As a hardware person, both the Rabbit R1 and the Humane pin are great examples of why I'm bored of today's technology in general. It feels like we have been caught in a cycle of minor spec increases and not much else (except maybe removing features/rights and shoehorning in a subscription) for the majority of technology we interact with day to day. Companies are desperately trying to come up with a new device class that will take off, but they all fail in the same way, nothing is solving real problems that people experience. Who wants to talk to something clipped on their shirt instead of pulling out the phone they already have? You don't see people in public talking to the assistant on their phone very often do you? Even if they worked well, these are likely destined to be niche products.
It feels like we need to wait for the underlying technology to advance before we can get to the next set of interesting products. I'm thinking unobtrusive AR, robotics, self-driving, etc. which are all going to take some time to mature to the point where they are practical.
Probably. I think we've hit a plateau with user interfaces for most connected gear. Sticking a physical scroll button on what is essentially a phone buys me nothing but annoyance. Phones just kind of do what we want and have an acceptable enough interface that the alternatives get in our way.
The Humane Pin added the laser display but ... nobody wants it? This is starting to feel like sticking a spoiler on a phone and pretending it adds value. Maybe VCs are impressed by this stuff, I don't know.
If there were a magic version of this it would have the model and processing onboard. That's obviously extremely cost prohibitive right now, but that's what it's going to take to create a real AI-powered assistant:
* I need it to work without network latency.
* I need multiple forms of input/output depending on the context.
* I don't need Teenage Engineering's usual form-over-function design. I need the form to be out of my way most of the time.
It's great that people are experimenting in this space. It's less great that people are getting multiple millions in funding and selling a phone and web service as an "AI device." This will hurt future development.
Strange thing is, it's actually a touchscreen. But the only thing that they let you use it for is the input keyboard. Saw this in a review on YouTube, I don't have one
This is the best two-sentence summary I've seen so far. People tend to forget just how many wacky flip phones and PDA-type devices had to come and go before the touchscreen smartphone as we know it came into being. Which is what these new AI gadgets remind me of.
However, the difference I see now versus 20 years ago is that companies today are just trying to get the most cash and deliver the cheapest product (and pocket the difference), rather than stick with their product and reinvest into R&D.
Same thing for ultra wide fov ar glasses, I'd kill for those.
Fundamentally it's the model of the boss giving orders to its secretary, and people dream of it as some pinnacle of evolution (perhaps it is, I don't have a paid secretary to assess the point). That's actually pretty close to the concierge services offered by phone for instance, and some people with enough money seem to enjoy it, so why not (try to replicate it, even if it miserably fails for a while) ?
Devices will stay the same until we invent new technologies.
We've certainly progressed semiconductors a lot, but it's still just the same old process for the most part. I think it's areas other than semiconductors that will drive innovation, new optics/displays (foldable, holographic, wide fov ar on a contact), solid state batteries, etc.
Would sure be nice to have properly 3d semiconductors tho. As many transistors as there are across x and y, there should be across z too.
I think it shows a lack of creativity to say this, because existing technologies can definitely be used in novel ways to create value. Perhaps we just love to imagine new technologies as engineers.
But for example, the iPhone was nothing "new" in terms of hard tech but it definitely changed the world.
The iPhone is a great example of technology catching up to make a product viable. Between phone networks, display technology (especially touchscreens), SoC efficiency, etc. the iPhone only became possible when these could integrate together in a practical package along with the vision to bring it to market.
I think people do seem generally annoyed by the scroll wheel, but I think that's just a bad implementation in this case. I don't think that means that all physical controls (meaning: not a touch screen) are just pointless gimmicks. It's a bit myopic to think we've hit some perfect un-challengable UI paradigm with flickable scrolling and sheets of software buttons and fields that you tap.
> The remote-access computer transponder called the "joymaker" is your most valuable single possession in your new life. If you can imagine a combination of telephone, credit card, alarm clock, pocket bar, reference library, and full-time secretary, you will have sketched some of the functions provided by your joymaker.
(https://en.wikipedia.org/wiki/The_Age_of_the_Pussyfoot)
Interestingly, the author came up with the idea after he got acquainted with the details of time-sharing mainframes common of his time. Consequently, the devices themselves are actually mostly hollow, just processing input and output, and all actual smarts are done on central servers elsewhere - which becomes a major plot point eventually.
It even got the social aspects of it - i.e. it being a standard device that people carry around them everywhere, it's used for most day-to-day communication etc. Just about the only thing about it that feels off from a modern perspective is their size - they are large enough to require carrying them on the belt (like 90s cellphones) rather than in the pocket.
Now, one thing that is still missing compared to the model is the "pocket bar" part (which in the novel also includes recreational drugs). Now that would actually justify a new device like Rabbit. ~
This doesn't convince the tech fetishists. In fact, I think te's contribution to music is very limited, even harmful, especially when I see the latest Yamaha even imitating te, which feels a bit funny and ridiculous. We need real innovation and democratic pricing.
By the way, if you care, you can learn about the history of rabbit’s founder. Let's just say, in certain circles, this is a recognized liar. So I’m not surprised at all when it was said a few days ago that Rabbit stole everyone’s passwords.
This is just the same guy, over and over. If he isn't too old, I bet he was into dropshipping too. Always on whatever new grift, always ready to start selling shovels to gold miners. I wish people were smarter about this stuff.
Update: They've barely tightened up, now the only missing piece is the OS build fingerprint. https://twitter.com/uwukko/status/1785626783900930447/
As for OP1, packaging a tape effect when it can't produce good sound quality is a fraud. Sampling from FM? Check Klaus Schulze.. Music startups can certainly innovate, such as Roli (Seaboard and Rise), moog. make noise. Open source monome. and, to my sorrow, Émilie Gillet (Mutable Instruments). TE's innovation in music itself is far inferior to them. If you praise te, I think it is unfair to other people who are obsessed with technological innovation.
I have always felt that TE is a electronics company with no innovation in hardcore technology. The appearance and packaging are different from previous ones, but there is not enough practical integration. Of course, this is just my personal opinion. If you disagree, you can write down your opinion. Thank you.
I bought an R1 during the first pre-order and it arrived a couple days ago. In it's current form, it's not the AI from the movie "Her" that's going to manage your life for you. It's neat. It's a cool little toy that does interesting things and has potential. It definitely needs work: before today's software update, I couldn't type all the ASCII characters in my home's Wi-Fi password. I still like playing with it and I hope it improves.
I don’t blame Rabbit for wanting to control the hardware they run on.
The idea sounds I mean there’s Alexa (which is way cheaper) and shipping an AOSP with a Chinese vendor hardware is indeed manageable — but do we really want to have a ChatGPT-only 24/7? I highly doubt that, at least for me. I’m probably only chronically online for the social aspects of that, but had my interactions been only to an llm agent, I don’t think I’ll be using it that much.
I like to think 5 or 10 years later in the future and see how us and the newer generations would interact with technology but I find it difficult to envision that would be rabbit r1 device
> We didn’t bother testing out any other functionality, such as Spotify integration, Vision, etc., but we wouldn’t be surprised if some of them didn’t work. After all, the Rabbit R1’s launcher app is intended to be preinstalled in the firmware and be granted several privileged, system-level permissions — only some of which we were able to grant — so some of the functions would likely fail if we tried.
I mean, even the hardware that Meta and Amazon produce are built on top of Android / Linux variants, so I would be surprised if smaller players were writing their own custom OS these days.
There is nothing worthy of criticism in the fact they choose android as a base.
If it's a portable device with connectivity and a battery Android is the only choice if you want to ship it on any reasonable timeline. Even going a Linux route you'll burn months just getting power management and wifi/cellular working to an acceptable (not even great) level, so much struggle with poor to no driver support.
I'd love it if there was a better Linux solution because Android certainly isn't without some pitfalls but from my experience Android is the only viable way to ship a device like this today, not even Linux is a viable option.
The dumb part here is a company getting attention for making a virtual assistant just because they added "AI".
It mostly does simple commands and for the rest kicks you to web search results.
Apple surely understand the opportunity here, maybe their view is LLMs are just too variable / hallucinate too much to go all-in on it?
So much “disruption” with no clear product use case. AI is the new dotcom boom. 190 percent hype. 10 percent actual implementation.
The startups have expectation to capitalize on testing with early adopters and naive consumers. The big companies fake their demos for likes and stock prices.
What a conundrum. We need people with real skills and clear vision. We need Skunk Works team quality to achieve something of substance.
Ovens that leak gas, cars where the wheel comes off, games where you only get 10 minutes in etc.
Personally, I prefer how the world actually works where you have user testing sessions, beta programs, quality management etc.
There was a time when here knowledge and experience did not hinder the new ideas at all. My entire career was a result of this process in early 2008.
Today we live in inflated VC realm of promises, big statements and low delivery. It is not about demanding on a side lines.
Some of us still produce quality by following the proven methodology from the past. To push to market R&D projects and demand applause is pathetic. And no. There is no place for PC in this. It is about time to wake up from the AI Utopia. And the cleansing process is going on as we speak. Corporations reached the limitation of participation trophies and DEI agenda.
The problem is marketing and severely exaggerating what these devices are and can do thanks to all of the noise about "AI" recently.
This is a clear miscommunication (or intentional miscommunication) internally about what this thing actually is.
You mean the kind where plane literally leaks fuel mid flight? I think they're already there
The A-12/SR-71 was leaking fuel while it was sitting on the apron/ramp, because due to the insane heating that occurs at Mach 3 you have to compensate the heat contraction of your tanks and everything else. So while they do leak on the ground, as soon as they start getting anywhere close to operating speed and the structure thus heats up, everything seals shut. No leaking midflight. The Skunk Works director Ben Rich (who took over after the original director Kelly Johnson stepped down) wrote a book called "Skunk Works: A Personal Memoir of My Years at Lockheed", it's a good read.
And towards that end, it seems these devices are underpowered, drain batteries faster, require an Internet trip to do anything and that anything could often easily be hallucinations.
And that's where these AI devices don't seem to have a chance to ever even a dedicated user base.
It would be much more appealing to me as a watch. Since it can at least tell time, and I don't need to carry it around in my pocket, which I'm already carrying a phone and a wallet. I'll need a 3rd pocket for this?
Much quicker to access for quick question, take voice notes too.
So I guess all depends on how next evolution of these devices are going to be, and cannot see it's replacing the mobile anytime soon. "More devices" is never the answer.
I don’t know how they can possibly make money with this but i’m looking forward to having a new toy on my desk and I really like the teenage engineering vibes of it.
Definitely not hooking it up to Spotify or any personal accounts after hearing how they handle security.
I want more companies to try hardware!
To quote the title,
> a thing that should just be an app, is just an Android app
Except the company doesn't own the hardware for distribution by being an Android app! People want developers to be subservient and taxed forever, as if Google and Apple own all of non-desktop computing. It's an unfortunate place we've arrived at.
We need many more hardware options. The cellphone duopoly is harming and taxing innovation.
From the first looks of it those will just end up in the drawers or landfill in no time which is a little sad.
The whole device seems to be “it’s Siri but as a standalone device” and since you still have to take your phone with you, it seems to provide no value.
Smartwatches are exactly that, and some people buy those.
That said long term I want them or others to succeed.
The last thing I want (and most others should want) is a world where only Apple and Google are the only ones hosting mobile AI products.
As any phone OS integrated Apple AI or Google AI will beat out any shipped apps store AI app long term.
If a new hardware form factor is the way to break that duopoly then I wish them all the best.
I think the thing people should want is on-device AI.
Because I honestly don't see the advantage in terms of privacy or performance to have Rabbit R1 proxy my requests to chatgpt or other cloud LLMs... At that point I might as well use Google AI or Apple AI instead of adding 2+ parties to my private AI use.
The only real allure of a separate device, for me, is isolating the bot from my data. I don't want it reading my emails and notifications and who knows what else. I suppose you could also market this for kids but they didn't go that route.
Purely from a hardware perspective, a phone has all the sensors, chips etc needed for AI applications, so I'm puzzled why Humane and Rabbit thought that a standalone device was a better idea.
99.9% of the HW projects that have a modestly complex display/networking need run on Android. It's a no-brainer. OEM the HW from China, even if moderately custom and they can get you something for cheap.
The only thing Android sucks at is native support for keyboard interaction, anything big screen or touch screen related may as well be presumed Android until proven otherwise.
There's one exception, which is Samsung, who is still pushing Tizen to its products, though their smart watches switched from Tizen to Android not too long ago.
Bonus points because then they get to re-implement Spotify integrations et al instead of using an existing APK.
I don’t really care that it runs Android, but it seems like it runs on Android _and_ locks you out of said OS which I don’t love. By all means, build on Android, but then let people use Android. Let me slap a SIM in it and make calls, or install a stupid clock face or something.
I’d say the same thing of most appliances. I don’t care what OS it’s built on, but I want access to that OS as much as possible. Eg I liked that Bluecoat devices have a custom SSH shell for managing them, but you can drop into a regular bash shell with the root password. Can’t say that I used it, but it was reassuring to know that I could get vi or something if their shell fell apart.
Personally I figured it was running Android but likely a heavily modified fork.
Especially after how many times they seem to have buckled down on it not being possible as just an app.
(Unless I am misunderstanding and it is indeed a fork and not just an App? ).
Using Android for this makes a ton of sense. Lying about it does not.
That being said, their marketing terminology of "a revolutionary AI-based OS" is what's more problematic. If they had just mentioned it was an android platform none of this would have been shocking to anyone.
There's never been a company I've personally seen that is both truly innovative that also uses the term "operating system" to mean something other than an actual operating system.
Rabbit R1: Barely Reviewable [video]
It seems they want the compute to run on device, but that doesn't seem to be the case right now[0]. I don't know the specs of the device, but pretty sure it'd have to cost more than $200 to run an LLM locally.
edit: 128 gigs of storage and four gigs of RAM. For llama3-8b (the smallest llama3 model) you need at least 8GB of RAM.
0 - https://www.theverge.com/2024/1/18/24043490/rabbit-r1-ai-per...
* Everyone else can buy it when the pre orders are delivered
* Due to monetary constraints, regular orders are actually shipped before pre-orders
* The device is nothing like it was promised
* The device doesn't exist
The intentions are good (funding without VCs) but I rather the VCs lose money than consumers
I can't judge them too much considering how much actual grownup money I've spent on Legos...
In 5 years this thing will be poisoning some small trash picking child in Africa somewhere.
based on that appeal, i’d say you’re probably correct in that assessment.
So this was immediately obvious that it was running Android. It is just that this was a nice and perfectly packaged scam, but not as expensive scam like the Humane AI Pin.
and yes. Humane is also using Android for their AI Pin devices. Unsurprisingly. [1]
The Playstation Portal is another good example, it's a single-purpose device just for streaming games from a PS5 but it runs full blown Android, locked down so you can't use it for anything else.
The Rabbit thing is literally just a terrible phone.
What led you to believe it was rushed and/or unfinished?
I agree with the sentiment that it's a _bad_ product with _bad_ choices and _bad_ strategy. It was likely caused by shitty ZIRP-era "MVP product without the 'viable'" mentality, rather than caused by rushing an unfinished product.
You can just use Chat GPT on your phone and get 90% of the same experience. But then it's a matter of branding. To some it's cooler to use a toy like this. Sorta like how Beats headphones are often beaten by headphones costing a 1/3rd.
Edit: At 200$ I'm not mad, it's ultimately a toy. Much better than the AI pin costing 700$ + 25$ a month.
But I can't see how they can sell this device without a monthly subscription? Even if you don't make many AI queries, you're still consuming resources on their "rabbithole" web services. If the company behind Rabbit closes down, I'm guessing the devices will become near-useless? Although, knowing that it runs on Android gives hope for hackers to extend and modify the devices.
It's also a lot easier to stomach a beautiful shiny teenage engineering toy that doesn't do much if you don't have to pay every month.
But now we can see that it is just a weird shaped Android phone hard-coded to only run one app. It shatters the illusion that the R1 hardware was worth the $200 price tag, as the software under-the-hood could be equally if not more functional if it were just distributed as an app instead of as a piece of hardware.
It would be different if the hardware had special sensors and processors that help maximize its functionality. There would be no issue for AOSP to be the base OS - in fact I would agree with your assessment that it was a good decision that should e celebrated. The hardware itself having novel capabilities would be the important aspect, not the OS they chose to build upon.
Absent any special hardware functionality, it throws the entire Rabbit business model into question. They aren't charging licensing fees, and the service itself seems to be a lifetime subscription, so they aren't planning on making money selling software or SaaS services. The profit-making part of this business is selling hardware - and the hardware is just a worse version of the phone you already have...
The software also seems barely useful. Pretty much just a bare-bones implementation of speech-to-text, GPT via REST API, and text-to-speech, with a handful of basic integrations.
So once they've sold you the hardware, do they have any reason at all to improve the software or the service's offerings? If the R1 hardware can't do anything your phone can't, can they really compete with an app (even a paid app) that does the same thing on your almost certainly higher-spec'd smartphone?
Given the founders' history of grifting, I am going to guess that the R1 hardware costs Rabbit <$50, and that the company will soon disappear, taking down any expensive cloud-based AI functionality with them, as soon as they feel they've sold enough units to line their pockets with the profits, and the R1 owners will be left with a brick.
Ever since I saw that hardware, I've been wondering if it could be repurposed as a tiny phone with smart features, or as a 'connected' MP3 player. Pity it does not have a headphone jack.
I remember a post, here, some time ago, that was an explanation about why someone wasn’t going to be making a webcam, even though they had a great idea.
The post talked about all the various details and hurdles, involved in sourcing parts, making, promoting, delivering, and supporting the product.
It basically said it wasn’t worth it.
Backing a hardware project requires substantially deeper pockets than an app.
Sometimes, a good hardware wrapper could make all the difference, but it’s a really big deal.
Followed by a demo of someone copying the APK and the whole thing more or less working... I think Lyu forgot that the statement was for androidauthority.com (where people who understand android hang out) and not for his 80-year-old-uncle...
Isn't this how Android works though? The interface you boot into is an APK that has some more entitlements and hits some APIs. Like how should they have made it?
(via https://news.ycombinator.com/item?id=40223247, but we merged that thread hither)
And end up producing their own custom interface which has been universally panned.
Is it relevant whatsoever? Is the product relevant whatsoever? The answer to both seems categorically no at this point.
I really don't understand why this device is getting attention while hundreds of hardware products launched each year go mostly unnoticed. What am I missing here?
No, in that case they would deserve to be called incompetent.
It would be much weirder to learn they built an entire tech stack specifically for the device and that it was technically impossible to port it.
A.I. itself is amazing... but still kind of half baked
AI apps, particularly voice assistants, are designed to give you text and data via a screen. I can't tell you how many times I've asked a perfectly simple question and my android assistant responded, infuriatingly, "here's what I found on the web" or the dreaded "please unlock your phone" prompt when it relates to anything remotely personal.
If I wanted a web browser experience or to find the answer on my own, or to follow up with focusing my attention and interacting with a digital keyboard, I would have!
The rabbit interaction is for a purely responsive 2024 AI experience, which doesn't try to shovel me back to the 90's at earliest convenience.
the real value is in how it functions in the life of the users. if they put a new llm based shell on top and built a new app ecosystem for it that makes users happy, then they've done something useful!
Rabbit on the other hand...don't give them your passwords!
Saying this as someone who won't waste money on the Humane pin but ordered an R1 for fun. Wish Humane got the price point (and a lot of other things) right. Like focusing on device latency and integrating Spotify instead of trying to boil the ocean and integrating only Tidal.
