▲ 36 points
back
82 comments
Does anyone actually use ChatGPT’s voice mode? It seems technically impressive, but when I try using it, it feels… deeply uncomfortable talking to a machine like you would a person.
Treating it like you would a person feels silly: It’s not a person, it doesn’t have any life experience or anything interesting to say, it doesn’t have real opinions or anything worth talking about that isn’t a direct answer to your question. So having a “conversation” with it feels like I’m doing role-playing with a computer. And it gets into “weird” territory, fast.
But if you treat it like you would an LLM chatbot, to get specific answers to questions, or to explain something to you, it’s so. much. worse than having the answer on a screen for me to skim and reference directly. Either the answer is too brief to be helpful, or too long and it’s talking forever and I have to make it stop. I suppose certain highly specific questions that have brief and concise answers, it may be good at?
Maybe I just haven’t found the right use case for it yet. Or maybe my discomfort is just my introversion coming through, and normal people don’t have this problem.
I've talked to ChatGPT in the car whilst driving to learn various topics. It easier to talk than type and the real time verbal feedback helps with the learning.
It is a bit strange, the outputs are the result of complex mathematical computations and math doesn't have feelings, but it does occasionally generate genuinely useful outputs.
If you're aware that you are talking to a bot, then I see no problem with it. The problem is the dissonance people get talking to an AI like its real.
> it feels… deeply uncomfortable talking to a machine like you would a person.
That's why the form factor should be a wand. Flourish it to activate, and speak your spell.
I wonder what exact features this would provide that the mobile app cannot. I don’t wanna carry another device with me all the time
> I don’t wanna carry another device with me all the time
And that's why this (and any other stand alone "AI device") is doomed to fail.
We solved this problem already in 2007. "An iPod. A phone. An internet communicator. Are you getting it? These are not three separate devices."
The industry has been chasing that high ever since, failing to stop and consider that maybe the smartphone is effectively peak consumer tech as a delivery vehicle for various services.
OpenAI is just upset that they don't get to own the delivery vehicle, Google and Apple do.
The description makes me think that it's targeted at the "screenless" trend. Especially among young people there is a trend to cut down on smartphone usage. This device would fit right in with that: no screen, status display with LEDs and oldschool moving parts (presumably engineered to be cute or endearing). Something to carry around when you don't want to be distracted by the temptations on your phone screen
The irony of selling an AI assistant with that premise is not lost on me
> I don’t wanna carry another device with me all the time
Hopefully they've learned from the mistakes of the Humane AI pin and/or the Rabbit R1.
https://en.wikipedia.org/wiki/Rabbit_r1
https://en.wikipedia.org/wiki/Humane_Inc.
EDIT:
From the article "The idea is to make the device easy to carry around the home with one hand".
I don't understand the market target at all for this device. You want the equivalent of the Amazon Echo Dot but you also carry around your house?
Following the same reasoning, smartwatches would have been a failure. They might be niche, but sales prove there is a market to explore/deepen.
Overall, openAI can sit on a mountain of money and bet mobile phones are the final form factor for human/AI interaction, as they have been for human/internet.
Or deploy some of that money and figure out whether that is true.
It’s also about creating an environment for those features to emerge, without being constrained by someone else platform, that could natively embed it at any time, and (probably more important) app marketplaces.
A signal for opportunity is also given by Apple’s inability to give their users Siri and Google sitting on transformers while OpenAI grabs the chatbot marketplace.
Also, anything they build does not cannibalize their other product lines.
There are lots of incentives for them to try out and figure out features on the fly, which arguably also happened with the IPhone (app store).
It will probably be pretty.
in the US you may ask Siri about stuff in Europe it still takes about half a minute to activate GPT and talk to it. Investing $1000 in a new iPhone might solve the problem soon but at this point I'd rather try a new $300 device just out of spite
A device thats worse than the phone you already have, which is dead the second you stop paying your monthly OpenAI sub, and is useless if you decide you prefer other LLM provider. Where do we sign up?
Voice is almost always the worst, most frustrating interface.
Think about how much scrolling through UIs, settings menus, and different interfaces we have for every app that are all designed differently. Is this really peak technology, or something that is just the best we’ve been able to do without “smart” agents that can manipulate these programs themselves?
Is swiping on a flat screen really the endgame?
It's going to be a phone, a screenless phone. Which is fine I guess, but only the obnoxious people that watch youtube on speaker with everyone giving them a death stare are going to use this. Maybe that's a bit too harsh, it will definately have a bluetooth connection for earbuds, so it might not be so grating. If they roll the data into the ChatGPT subscription, it will probably pick up a decent # of users. I can't imagine having a main device only use a voice interface though. Skimming an email visually is way more efficient than having an LLM summarize it.
Isn't this a plateaued market?
Amazon, Google, and Apple are already entrenched in the space. It's not like any of them have toxic brands like Oracle where everyone is dying to free themselves. Most people are happy with their products.
This seems like a tough move...
I haven't kept up. Are Siri, Alexa or Google's smart speakers any good now? Or are they still mostly for setting timers and controlling spotify?
The other article pitched it as an "AI Companion". I can see the market already; virtual girlfriends or virtual friends for the homebound. People will buy a "Companion subscription" that will allow them to talk their friend all day. It will be very successful. I feel bad for any human moderators that will have to wade through these conversations.
Judging by how much effort Apple puts into its speakers it isn't a large market, probably not even on the scale of their headphones.
Between Apple's poor quality AI, Amazon continually trying to sell you more stuff and Google flogging your attention to every corporation out there I'm surprised anybody buys these products anymore.
Gotta start v1 somewhere and ChatGPT was and is a big enough threat to Google Search.
Privacy nightmare + worse UX than a phone or sitting at your desk? Dead on arrival. Not a fan of Alexa but at least it can change what you're currently viewing on your Fire stick.
This seems like a more expensive version of the rabbit or friend ai pins.
What is going to differentiate this above any of (1) my phone that can use any AI provider and I can already carry around the home, (2) a set of smart speakers that can talk to an AI provider like google home/homepod/alexa, or (3) my computer?
1. You can't just press a button on your phone to instantly start an ai voice conversation
2. You can't bring your smart speakers with you everywhere, also same as 1
3. See 1 and 2
This thing is dead on arrival. The space is already crowded and no one needs another device listening in your house. Already plenty to choose from.
Counterpoint: the number of product lines in the space shows high customer demand, but the existing products are all pretty bad. If OpenAI can deliver a better product with a business model that works that's a huge market opportunity
This feels like the obvious future. Everybody doom scrolls on their phones too much, but we can’t get rid of them because of the convenience of the productive things we can do when we need to.
So if there was a way to enable the productivity (or even multiply it, let’s be real) without the doomscrolling, a lot of people would go for that.
Not to mention the endless fiddling we’ve had to resort to with phones. We’ve gotten used to it, but apps for everything is such a cognitive burden on people. A screen-less AI assistant is the perfect solution for this. I can’t wait.
People like doomscrolling. They could uninstall their doomscroll apps, but they won't. This puck wouldn't stop most from doomscrolling either. Probably a few, but it's already possible for a few people to stop doomscrolling.
I keep seeing this argument, but there is nothing preventing anyone from purchasing a "dumb" phone. How would you envision multiplying your productivity in the day-to-day with a voice interface?
Why do I need to agree to arbitration and waive class action lawsuits to read an article?
Maybe based off the proprietary Apple expertise that led to the hockey puck mouse.
I think a truly open source OS and phone is what both Facebook, OpenAI, Amazon, Netflix, Epic, Microsoft, Spotify and many others need to get out of the grips of Apple.
It’s quite ironic that all these companies pour billions in failed attempts of stores, hardware, platforms, lawsuits yet fail to collaborate in any meaningful way to establish a new standard that is truly better.
It implicitly also proves that their problem is not with the app stores being closed, but that they don’t own it.
There’s nothing wrong with the phone, we don’t need glasses, hockeypucks or new contrived ways to interact with the digital world, we need a truly open hardware and software stack for the benefit of humanity, no lockdown, no one company in control, just a healthy platform where apps can thrive in a fair competition.
I feel like there will be a minimal UI/no screen but it will have 9 far field microphones. 2-3 cameras , an iris scanner, GPS, and lidar. This is because it is a data collection device for use in OpenAI's future models, more than it is a an assistant.
> The idea is to make the device easy to carry around the home with one hand.
So their target audience has a large enough home that they need a completely separate device since they have areas that aren't covered either by smart speakers or their phone. But the device has little functionality outside the home, because it requires talking to it which usually cannot be done in public. How large is this group of wealthy shut-ins?
The previous "smart pins" made more sense. The first thing I would want to do with this new device would be to put a chain around it and hang it around my neck, but it seems specifically designed to not do that.
> The first thing I would want to do with this new device would be to put a chain around it and hang it around my neck, but it seems specifically designed to not do that.
How do you figure?
Assuming instead of being another phone or voice assistant: It might make sense as a local AI server/proxy expanding to home assistant/TV. If they could claim to need to have device id's to avoid their models being scraped, they could force users to use their device proxy and avoid middlemen (likely by making it cheaper than direct access), and likely save a lot of server time by doing some processing locally on client's dime. Then they could start to play the device-id game with 2-factor authentication, etc. and their stock would track their install base like Comcast STB's.
If Johny Ives delivers a hockey puck then he is diminished in my eyes.
Of all the things you could breath AI life into, what could be more boring, bland and already done than the Alexa hockey puck.
Sadly, zero interest on my part.
If I need information, it is already within easy reach:
"Hey (Google|Siri), <question>?" or Just get my phone out and use website/app of your choice.
The device is presumably listening and recording everything and you can bet your last dollar there is fine print in the TOS that says Altman actually owns everything you say near the device and can use it for whatever he wants.
Oh, you thought you bought the device! Silly peasant! you just have the privilege of paying me to rent the device from me. Remember you will own nothing and will be happy.
Given the almost universal negative press about Altman, I wonder how many people would actually want a device from Sam 'let me scan your eyeballs for make believe money' Altman in their home.
A real life companion cube, even with the same corporate juggernaut flavor.
So its a competitor to Homepod and Alexa, but more capable, possibly through Computer Use among other agentic workflows people have today.
DoA and won't be allowed in any space that values privacy.
Is this the device they picked up Johnny Ive to design?
I hope it has little arms that gesticulate while it talks, like Stan from the Monkey Island games. Maybe it could have a Trump mode as well.
“ No wireless. Less space than a Nomad. Lame."
Related:
Apple sues OpenAI, accuses ex-employees of stealing trade secrets