So the AI Pin comes with its own phone number? Couldn't it be paired to my current eSIM? I don't need a second phone number... In any case, 700USD plus 24USD monthly subscription is a bit too much IMO.
A lot of the interactions seem to take longer vs me just taking my phone out and quickly doing what I need to do. Its compelling given this is a v1 of the product so it will only get better from here but not completely sold on it just yet.
Absolutely. It was a snooze fest but as a presentation, even a few minutes in I thought I'd missed the part where they told me what it does and why I needed one asap.
Not that it matters either way, but I hate the recent Apple style of presentation with everyone being overly happy, hand-wavy and eerily excited about mundane features. This at least felt different and somewhat more honest.
They ask it where the next eclipse is, and where best to watch it. They got the date right, but the suggested locations of Timor and Australia are not in the eclipse path. https://www.timeanddate.com/eclipse/solar/2024-april-8
That also doesn't look like enough almonds for 15 grams worth of protein. 10 almonds have approx 3g of protein, requiring 50 almonds for 15 grams of protein, I don't think there's that many in the video.
Makes you wonder about the veracity of the AI (or the accuracy of the demo). Looks like a cool product either way.
so it's like a phone, but there's no apps or functionality other than what's built into the base OS, and there's no screen or input controls other than voice and gesture, it only plays music from Tidal, and only connects to t-mobile, and looks to everybody around me like i'm always wearing a camera pointed at them?
Very Star Trek-esque. It seems a bit clunky but a good first version. I could imagine this being pretty powerful in 5-10 years when everything is integrated and the AI can do almost anything you ask it. Until it can 100% replace the phone (which seems to be their aim) it seems like an extra device that’s a bit unnecessary.
Says there is no wake word, so it's not always listening, but responds to your voice... aka it's always listening then no? And rather only processing info when it thinks it hears a wake word it would always be processing no? Having to hold up my hand to get a limited screen, all seems to be solving a problem that is already solved with a phone, and if one wants a smaller device use an Apple Watch with on device Siri and cellular. This seems far nitcher than even Google products that always appeared dead on arrival.
Strange given it's placed in a way where it's always there. Did seem like a bit of a contradiction when he said that. It's always listening but it's not always listening but it is.
It's got a hardware light that tells you its active, what more do you want? You have your phone out all the time and it's got a camera facing people too
That's a funny comment. I don't wear my phone on my chest all the time. If I'm outside it's in my pocket. If I have it in my hand, the camera is facing down. There's no comparison to this consumer grade bodycam.
When I first heard about the company I knew immediately it would be a Star Trek badge type device.
> Google learned a similar lesson in 2018 after it launched Google Clips, a body-worn camera that used algorithms to automatically snap photos. Female users tended to end up with an abundance of cloud shots when they intended to record what was in front of them, because the device was not designed to account for bodies with breasts
This feels like it could have been a classic Silicon Valley scene but I'm going to give them the benefit of the doubt and assume that it was designed at Google's secret Castro office.
> Tapping the Pin and then moving a palm into its field of view activates its laser, which projects images and text onto a user’s hand at a wavelength that produces a blueish-green tinge, a 720p-resolution system Humane calls a Laser Ink Display. Tilting the hand navigates between displayed options and a swatting gesture swipes to a different menu. Users “click” on an option by tapping their thumb and index finger together and close their hand briefly to return to a home screen.
Okay, that's freaking cool. Anyone here willing to take the $700 plunge to test this? I want to know if it's as ridiculous as it sounds.
01:49 > The AI pin privacy chip also protects it from being exploited, which means if it's ever physically tampered with, it will require service from humane to restore operation.
Is that real or CGI? I looked at it for a long time and I didn't find any smoking gun (perhaps too much shacking, it looks like fake shacking). I've seen too many videos of Captain Disillusion, in particular https://www.youtube.com/watch?v=KbgvSi35n6o
Their demo of AI Q&A in the announcement video (@3:36) is totally wrong. The Pin says the April 2024 total eclipse will be best viewed from Australia and East-Timor... Except the eclipse passes over the US and will be nowhere near Australia. The answer seems to be about a partial eclipse in April 2023. How was that not fact checked???
That's what you get when your futuristic "AI pin" works by making an API request to an off-the-shelf LLM (ChatGPT I assume) that was last trained a year ago.
They're not "color options" they are "colorways", you rube. And it doesn't come in three shades of grey, it comes in "Lunar", "Equinox", and "Eclipse". Those colors are worth easily $100 of the MSRP right there.
Ha you're kidding right? This is almost an attempt at adding another product to Apple's existing range of hardware, which is all phenomenally close to this hardware by being a chip, a camera, audio IO, and in their case a screen rather than a projector. You couldn't get closer if you tried.
The demo was showing how you can interact with it. I don't think this is a big deal. If they're using GPT-4 or another model, then it will be resolved when the AI model is.
What reason do we have to believe that the underlying technology will improve enough for the product to be actually usable? Many of this product's features require a degree of trust that clearly isn't merited at this stage. An assistant that tells me to go to an event at a place it's not happening is worse than useless. Their advertising shows a person asking "can I eat this?" with a piece of fruit -- that's not a question you want a hallucinated answer to!
They did an excellent job. The small delays are close to impossible to avoid at this point. Although I may have seen one or two demos that managed to avoid them for conversation only. But those systems could not take any actions.
It would be difficult for them to be more poorly positioned in the market. They don't own their network, are leaning on powerful incumbent players for the base of their tech stack, and are entering a market where any one of a number of existing global brands are poised to crush them the second the concept shows the first sign of gaining traction in the market. Not sure if this is a naked attempt to grift early adopters, dumb money, or an acquisition play. Hamhanded in any case.
Good exploration of the clip/pin/brooch form-factor that was previously explored by the Narrative Clip (http://getnarrative.com/). Can it be worn on a necklace? That seems more convenient than attaching it to different pieces of clothes and having it tear/wrench at the fabric.
The Narrative Clip is 18g, this is 55g... I can tell from experience that 18g on your t-shirt is OK, but just barely. A 55g is clunky like crazy. It's OK if you're going to do some action cam footage in a session, but not wearing all day "without thinking of it".
I’m a Humane employee who has worked on nearly all the Ai features in the product. We’re very happy to get this out into the world and start getting feedback!
The Comm Badge was never meant to be a full interaction device, just a point to point communicator. Since it has limited UX capability, a single use case made sense.
Why are you trying to do something totally different with what appears to be an obviously constrained interaction paradigm?
I get that this has and needs a constant cellular connection, which ain't free, but $700 + $300~/year is going to make this niche.
For comparison, my main smartphone is cheaper by both measures. You can go buy an iPhone 13 ($600) and hook it up to T-Mobile's Connect plans for $180/year ($15/month, 3.5 GB/month unlimited talk/text).
This frames itself as a smartphone replacement, but realistically it is a rich person's toy.
Your AI hallucinated it's answer about the location of the eclipse. Why was this a) allowed to happen, and b) included in the video without fact-checking?
From a consumer standpoint, I do not see the main advantages of this device over a better designed smartwatch or a smart wristband. The main interface, i.e. the projection onto the hand is a very novel and innovative concept but also blurry and cumbersome. A watch would also technically be able to perform all the function shown here. And personally, with how much we are all in need of a smartphone, might as well make something like a smart bracer. That would give all the space needed for the battery and screen and computation and roll everything into one, phone, watch, AI, etc.
They likely went with this form factor because it allows for the camera to look forward and use that context in its responses. A watch wouldn't easily be able to do this.
You could make a smartwatch with a camera facing up from the screen. The user could bring their palm to their chest (so watchface faces out) to activate it. Then the camera can see forward and the microphone is close to the user.
And then you could do video calls on the same device too.
But how practical is it really? Let's say it's winter, you have your AI Pin on your winter jacket. Then you get inside and take off your jacket naturally. Then you take off your AI Pin and somehow put both parts of it into and onto your sweater? This sounds very cumbersome. A smartwatch just stays on your wrist. You can even take a shower or go swimming with it if you want to. And a smartphone has a screen you can use in many situations – sitting, standing, lying, with the phone on your hand, lying on a table, attached to a stand. All of this is not possible with the AI Pin. It is meant to be attached to your clothing or you can't use its projector. How do you read your emails? How do you read a book? How do you frame a shot? How do you scroll through TikTok? These are all things people do with devices that cost way less than $700 today. And many, many people love to do these things.
This has the added benefit of having some recognizable sign that someone is using the camera... which despite proclamations that "the public" is ready to accept being on camera all the time, I'm not convinced is true when it's someone wearing an overt device pointed at you and possibly recording, but you're not quite sure, all the time.
I for one have no particular desire to be part of your "context" (nor the company's training data set) without knowing it.
If so, they made a big bet. Vision LLMs were literally made this year. Before that, parsing images to get a coherent response is pretty resource intensive and not really reliable at all. Designing the entire device around image capturing for context seems like a very risky approach so I doubt that was their main reason.
I wish them good luck and I wish good luck to the guy who develops the tap which is a similar concept without camera laser just microphone and GPT connection
They say they don't listen for watchwords and only record video on demand. Good.
But that means their hypothetical calorie-count example only works if you actively remember "I am eating, I should tell the pin to record my eating". Same with many other things; the wrist-worn smart devices are useful exactly because they are always recording, they are somewhat ambient.
I understand why they do it, I even applaud them for thinking about privacy, it might even be necessary (cough..Google glass...cough), but I feel they have to walk a tightrope between a rock and a hard place to bring privacy-consciousness and useful features together.
Yeah I was thinking on similar lines when it comes to food tracking. I did a study last year on how food consumption correlates with certain psychological features, but the lack of good data made the study very hard to do.
Self reported data is always annoyingly unreliable.
This product would probably be better than self reporting, but not good enough.
The privacy issue is annoying thing. If it was a on-device ai model things might be easier to accept. But I such a device is further in to the future.
194 comments
[ 4.9 ms ] story [ 244 ms ] threadhttps://vimeo.com/882968794
A lot of the interactions seem to take longer vs me just taking my phone out and quickly doing what I need to do. Its compelling given this is a v1 of the product so it will only get better from here but not completely sold on it just yet.
I'm sure Steve Jobs rolled over in his grave when they lead with product specs over "why buy this."
Makes you wonder about the veracity of the AI (or the accuracy of the demo). Looks like a cool product either way.
pass
“No wake word so it’s not always listening”… you just activate it with your voice?
You press it to activate it. That’s when it starts listening. It’s not always on.
> Google learned a similar lesson in 2018 after it launched Google Clips, a body-worn camera that used algorithms to automatically snap photos. Female users tended to end up with an abundance of cloud shots when they intended to record what was in front of them, because the device was not designed to account for bodies with breasts
This feels like it could have been a classic Silicon Valley scene but I'm going to give them the benefit of the doubt and assume that it was designed at Google's secret Castro office.
> Tapping the Pin and then moving a palm into its field of view activates its laser, which projects images and text onto a user’s hand at a wavelength that produces a blueish-green tinge, a 720p-resolution system Humane calls a Laser Ink Display. Tilting the hand navigates between displayed options and a swatting gesture swipes to a different menu. Users “click” on an option by tapping their thumb and index finger together and close their hand briefly to return to a home screen.
Okay, that's freaking cool. Anyone here willing to take the $700 plunge to test this? I want to know if it's as ridiculous as it sounds.
01:49 > The AI pin privacy chip also protects it from being exploited, which means if it's ever physically tampered with, it will require service from humane to restore operation.
Cause they aren't Apple, just the cheap knock-offs
Why are you trying to do something totally different with what appears to be an obviously constrained interaction paradigm?
- What's the argument for using this instead of a full-fat computer with ambient capabilities? (think iPhone/Siri or Android/Assistant)
- How does Humane expect the $699 price point to develop in the future? Is the intention to bring prices down, or capabilities up?
For comparison, my main smartphone is cheaper by both measures. You can go buy an iPhone 13 ($600) and hook it up to T-Mobile's Connect plans for $180/year ($15/month, 3.5 GB/month unlimited talk/text).
This frames itself as a smartphone replacement, but realistically it is a rich person's toy.
One question is does it work over Wifi when there is no wireless coverage?
Also what would roaming costs do when you travel?
Do we have holograms tech yet?
And then you could do video calls on the same device too.
Anyway, it's cold season now in Europe, and it's interesting how much less useful my Apple watch is when it's almost always covered with a sleeve.
Something that goes on top of your clothes makes sense.
I for one have no particular desire to be part of your "context" (nor the company's training data set) without knowing it.
But, like the Apple Vision Pro (which I don't think is as DOA as this), I'm most curious to know where this goes in 5-10 years.
But that means their hypothetical calorie-count example only works if you actively remember "I am eating, I should tell the pin to record my eating". Same with many other things; the wrist-worn smart devices are useful exactly because they are always recording, they are somewhat ambient.
I understand why they do it, I even applaud them for thinking about privacy, it might even be necessary (cough..Google glass...cough), but I feel they have to walk a tightrope between a rock and a hard place to bring privacy-consciousness and useful features together.
Self reported data is always annoyingly unreliable.
This product would probably be better than self reporting, but not good enough.
The privacy issue is annoying thing. If it was a on-device ai model things might be easier to accept. But I such a device is further in to the future.