsense would be even more distracting.
Re: (Score:3)
Re: Not very "Innovative" (Score:2)
Re: (Score:2)
My wife uses the kitchen one to set and check timers whenever she's cooking. ... She tried using the screen to look up recipes, but it was too small to be really useful . . .
Seems like it would be useful to have kind of a touch-screen tablet device which could integrate timers with recipes. So if the recipe says, "Bake for 30 minutes", touching the "30 minutes" would start the timer. Or saying "start timer" in case your hands might be dirty. Scrolling in response to voice commands would also be good.
And it sounds like you don't cook. Or use timers anyway.
Re: (Score:2)
Oh, it gets worse ....
* Papa Johns Transforms Digital Ordering with AI Ordering Assistant with Google [qsrmagazine.com]: "Papa Johns announced a major transformation of its digital ordering experience, designed to deliver speed, accuracy, and real-time personalization .... the first partner for Google Cloud’s newly expanded AI solution, Food Ordering agent, which is helping the pizza brand deploy a fully unified voice and text AI ordering system to remove friction across customer touchpoints." [Don't you just love mark
Creepy animatronic HomePod? (Score:5, Insightful)
Prior Art (Score:4, Funny)
The goal is for the device to feel like a companion
Hasbro should be the ones suing since they have the original prior art on creepy digital companion with the Furby.
Re: (Score:2)
waitaminute- (Score:3)
it's unlikely that it violates trade secrets
Since when are trade secrets something that can be violated? You either get a patent and enjoy 20 years of patent protection, or you lock your secrets in a vault and hope nobody else comes up with the same novel idea.
We're really leaning all the way into the whole corpo dystopia now huh?
Re:waitaminute- (Score:4, Insightful)
Even if the features appealed... (Score:5, Insightful)
Why would I want a device that deliberately handicaps itself by not bothering to have a screen?
Especially with how verbose AI responses are, I couldn't imagine having to just wait for it to read out the information asked of it.
I don't get the AI companies' fascination with voice-only input and audio-only output. It's a strict subset of what the device in people's hands can already do. Further, every single product that has aimed for this has flopped and you would think they would get the hint by now...
Re: (Score:3)
I think they feel it would be more like humans communicate. But humans invented blackboards and certainly found them useful to improve communication.
Re: (Score:2)
For what, math problems? The last time I used a blackboard to communicate was in Calculus II.
Re: (Score:2)
The whiteboard is the modern "blackboard", and many people use them to communicate in a home setting. They just tend to be smaller than 3' x 9'.
Re: (Score:2)
The whiteboard is the modern "blackboard", and many people use them to communicate in a home setting. They just tend to be smaller than 3' x 9'.
Maybe in your home .... there also exists paint which can be used to turn entire walls into erasable white-boards, with the additional advantage that the wall is easy to clean. Especially nice if you have children (or whoever) who like to draw on walls. (Not a good habit to get into, though.) The paint is rather expensive and requires a primer ... "fully cured in four days" ... OOO! Magnetic whiteboard paint! I didn't know about that! "Holds four pages of printer paper."
Re: (Score:2)
Besides, even with text, waiting for a person to finish talking can be tiresome. "This could have been an email" being a supremely popular sentiment.
Re: Even if the features appealed... (Score:2)
Desperate attempt to excite more investors to make them pump in more money that they'll never see again.
Re: Even if the features appealed... (Score:1)
Re:Even if the features appealed... (Score:5, Interesting)
I don't get the AI companies' fascination with voice-only input and audio-only output. It's a strict subset of what the device in people's hands can already do. Further, every single product that has aimed for this has flopped and you would think they would get the hint by now...
Voice input will always be less accurate than writing. Voice output will always be slower than displaying writing on a screen. Input accuracy is even worse if you don't have a perfect mid-atlantic accent.
I can only come up with two explanations for why companies (not just "AI" companies) keep trying to force this on us.
1. They have so much money they don't care, or they are using investors' money and have a robust contract in place so it's OK if they waste the money.
2. The profit from selling all the data from inside your house is worth loads of money. I can see how knowing exactly what is being said in every household could be valuable to many different customers: advertisers, governments, health insurers.... note also that none of these customers have your best interests at heart. It gets even worse if you work from home and these devices can listen to confidential business calls.
I don't allow any such devices in my house. If I visit people who do have such devices, it's fun to mess with them. Obligatory XKCD: https://xkcd.com/1807/ [xkcd.com]
Re: (Score:3)
I guess they would say there were a lot of iphone like products that flopped before the iphone. But no, the iphone is obviously useful, the smart speaker is just obviously annoying.
Now I do see the possibility of great smartspeakers someday in the future but they would need to
A: be fun to talk to instead of grating (OpenAI has a long, long way to go with this)
B: know you very well without it being creepy (probably by having a local ai running in a box in your home)
C: And maybe tailor their personality to on
The next set of users may be illiterate (Score:2)
Not saying it's healthy, but if there is an untapped market for computers in general, they don't read, and cheap is mandatory.
Yes, wifi availability, IDK about pricing, or why you'd bother making a ubiquitous $5 Alexa. Having said that, in this (kinda awful) use case, you don't need a screen.
Re: Even if the features appealed... (Score:2)
I don't get it either. I use AI a lot, for all sorts of different things. In none of them do I want to talk to it. The only exception I could imagine would be if I wanted to brainstorm about something while driving.
Also, all of the voices have (to my ear) an annoying, fake-bubbly-happy California accent.
Is there an Echo in here? (Score:2)
So a crappy Alexa? (Score:2)
Seems somebody failed the creativity exam.
Re: (Score:3)
"Alexa, this is our new family member, GPT Live. You both have a lot in common, so I hope you'll get along. I have to run some errands, so I'll leave you two to get acquainted."
[several hours later...?]
John Ivy will give it that million dollar look (Score:2)
which you'll happily spend way more money on than you'd normally would.
Re: (Score:2)
Obviously. These companies are only about money, nothing else.
No (Score:2)
Really? Exactly not what they should double down on.
Not in my house (Score:2)
This thing could be always listening.
Things like this need a mandatory push to talk button which mutes the microphone.
Sounds ghastly (Score:5, Funny)
OpenAI envisions the device anticipating needs, surfacing information proactively and serving as an expert on its user, they said.
Will it have a Genuine People Personality? Perhaps a cheerful and sunny disposition?
Re: (Score:3)
Can it be programmed with a very large ax?
Re: (Score:2)
Mod parent extra funny, though the story deserved more.
Isn't it just a phone app? (Score:2)
The customer already has a high-end, power-managed speakerphone in their pocket.
Nobody needs this object.
They should be making The Orb of All Truth, Facts, and Wisdom,
instead of making Microsoft Intimate Intern - Home Office Edition.
At least OpenAI has that Jonny Ive glass and aluminum money.
No Teenage Engineering plastic trash for them.
Microphone always on (Score:3)
Since it's a screenless device, that means that it's primary microphone interface will be active all the time.
What would you like to bet that they're going to sell recordings from your home to advertisers?
Re: (Score:2)
What would you like to bet that they're going to sell recordings from your home to advertisers?
You have a device with a microphone and a cellular modem in your pocket that's always on. What's to say they're not already doing that?
Yum yum (Score:1)
Stupidest product I have ever heard of. (Score:3, Insightful)
It is designed under the assumption that AI works better than it does without any need for verification.
Screens are cheap, put one in just to list what prompt the AI heard, you idiots.
Smart homes are dying because they were stupid BEFORE AI came about. No one is willing to let an AI control their temperature, let alone the lighting, doors, etc.
What is going on is people are trying to create products for a Science Fiction version of AI, when what we have is closer to a Horror movie version of AI.
Make me work? (Score:3)
Why does it have to be locked to one device? If I have a few speakers around, maybe a laptop and my phone as well, why can't that AI software follow me around instead of making me carry it? This push is just to make people anthropomorphize their devices for emotional lock-in and a whole new level of brand loyalty. If it's so useful, why put it in a cage?
Re: (Score:2)
Because this smart speaker is likely to be much, much more expensive than most speakers you own. If they made the AI just a service that you could just use with your existing devices in your home they would lose the chance to sell you the expensive gadget. Exactly the reason the Humane AI Pin was created when an app on your phone could have done the same thing.
A shittier version of Siri, Echo, Alexa (Score:2)
I've used all the old voice assistants, every time I asked a question they all seemed to default to a wikipedia search. A wikipedia search with a laggy, mistake-prone voice input.
So now there's a new "AI" version that won't even use proper wikipedia, it'll just make stuff up.
Unfortunately, it will probably be wildly successful.
Re: (Score:2)
Alexa is the shittier version of Alexa. It used to be good, but nowadays it is barely useful.
Re: (Score:2)
100%. My wife and I were perfectly fine with the old voice and "attitude" - the new "sassy" version really just grates on our nerves, and all her responses take 10-15s instead of the 1-2s they used to be. Just a complete downgrade.
It's hard to imagine (Score:3)
...anything I would want less than this.
Re: (Score:2)
Well, I can think of some uses if it were actually up to the job. I'd often like to have it read something to me, but it would need to be able to understand the emotional background.
E.g.:
Strange thing are done neath the midnight sun by the men who moil for gold,
The arctic trails have their secret tales that would make your blood run cold,
(etc.)
Or possibly have it read me The Lord of the Rings. And I have some fond memories of the radio program "X minus one".
But I *really* doubt that current AIs are up to
Re: It's hard to imagine (Score:2)
This is going to be awesome (Score:2)
This is going to be awesome. I will be first in line to buy it.
With house full of guests (Score:2)
It sounds stupidly redundant. (Score:1)
When the version of iOS is released iPhones will be able to do everything this does, be more portable, and they have a camera for uploading photos to ask questions about. Google and Samsung phones already have that functionality. So do a lot of Chinese phones. Why would anyone buy a low end speaker that does less than the phone they already take everywhere? I cannot see how this appeals to anyone but industrial design geeks who idolize Jony Ive.
TJ Bass Companion Cybers (Score:1)
In The Godwhale and Half Past Human novels by TJ Bass, companion cybers were human level intelligence advisors that at least some humans carried. The two I remember are "belt" which was exactly that a wide belt and toothpick a "javelin" about 3' long. There was a less portable one called "ball" that was a basketball sized "crystal ball" as well. They were very logical and well educated advisors and did not "appear" to be reporting all your activities unless they had to to get you necessary help. These b
A screenless speaker? (Score:2)
What a remarkable invention.
I mean you look at all those speakers with screens they kept building in the Olden Days between the invention of the radio and the invention of the television and think to yourself "If only we had AI back then, they could have saved a whole lot of wasted resources by just not building all those useless screens".
Thank god for Sam Altman.
But will it blend (Score:2)
I sense a million seller everybody wants and needs (Score:2)
Bluetooth speaker (Score:2)
ChatGPT didn't tell them bluetooth speakers were invented 25 years ago?
So they're making an alexa? (Score:2)
Hot damn (Score:2)
Oh great (Score:2)
Sigh. (Score:2)
"OpenAI's success in hardware will hinge on bringing a novel approach to the market"
It says as it releases an Alexa....
Another AI competing against your cell phone (Score:2)
So, OpenAI will access to your e-mail accounts: Do you really need OpenAI to know your banking details and medical history?
Notice how they talk about the 'hardware' on the kitchen bench doing stuff, not the OpenAI database with a history of your e-mails, at-home conversations and search requests.
The cell phone and Alexa can answer simple "just the facts" questions. What else can people ask a audio-only lump of hardware to do? Like Alexa, this just stops people picking-up their phone, at a much higher
Alexa By OpenAI? (Score:2)
So, OpenAI is launching their own flavor of Alexa?
And?
So, a robot! (Score:1)
China just banned them (Score:2)
China just banned AI companions ... now US starts building them.
https://finance.yahoo.com/tech... [yahoo.com]
I hope the new companion will at least have a cube shape and be weighted.
You are the product (Score:2)
It will be the Zoon of our time (Score:2)
This is the future, not 1984 (Score:2)
All citizens will be required to carry an AI clone of Big (Corprate) Brother with them at all times.
Integration with ChatGPT apps will be critical (Score:1)
The limitations on the smart speaker in the past have been that it doesn't have access to enough services to be terribly useful. The end-user value will be linked to all the services the speaker can access, such as podcast player, subscriptions like music streaming, Netflix, other podcast-hosting apps, social media apps, email, text messaging, and web browsing history. If the app ecosystem can funnel audio streams to the smart speaker, it can capitalize on all that media and essentially curate your daily li
6 billion well spent (Score:2)
Or probably more 5 billion well spent. The soft-porn with Jony and Sam probably cost 1 billion alone,
Codex Micro announced today, not a speaker (Score:2)
https://openai.com/supply/co-l... [openai.com]
AI Companion for Mentally Ill People (Score:2)
Pack it into a Pokemon (Score:2)