I mean, the CEO literally says if you read between the lines, “this product could exist as an app, but if we did it as an app, someone else could too, and then we couldn’t force everyone to use us”
Every company want monopoly duh, the only reason Microsoft isn't one in multiple industries is because they can't just buy everything without prior agreement with countries and such@@ReigoVassal
If we only had some rectangular device, with a camera... No, a camera for forwards and one backwards... With a speaker AND a microphone... Yeah... And a screen... No... No wait... A TOUCHSCREEN!!! And can communicate ✨ wirelessly ✨ With sensors that can detect shaking and stuff... And runs android! Where else are you going to get all that?!
@@tezcanaslan2877oh mate.. at this point we just gotta understand the old mate craig is leagues above everyone else it's not surprising they have made it! AND to top it all off a cool blue disc with the silver ring around it
That would have been a better plan. Market it as a hybrid dumb phone in the sense that it is not optimized for social media and content consumption yet could cover basic phone features.
That's a great idea Many people try to download minimalistic apps towards social media or add ad blockers where instead they could just have a phone without any pre-built apps and social media Like the iPod for kids without the sim card option and without social media @@hrdcpy
The trend that really bothers me is people treating companies like they're their friends. "Oh it's okay it's their first product, give them some slack". NO. As a customer that bought a product that promises functionality you are in your right to expect that functionality.
@@dingus1720 LMAOOOOOOOO they don't even have the money for gpt4 and they market it with complex words to make it look like it is powerful this is truly a snake oil salesman moment
I really hope people realize the company that didn't secure their own server endpoints probably is not good at protecting any of the data you are sending it.
to be fair, I just don't know what that is. I think people tend to forget there are a lot of us who just don't know these things. for the most part I just kind of have to have faith when I'm dealing with technology because I'm not going to understand what everything means. granted I never believed that this was at all realistic but even if I did and I tried to do my due diligence there's a good chance I wouldn't know to look for that. I don't know anything about steam or windows but I use them.
@@moamber1 I do not trust Microsoft or Google but to function I have to utilize their services. As for Facebook Instagram and the other wonderful social media giants I have not had them for years. They are just ad machines. Choose wisely what you give your information to is the point.
@@moamber1 I don't trust them either but at least a breach is less likely compared to some startup with "questionable" security practices potentially leaking sensitive info. At least google protects the more sensitive data like payment info from theft/breaches properly... 😅
@@dizzylilthingit takes time, just start reading, watching and learning from people. I'll help you understand what an end point is. It's semi self explanatory. Say you want to send data in the real world. You can input information onto paper and store that data in a letter. Data on the Internet works like that. Your computer is making something like a letter with your info in it and sending it to the company. The company will receive it also in a similar way. In the real world you send that letter and the company will receive it wherever their mail is delivered, that's an end point in real life. The point of delivery is the end point aka the mail box hole. Digitally the company also has a place where the data must go which is like a digital mailbox hole, that's the end point. TLDR the end point is where the company receives data from you or your devices. It's like a hole in the door that lets you send and receive stuff to and from them without. The op is saying that they didn't use any security to protect that delivery hole. I hope that helps. Everyone has to start somewhere. Think of computer knowledge like a painter painting an image. He doesn't just spit it out like a printer, humans don't work like that and you would be overwhelmed if you tried. Instead they focus on one small part and then expand from there, they start with nothing, but end with everything...
@@ilonachan Agree. It's moreso the fact that they're pretending it's different. Like it's literally just a small android box that can only run a single app, and if wanna run more than 1. Prepared to get harassed... Just call it what it is.
Feel he said it best with going over the Marquis stuff " we want something that hasn't been done before". Asides from accepting criticism and having good PR. Humane at least has a reason to need their hardware to get the full experience ( assuming things worked as intended). You ( at least not yet) can't get a laser projection working from your phone, and as a wearable it should save you time without having to go through a screen, and if you need a screen for something lol.
They weren't really being shamed for using Android in the first place though. They were being shamed for lying about it. They told a bunch of stupid lies about how this couldn't just be an Android app, when it obviously could, and then it turned out that, surprise, surprise: it's an app.
Tony Fadell, the inventor of the iPod said in his book that he won't even look at the hardware of a startup until he's convinced it can't be done any other way.
werent there plenty of options when it came to music players back then. seems like there were plenty of other ways. that quote kinda comes off snobbish ngl
@@wojtek-33 Creative wasn't exactly a startup. I'm no Apple fan, but I think his point still stands; hardware startups are most likely in search of a problem.
@@pikkyuukyuun4741There were options, but none of them could do what the iPod could do. The scroll wheel, the screen and the software all made it easier to put a lot more music on a portable device. Eventually people were storing their entire library of music on that one device... That's not even the iPod touch, that's just the iPod...
The same critique would still apply, if they built the Rabbit R1 on top of embedded Linux, or if they built "RabbitOS" from scratch entirely. It could still just be an Android App, there's no reason to create extra e-waste for this.
After the recent Coffeezilla video exposing the "LAM AI" as just some ChatGPT scripts along with Playwright, an existing non-AI browser scripting tool, this looks even worse on their part. What a disaster.
I had the misfortune of working for a manager who was very much like this CEO. He had a background in technical sales and was 'that' kind of salesman... you know the ones who only care about closing a sale... with not a care about whether what they sold is deliverable. Some people seem to constantly be recreating a reality that was good for them and then gaslighting everyone and themselves into believing it. The ones you mean in employment have amazing manipulation skills, seem to be shameless and incapable of learning from their mistakes....pp In the case of my manager, he had got the company to invest millions into an app, putting the business into serious financial problems, for an app that was a wrapper over opensource and barely better than a proof of concept prototype.... the lessons i remembered from this, never work for a manager or a small company owned by someone with a sales background.... And most importantly buy anything from a business that engages in gaslighting marketing techniques like rabbit...
My life lesson is : never trust sales person, they're paid to lie. I only trust word that come from the engineers (whenever I buy product I try to find the engineering manuals for details or even talk to the engineers).
I suspect the real reasons for releasing the hardware and not the app is to: 1. Get more money and hype from a physical AI device. 2. Avoid paying 30% of subscriptions / usage fees once they have to pay for their clunky LAM system.
Good video! We managed to get the updated app working with valid details now, so yeah they still can't do proper server restriction and as long as they don't charge a subscription or completely revamp the os, there will always be ways around this.
About the first point. I think that people arent making fun of it because haha "funny crappy android app they are so bad" but because this whole device is useless and easily could have been a paid app in your phone. Especially considering all its weaknesses (like being not performant on this weak hardware), this whole idea of an extra device just seems pointless. Now, you made the example of e.g. your camera also running android. Or the human AI Pin. But for those two, I'd argue it does make sense having an extra device. Because they all have something your phone doesnt. E.g. an extra OP camera with a lens and such or an projector, etc.
Exactly. The main problem is that the Rabbit app works with 100% compatibility with an Android phone. This is the reason the R1 is a laughing stock If your app requires specialized hardware (or unique enough that a phone doesn't have it) to get it working, then it doesn't matter even if it runs Android. Like an Android TV box, I can extract all the apks and install it on a phone. It will run, but the experience sucks because your phone doesn't have an HDMI port, nor does it come with a TV remote. By the way I just rolled my eyes with this LAM bullshit. I bet this is just some system message telling the AI what actions it can do, with the actions have to be manually coded in. Something the regular ChatGPT 4 can already do.
@@rigen97 I highly doubt they are doing that at all. If they are doing this right now then why is it only right now limited to 4 “apps”? If it’s truly universal it should be able to use a lot more services than just those 4. Heck, it should be able to google search a timer to add a timer functionality. No, I strongly believe this is just some system message for GPT that triggers Spotify’s API, or Midjourney’s API, etc.
@@shamringo7438 shrug, that's what they _wanted_ it to be. they clearly talk big. we knew from their past project that they knew how to make things sounds plausible and then just...not do it.
Electron is not a good thing. Apps leaking a lot of memory (especially commit size, which is a huge issue on Windows), and separate runtimes with developers of each individual application being responsible for updates, which means that a vulnerability in Electron is a huge issue. This is a problem with browsers too, of course, but they are a lot better at pushing out updates when ACE vulnerabilities are found. This gets extra scary when people write chat applications, where you can push exploits without user interaction, and use Electron. /IT security specialist who is stuck on Windows and sick of having to pay for RAM sticks, just to literally be idle for "committed memory"
the only good thing ive ever found of electron without ever being in web development is that is for some home automation or game that has html, launching via electron is much easier.
A correction. OS/2 was not an experimental joint project it was a multitasking operating system that also hosted Win 3.1 and DOS applications natively. While it did die off it was mostly due to cost as you were also buying a Windows licence. After 4 major versions it saw lots of use on non-consumer systems where stability was important.
Still even sees use today. I encountered POS systems using it in 2016 which are probably still running today and now I see industrial machinery running it in 2024. But of course I see industrial machinery running CP/M in 2024 as well so there's tons of old stuff out there. Somewhere in one of the stairwells of my father's old office building, on the 5th floor, there is an Altair 8800 running a rooftop weather station that has been operational since the 1970s (the campus has its own power plant and is not on grid power, so the 8800 has presumably been running since the 1970s).
I‘ve been following „AI“ developments for a while now and there were some really cool projects i tried. One i vividly remember was a project from Facebook Labs where it could take a single picture and make a 3D model from it. Impressive stuff! Nowadays every „AI“ is just the same API from OpenAI with a fancy skin on it. It’s disappointing and frustrating. With the Facebook Model thingy it was an actual engineer behind it who put his heart into the project. Actually when i posted the results i got from running his algorithm to twitter he liked my post pretty much right away. I want to see cool innovative projects like that again, not everyone packaging an API and calling it a day
Yeah ikr? There's so many exciting things AI can do, contrary to what apparently everyone is thinking, AI doesn't mean to just get some text from ChatGPT lol
Contrary to what is commonly repeated, capitalism doesn't reward innovation. If you want to see more passion projects, join a communist party and help build an economy that will give people enough free time to pursue their passion projects.
@@justinwatson1510 i mean yeah, you are right on that one. sadly there is no communist party in my country, and it doesn't seem we'll get a UBI any time soon
Thanks for the video… Personally I got into deep learning 11 years ago, focusing a huge part of my life on solving really hard engineering problems, and it is unfathomable how many low effort grifts these days seem to be showered in VC money…
When I saw this video on Sunday, I immediately submitted for an RMA to return my Rabbit R1. I received a response Monday morning stating: "Your email is important to us, and we want to ensure that we address your inquiry promptly. Our team is working diligently to provide you with the assistance you need." ... As of late Wednesday, no RMA or any other information has been provided. IMO it should not take longer than 24 hours to provide an RMA to return a device. It will be interesting to see how long it takes or if I will get one at all.
This gets deep. In 2018 Bidu and Teenage Engineering released a similar project with Jesse Lyu (working at Bidu). Look up The Raven H. It’s a real thing!
There is no shame in using AOSP for the underlying architecture for your new product. It has software for the screen, wireless connections, touch...why bother reinventing the wheel?
Then why reinventing the smartphone ? I think it's the point of the video: if it run on Android, on a cubic object with a touchscreen, then why not make an app ?
The ridiculous thing is that there is actually a potential market for an AI assistant device that's not an app in a phone - medical alert devices. A small, lightweight device that doesn't have a million apps running in the background draining the battery, that could perform basic functions like reading labels on meds, monitoring vitals & alerting when to take meds, and a panic button to call 911. The only problem is that it would have to be a pretty bulletproof system, and that would all require FAR more actual effort than someone like Jesse "I threaten to sue 14-year-olds" Lyu could cope with.
- technology gets genuinely hyped (computers, internet, lab-on-chip, crypto, AI) - scammers arrive first, (because it is the easiest to pivot if you are not actually doing anything useful) - we are swimming in poop and can barely see the actual real good products This cycle is so bad for crypto that I literally cannot name anything that is not somehow scammy. Elizabeth Holmes made lab-on-chip infamous, but never actually (mis)used the technology. AI scams are just beginning, at least there are some products which are usable, we should mentally prepare for a big avalanche of AI scams.
Nah, Holmes at least wanted to make something that benefited people, her screw up was when she doubled down instead of just shutting the company down when research doesn't pan out like the rest of healthcare / biotech start ups. Web3 and NFTs are just morons chasing trends against business fundamentals, it's the same nonsense as the Dotcom bust when people buy internet stocks even if the companies had nothing to do with the internet.
I do understand why they went the hardware route as yes, an app can't lock people in for apple or google to replace with their own. But it's just a bad idea especially if it's that fragile it should have never been considered to be made
There is nothing stopping apple or Google from doing that now, in fact they already have their own AI assistants. Soon those AI assistants are going to get big updates...
The original iPhone didn't even have a flash that could be used as a flashlight - the original flashlight app just turned the screen white. Which you could do with the web browser.
11:37 - for the non-technical people watching, the researchers and experts and hackers have actually become much better at reading this compiled, obfuscated bundle of letters and numbers and turning it back into actual code but still, submitting compiled code to an app store is not even a moderately significant problem in the grand scheme of things... whenever a user recieves software, whether it's through an app store or through firmware installed on hardware or otherwise, they'll receive that compiled package
I'm not a genius. I will never develop a genius device, nor a genius app. That's why I don't seel genius promises to other people. People nowadays should: a) Accept their limitations b) Take responsibility for their actions c) Be held accountable for their mistakes Scamming people shouldn't be so easy.
This man does not remember the days of the flashlight app setting your brightness to max and filling your screen with white because hardware flashlights didn't exist back then.
I work at a robotics automation scale-up and one of the best things we did was move away from manufacturing our own hardware product. The sheer monetary and personnel overhead required to start manufacturing even the simplest product, heavily outweighs the margins on selling the product itself. I could speak at length on those tweets from the CEO but this is a comments section and nobody actually cares. This is a CEO that has been sold a dream by a Product Manager that doesn't own a smartphone.
Honestly if what comes from these junk AI in a box devices is Apple pulling their finger out and making Siri decent then it will have been a positive in the end.
That's not why apple is working on upgrading siri, it's because Microsoft was working on upgrading cortona aka co-piolat. Microsoft has been working for a long time on this and working with partners like Intel, AMD and Qualicom...
remember: this community will encourage you to try and build new things and that it is okay if you fail, but if you fail, they'll make sure you regret it dearly by berating you in a 30 minute highlight reel video of all your mistakes
everyone is just being so wholesome and building their community, right? it's all for views guys ... it's all for views ... and they'll get you riled up and teach you to not try to break the mould by making an example of someone else in their videos ... toxic
@@tomatoslavnobody makes videos like this about the poke walker released for the Nintendo DS, that square smartwatch that got bought by Fitbit, the micro PC community, the raspberry Pi community, dozens of communities are propped up around actual good hardware. this wasn't people trying something and failing, this was people selling a bad version of Siri on a $200 device that has next to no function
Jesse's reply sounds like chatGPT, just that he made all of it lowercase so that it "looks" like it's not, together with the "its" and "it's" mistypes.
I saw some dude hyping this up a couple weeks before it came out. And immediately I thought this is going to fail Google lens is a far superior product And it's free
The thing that had me worried pretty much right from the start is when Marques mentioned there is no subscription in contrast to the Humane Pin. Given that AI has running costs, how can that ever be a sustainable business model if you're not one of the big ones and this is just a side project?
I'm impressed they've managed to jank their way to getting GPT-4 to talk to a back end on a remote device. But due to the nature of prompt engineering with extremely large LLMs like GPT, the moment OpenAI moves on to GPT-4.5 or whatever, it's going to break possibly forever. Worse, they implied the reason the Rabbit r1 even needed a dedicated device was to make use of a Qualcomm chip with it's own tensor processing unit to run some component of their app locally, which now the specs and app are out there and working on other devices, was obviously a lie. Worse still, the hardware there is about $80/device, after losses on shipping, tax, and etcetera, there's maybe $25 of income left on their end, how are they paying the fees for GPT-4 access for more than a year for every one of these little Android doodads? I smell an incoming rug pull again.
Canceled my preorder, I agree, this needs to be an app and not a dedicated hardware device. One in which I have to carry, and keep charged, etc. This combined with their shitty behavior, I am done.
Jesse should have said only one thing: We made a device because we wanted to. Not only the profit margins are higher, but we also wanted to minimize the amount of time people spend on their phones. The R1 is a cute little rectangular that can do lots of things and could help with your mental health by minimizing smartphone use.
Tbh the thing with it being android based is just funny and kinda ironic *because* everyone has been saying for a while that these AI hardware devices could just be an app. It’s not that surprising it’s built off base android. Then bundled with the fact the app they built is actually (to an extent) compatible with android phones with not too much tweaking - is also pretty funny.
At 14:49 the "even if we built RabbitOS as an app..." is discussed. Theo reacts to it by giving an Uber example. To be fair to Jesse, he has a point. He's basically saying: "If RabbitR1 was an app, rather than a separate minimalist hardware product, the experience would be way worse, as in comparable to Siri". What he means is that as an app, you would have to launch it, be interrupted by the notifications from other apps on your phone, etc. He's arguing that by giving you a hardware thing without notifications the UX is better as the assistant could be immediately available and not get its volume reduced every time a notification comes in etc, distracting you from the assistant / intended UX. All other points made by Theo are valid, just being fair to Jesse here
they didn't want to make an app because they knew how badly they would be roasted for it and likely would also have been denied from appstores for data collection policies
That Jesse Lyu thread is basically: "if we me make it an app no investor or customer will pay us an insane amount of money and we don't want that. Of course that's bad to admit outloud that it's basically a scam so here's a bunch of barely thought out excuses."
The moment I saw it I said so it's just a GUI app and they did the hardware to stand out and raise funds. It has absolutely no other reason whatsoever.
I love that they are like trying to hide behind the thought of companies stealing their code to integrate into their OSs, only for all the companies to be midway through doing the job better than they ever will by doing it directly on the phones all on their own... not that anyone will really use it anyway.
I want to rephrase the first point here: An App is hard to build but easy to copy. To make a good App you need two versions and maintaining both well is a challenging task, as stated. But your competitor can make a crappy App and claim to do everything you can do and undercut you in price to a point that users are not trying your App anymore. They would get the competitor App, have a bad experience and go, 'well, that's bad, why should I spend even more money giving this other one a shot?' Result: in the end, your great idea fails and the bad actor gets more money out of it than you.
Obviously both of these "AI in a box" projects are a way to make a quick buck on AI hype. They don't need to be a separate hardware, but if they release as an app they would have to compete with millions of other apps and they just would get overshadowed by Siri or Google voice assistant. And they know that they don't have anything worth of value to offer with their service, because the app model is more sustainable in a long run and if they were confident in their service they would release it in as many forms as possible (like app, web, PC client program, etc) like any other big service does.
that flashlight he was referring to was literally a white screen with the brightness on max. Because the first iPhone didn't have a flash led on the phone to function as a flashlight!
~ 29:00 Imagine trying to find a video to back up your point that MKBHD wouldn't have gotten here without iteration.. and choosing the video that is probably the most opposite your point possible because it highlights a thing no one talked about, it covers something that no one would've even paid attention to if he didn't. It's like the exact opposite of a valid example.
A typical example is that when I make an online purchase with my credit card I get a notification from my bank requiring me to approve it, then I get a notification from the online store which arrives through the mail app, and finally I get another notification from my bank that the amount has been charged to my card. What I find even more ridiculous is that this bespoke device of theirs is so expensive for what amounts to a very crippled phone. I can buy a decent Android phone for $200. I think his English is so bad because he's actually from China and it appears his funding has primarily come from Chinese investors. He wouldn't be the first to bamboozle venture capitalists though. OS/2 Warp 3 and 4 were actually really good.
ToS barely changes - it's more the user end, rather then the programmer end - and while i'm not sure if i'm CORRECT i'm. going by the fact that Google's ToS even for colab BARELY if ever changes - but the Acceptable Use Policy gets vague and unvague -- The laughable points that you could in some ways translate most of it to "You can't do anything on colab until you just pay us all of your money and still can't do anything" -- is why I laugh when people are saying "ToS keeps changing" -- A website's one could change, but a corporate one rarely changes.
the first thing i thought when i saw the Rabbit was 'this should be an app' and then 'i don't want more crap in my pocket'. You have to have a REALLY IMPORTANT use case to create more junk for my pocket. having to put more crap in my pocket and remember to charge it is just not worth it. This should have been a an app or i can live without it just to avoid the clutter.
Just to be clear: Jessie is a board member of Teenage Engineering. So when you say "its nice that they didn't try to make their own sht hardware", that doesn't actually hold much weight.
@@theorants "SANTA MONICA, Calif.-(BUSINESS WIRE)-Today, Santa Monica-based AI startup rabbit inc. has announced the appointment of Jesper Kouthoofd, founder of teenage engineering, as Chief Design Officer at rabbit. Together with rabbit founder and CEO Jesse Lyu, Kouthoofd will guide the future product design for rabbit while maintaining his roles as head of design / CEO at teenage engineering. Lyu and Kouthoofd first met more than 10 years ago and immediately connected over their shared passion for synthesizers, music and product design. Their collaboration officially began when Lyu’s previous company, Raven Tech, engaged teenage engineering to help design award-winning products like the Raven H and Raven R speakers. Lyu joined the board of directors of teenage engineering in 2020 and continues to serve in that role."
Although AI is incredibly useful, everyone is so focused around making it a product they can make money off of, rather than making a tool that can actually better society and vast amounts of work force, there is only so much an any form of AI can do, at some point down the line, a human is going to have to be overseeing everything.
Rabbit OS sells itself as an OS. That's a very special type of app. Rabbit would be a good app if running on the outer screen of a foldable, the form factor is just better and every single resource needed for the app is present, but no, lets make an overpriced plastic square designed by TE just because.
I’m glad you inserted the intermission commentary, because I totally disagreed with your generous defence to the criticism of R1 being an android app. Those critics are well aware that many devices run android, it’s common knowledge. The criticism is not for the fact that it’s an android app, but the implication that the Rabbit hardware adds nothing. On your comparison with android-powered devices (camera, TV, vacuum cleaner, fridge): well yes they run android, but you DO still need the hardware. In fact, the hardware IS the product. I can’t just install a fridge app on my android phone and suddenly have a fridge, can I? But I can install R1 on my phone and have everything Rabbit has, thus what the hell is the point of the hardware? THAT… was the criticism. The only value from the Rabbit hardware is for the company themselves (to paraphrase them: to control higher profit margin), but none for the consumers.
They really should’ve pushed this device as an educational toy for kids. It solves all the problems. Kids don’t (shouldn’t!) have phones. Kids ask a TON of questions which seems to be the only thing this thing can do. It can also scan random objects and say what they are which kids would find fun. Heck you could go on a hike and have the kid scan a bunch of plants. Will it always be right? Probably not since plant scanning isn’t even great on phones (I’m a plant addict) but it’s still fun! Now $200 for a kids toy is insane BUT it’s the only way to solve the “it’s just a worse phone” issue
This is the first time I came your channel and by accident, I really like the way you deliver your content, unrated and straight to the point talk. Now I'm subscribing and eagerly waiting for new content from you.
I believe Rabbit r1's biggest mistake was that they didn't know how to market their product because they kept focusing on showing the benefits of the software, and it ended up in the hands of tech reviewers who started roasting it as an Android app, wondering why I would buy a device designed for an app when I could install it on your phone. That's a reasonable point, but why would I need a watch when I can see the time on my phone? They may be better off presenting their product as a luxury item that blends the beauty of Teenage Engineering designs with the power of having an AI agent right in your pocket.
I didn't realize the design was Teenage Engineering. From the very first time I saw it, the only good thing I had to say about the rabbit was that it had great industrial design. I think anyone who's ever actually used AI would have known how useless this would be as a complete product just from the concept alone.
If you dont have permission to perform cybersecurity actions. Even if you are doing it for the good of the company. You are still breaking the law. The number one rule of cybersecurity is scope. No permission. No scope of work. Anyone REing the rabbit app is breaking the law.
14:25 Rabbit OS doesn’t need apps but 4 on device and have 800+ coming. Sorry, Jesse, you’re not just shooting your toe, but your whole foot in the process.
So they have a public endpoint which was publicly accessible but if you interact with it you’re breaking rules. Can I sue WebScrapers for even landing on my website? Did they accidentally find the infinite money glitch?
The business model is so clearly unsustainable, it's either a ruse to make some short term profit on the hardware, or it's a bait-and-switch exercise. The problem is that there are investors happy to look past this just to be in the AI space. Welcome to the next generation Dot Com bubble.
What I find especially weird that people call LAM "exciting." Did anyone really think about logistics of it though? I'm assuming it's something like gforce now where you rent a virtual environment your AI can run in. What else can it be, neither apple nor google will ever give root access to allow rabbit's AI to control your phone, and we all know they aren't releasing desktop/server software for self-hosting. So what then, you get a virtual phone where you gonna store your bank password alongside with your credit cards, addresses, IDs, social media, messaging and so on? You trust the big brains behind rabbit with all of that? Whatever, let's ignore the security nightmare for a minute, suppose they succeed and everyone adapts LAM technology. I can already see it, endless rows of GPUs spending thousands of hours comprehending fancy animations and performing screen gestures. Emitting tons of carbon just to let you do in 3 seconds what took you a single swipe before. Because somehow having machines talk to machines though a layer made exclusively for human convenience and aesthetics makes sense.
Fun Fact it had to be hardware because anyone can make that a app and they would have no moat around them so they made it hardware because you can't compete with them
I mean, the CEO literally says if you read between the lines, “this product could exist as an app, but if we did it as an app, someone else could too, and then we couldn’t force everyone to use us”
They’re literally a gpt4 wrapper. I wouldn’t be surprised if there’s an app that already does it better
@@sarthak-tishare an app that’s better, then without having to pay ai subscription.
there is, but like you said, a gpt4 wrapper will always be as bad as gpt4, which means you will always get hallucinations and stuff@@sarthak-ti
So they want complete monopoly. Got it
Every company want monopoly duh, the only reason Microsoft isn't one in multiple industries is because they can't just buy everything without prior agreement with countries and such@@ReigoVassal
If we only had some rectangular device, with a camera... No, a camera for forwards and one backwards... With a speaker AND a microphone... Yeah... And a screen... No... No wait... A TOUCHSCREEN!!! And can communicate ✨ wirelessly ✨
With sensors that can detect shaking and stuff...
And runs android!
Where else are you going to get all that?!
Truely a device does not exist. Thank goodness these clowns made one! Praise rabbit!
On the craig tablet from cashies
Wow! This is SUCH a great idea! Thank you rabbit for totally being the first to do this!!!! 😊😊
And a button on the side to raise or lower the volumes...no wait a complicated amount of movements because we want to save people time.
@@tezcanaslan2877oh mate.. at this point we just gotta understand the old mate craig is leagues above everyone else
it's not surprising they have made it! AND to top it all off a cool blue disc with the silver ring around it
AI is in it's Juicero era
Prime said this
@@TomNook.* The verge wrote it and prime read it.
@@monad_tcp so you know what it is. why ask what it is
@@monad_tcpAnd you failed.
LOL
Remove the scroll wheel and enable the touch screen and it becomes a phone. 2024, start up reinvents the phone.
That would have been a better plan. Market it as a hybrid dumb phone in the sense that it is not optimized for social media and content consumption yet could cover basic phone features.
@@hrdcpy Even better, market it as a kids toy.
That's a great idea
Many people try to download minimalistic apps towards social media or add ad blockers where instead they could just have a phone without any pre-built apps and social media
Like the iPod for kids without the sim card option and without social media @@hrdcpy
@@TempoFinoA $200 toy is a hard sell
hardware gpt4 wrappers are wild
The trend that really bothers me is people treating companies like they're their friends. "Oh it's okay it's their first product, give them some slack". NO. As a customer that bought a product that promises functionality you are in your right to expect that functionality.
hardware gpt4 wrappers are wild
according to the r1 though, it's not even gpt-4, it's a fine-tuned gpt-3 lmao
@@dingus1720 LMAOOOOOOOO they don't even have the money for gpt4 and they market it with complex words to make it look like it is powerful
this is truly a snake oil salesman moment
Absolutely savage
@@dingus1720gpt4 costs 10 times as much as gpt3
@@rethardotv5874the gpt3 API is $0.50 vs $10 gpt 4 per 1 million tokens. For an enterprise company openai can sell them a server for unlimited usage.
"Cloud Verification" === "Checking your user agent"
"R1" hardware == checking your device serial number for a thing that should have been an App
If your "device" functionality can be replicated with an app then you dont have a device, you have a gatekeeped app.
Gatekept
Gatekept
I really hope people realize the company that didn't secure their own server endpoints probably is not good at protecting any of the data you are sending it.
That's why we don't trust Google, Facebook, etc.
to be fair, I just don't know what that is. I think people tend to forget there are a lot of us who just don't know these things. for the most part I just kind of have to have faith when I'm dealing with technology because I'm not going to understand what everything means. granted I never believed that this was at all realistic but even if I did and I tried to do my due diligence there's a good chance I wouldn't know to look for that. I don't know anything about steam or windows but I use them.
@@moamber1 I do not trust Microsoft or Google but to function I have to utilize their services. As for Facebook Instagram and the other wonderful social media giants I have not had them for years. They are just ad machines. Choose wisely what you give your information to is the point.
@@moamber1 I don't trust them either but at least a breach is less likely compared to some startup with "questionable" security practices potentially leaking sensitive info.
At least google protects the more sensitive data like payment info from theft/breaches properly... 😅
@@dizzylilthingit takes time, just start reading, watching and learning from people. I'll help you understand what an end point is. It's semi self explanatory. Say you want to send data in the real world. You can input information onto paper and store that data in a letter. Data on the Internet works like that.
Your computer is making something like a letter with your info in it and sending it to the company. The company will receive it also in a similar way. In the real world you send that letter and the company will receive it wherever their mail is delivered, that's an end point in real life. The point of delivery is the end point aka the mail box hole.
Digitally the company also has a place where the data must go which is like a digital mailbox hole, that's the end point.
TLDR the end point is where the company receives data from you or your devices. It's like a hole in the door that lets you send and receive stuff to and from them without.
The op is saying that they didn't use any security to protect that delivery hole.
I hope that helps. Everyone has to start somewhere. Think of computer knowledge like a painter painting an image. He doesn't just spit it out like a printer, humans don't work like that and you would be overwhelmed if you tried. Instead they focus on one small part and then expand from there, they start with nothing, but end with everything...
"I don't think its worth shaming for using android as a starting point.... because they have soo much other stuff that we can make fun of."
Look, the AOSP is there for folk to fuck with, and thus, it's fine to start with it
It's fine to use as a starting point, but it's embarrassing if you literally don't go beyond that in any way whatsoever while pretending you are.
@@ilonachan Agree. It's moreso the fact that they're pretending it's different. Like it's literally just a small android box that can only run a single app, and if wanna run more than 1. Prepared to get harassed... Just call it what it is.
Feel he said it best with going over the Marquis stuff " we want something that hasn't been done before".
Asides from accepting criticism and having good PR. Humane at least has a reason to need their hardware to get the full experience ( assuming things worked as intended). You ( at least not yet) can't get a laser projection working from your phone, and as a wearable it should save you time without having to go through a screen, and if you need a screen for something lol.
They weren't really being shamed for using Android in the first place though. They were being shamed for lying about it. They told a bunch of stupid lies about how this couldn't just be an Android app, when it obviously could, and then it turned out that, surprise, surprise: it's an app.
Tony Fadell, the inventor of the iPod said in his book that he won't even look at the hardware of a startup until he's convinced it can't be done any other way.
werent there plenty of options when it came to music players back then. seems like there were plenty of other ways. that quote kinda comes off snobbish ngl
@@wojtek-33 Creative wasn't exactly a startup. I'm no Apple fan, but I think his point still stands; hardware startups are most likely in search of a problem.
@@wojtek-33 that was a BS software patent on hierarchical menus on MP3 players. how else would you make an MP3 player UI?
@@pikkyuukyuun4741 apple isn't a startup though
@@pikkyuukyuun4741There were options, but none of them could do what the iPod could do. The scroll wheel, the screen and the software all made it easier to put a lot more music on a portable device. Eventually people were storing their entire library of music on that one device... That's not even the iPod touch, that's just the iPod...
The same critique would still apply, if they built the Rabbit R1 on top of embedded Linux, or if they built "RabbitOS" from scratch entirely. It could still just be an Android App, there's no reason to create extra e-waste for this.
But then you have to give Google all your codes...
That's now how it works
@@bnorrish what?
After the recent Coffeezilla video exposing the "LAM AI" as just some ChatGPT scripts along with Playwright, an existing non-AI browser scripting tool, this looks even worse on their part. What a disaster.
I had the misfortune of working for a manager who was very much like this CEO. He had a background in technical sales and was 'that' kind of salesman... you know the ones who only care about closing a sale... with not a care about whether what they sold is deliverable. Some people seem to constantly be recreating a reality that was good for them and then gaslighting everyone and themselves into believing it. The ones you mean in employment have amazing manipulation skills, seem to be shameless and incapable of learning from their mistakes....pp In the case of my manager, he had got the company to invest millions into an app, putting the business into serious financial problems, for an app that was a wrapper over opensource and barely better than a proof of concept prototype....
the lessons i remembered from this, never work for a manager or a small company owned by someone with a sales background.... And most importantly buy anything from a business that engages in gaslighting marketing techniques like rabbit...
My life lesson is : never trust sales person, they're paid to lie.
I only trust word that come from the engineers (whenever I buy product I try to find the engineering manuals for details or even talk to the engineers).
I suspect the real reasons for releasing the hardware and not the app is to:
1. Get more money and hype from a physical AI device.
2. Avoid paying 30% of subscriptions / usage fees once they have to pay for their clunky LAM system.
Good video! We managed to get the updated app working with valid details now, so yeah they still can't do proper server restriction and as long as they don't charge a subscription or completely revamp the os, there will always be ways around this.
“Hopping back and forward and suddenly the name rabbit makes sense” 😂😂
I have to comment here. That made me laugh... 🤣
I've watched several of these kind of videos, but this is the one that made me cancel my order.
I've watched several of these kind of videos, but this is the one that made me regret my order. (Too late to cancel I already have it)
You can cancel orders? I am batch 6. Was still kind of excited, but this just really annoys me too much.
@@jasonjennings8465 i mean i have my device I was batch 1
it does work as a good toy for kids so my nieces' and nephews like it
based
The problem is not the fact that it is an Android phone, it's that they act like it isn't
About the first point. I think that people arent making fun of it because haha "funny crappy android app they are so bad" but because this whole device is useless and easily could have been a paid app in your phone. Especially considering all its weaknesses (like being not performant on this weak hardware), this whole idea of an extra device just seems pointless.
Now, you made the example of e.g. your camera also running android. Or the human AI Pin. But for those two, I'd argue it does make sense having an extra device. Because they all have something your phone doesnt. E.g. an extra OP camera with a lens and such or an projector, etc.
Exactly. The main problem is that the Rabbit app works with 100% compatibility with an Android phone. This is the reason the R1 is a laughing stock
If your app requires specialized hardware (or unique enough that a phone doesn't have it) to get it working, then it doesn't matter even if it runs Android.
Like an Android TV box, I can extract all the apks and install it on a phone. It will run, but the experience sucks because your phone doesn't have an HDMI port, nor does it come with a TV remote.
By the way I just rolled my eyes with this LAM bullshit. I bet this is just some system message telling the AI what actions it can do, with the actions have to be manually coded in. Something the regular ChatGPT 4 can already do.
Let's not forget the Rabbit also has terrible battery, and being an app, it would at least last as long as your phone.
@@shamringo7438 The LAM was supposed to be like "universal API" to communicate with apps without modifying the apps themselves.
@@rigen97 I highly doubt they are doing that at all. If they are doing this right now then why is it only right now limited to 4 “apps”?
If it’s truly universal it should be able to use a lot more services than just those 4. Heck, it should be able to google search a timer to add a timer functionality. No, I strongly believe this is just some system message for GPT that triggers Spotify’s API, or Midjourney’s API, etc.
@@shamringo7438 shrug, that's what they _wanted_ it to be. they clearly talk big. we knew from their past project that they knew how to make things sounds plausible and then just...not do it.
Electron is not a good thing. Apps leaking a lot of memory (especially commit size, which is a huge issue on Windows), and separate runtimes with developers of each individual application being responsible for updates, which means that a vulnerability in Electron is a huge issue. This is a problem with browsers too, of course, but they are a lot better at pushing out updates when ACE vulnerabilities are found. This gets extra scary when people write chat applications, where you can push exploits without user interaction, and use Electron.
/IT security specialist who is stuck on Windows and sick of having to pay for RAM sticks, just to literally be idle for "committed memory"
Electron is one of the worst things that ever happened to desktop imo
the only good thing ive ever found of electron without ever being in web development is that is for some home automation or game that has html, launching via electron is much easier.
I have to wonder what apps people are using that do this, I've used a grand total of 3 electron-based apps and none if them leak like that.
Agree. I hate electron. Its lazy development.
what do yall think of tauri tho
A correction. OS/2 was not an experimental joint project it was a multitasking operating system that also hosted Win 3.1 and DOS applications natively. While it did die off it was mostly due to cost as you were also buying a Windows licence. After 4 major versions it saw lots of use on non-consumer systems where stability was important.
Still even sees use today. I encountered POS systems using it in 2016 which are probably still running today and now I see industrial machinery running it in 2024. But of course I see industrial machinery running CP/M in 2024 as well so there's tons of old stuff out there. Somewhere in one of the stairwells of my father's old office building, on the 5th floor, there is an Altair 8800 running a rooftop weather station that has been operational since the 1970s (the campus has its own power plant and is not on grid power, so the 8800 has presumably been running since the 1970s).
CEO seems like they don't actually understand the tech and all they want is some sort of patent/to sell to some other company.
they sell nft too
They understand marketing. People get more invested in physical devices.
That's actually a lot of companies business plan, build something that will get you bought up by one of the big tech corporations
I‘ve been following „AI“ developments for a while now and there were some really cool projects i tried. One i vividly remember was a project from Facebook Labs where it could take a single picture and make a 3D model from it. Impressive stuff!
Nowadays every „AI“ is just the same API from OpenAI with a fancy skin on it. It’s disappointing and frustrating.
With the Facebook Model thingy it was an actual engineer behind it who put his heart into the project. Actually when i posted the results i got from running his algorithm to twitter he liked my post pretty much right away.
I want to see cool innovative projects like that again, not everyone packaging an API and calling it a day
Yeah ikr? There's so many exciting things AI can do, contrary to what apparently everyone is thinking, AI doesn't mean to just get some text from ChatGPT lol
Contrary to what is commonly repeated, capitalism doesn't reward innovation. If you want to see more passion projects, join a communist party and help build an economy that will give people enough free time to pursue their passion projects.
@@justinwatson1510 i mean yeah, you are right on that one. sadly there is no communist party in my country, and it doesn't seem we'll get a UBI any time soon
I think the issue is big ceo man sees ai is the hip new trend and wants to hop on it but ooga booga spend money bad, use openAI
Facebook/Meta has been doing some crazy stuff with their computer vision tech tbh.
Thanks for the video… Personally I got into deep learning 11 years ago, focusing a huge part of my life on solving really hard engineering problems, and it is unfathomable how many low effort grifts these days seem to be showered in VC money…
When I saw this video on Sunday, I immediately submitted for an RMA to return my Rabbit R1. I received a response Monday morning stating: "Your email is important to us, and we want to ensure that we address your inquiry promptly. Our team is working diligently to provide you with the assistance you need." ... As of late Wednesday, no RMA or any other information has been provided. IMO it should not take longer than 24 hours to provide an RMA to return a device. It will be interesting to see how long it takes or if I will get one at all.
This gets deep. In 2018 Bidu and Teenage Engineering released a similar project with Jesse Lyu (working at Bidu). Look up The Raven H. It’s a real thing!
Any updates?
There is no shame in using AOSP for the underlying architecture for your new product. It has software for the screen, wireless connections, touch...why bother reinventing the wheel?
Agreed
Then why reinventing the smartphone ?
I think it's the point of the video: if it run on Android, on a cubic object with a touchscreen, then why not make an app ?
Yup, that’s exactly what he says in the video.
The ridiculous thing is that there is actually a potential market for an AI assistant device that's not an app in a phone - medical alert devices. A small, lightweight device that doesn't have a million apps running in the background draining the battery, that could perform basic functions like reading labels on meds, monitoring vitals & alerting when to take meds, and a panic button to call 911. The only problem is that it would have to be a pretty bulletproof system, and that would all require FAR more actual effort than someone like Jesse "I threaten to sue 14-year-olds" Lyu could cope with.
More proof (Elizabeth Holmes, Web3, NFTs, anyone) that scammers have strong sway over investors.
- technology gets genuinely hyped (computers, internet, lab-on-chip, crypto, AI)
- scammers arrive first, (because it is the easiest to pivot if you are not actually doing anything useful)
- we are swimming in poop and can barely see the actual real good products
This cycle is so bad for crypto that I literally cannot name anything that is not somehow scammy.
Elizabeth Holmes made lab-on-chip infamous, but never actually (mis)used the technology.
AI scams are just beginning, at least there are some products which are usable, we should mentally prepare for a big avalanche of AI scams.
Nah, Holmes at least wanted to make something that benefited people, her screw up was when she doubled down instead of just shutting the company down when research doesn't pan out like the rest of healthcare / biotech start ups. Web3 and NFTs are just morons chasing trends against business fundamentals, it's the same nonsense as the Dotcom bust when people buy internet stocks even if the companies had nothing to do with the internet.
The issue isn't that it is an androip app, is that it could have been android app with no issues
I hope someone starts a Class Action Law suit.
Their comment section is turned off. That's telling.
I just realized something...
The Rabbit R1 is made by Teenage Engineering... So it makes sense that a teenager reverse engineered it lmfao
I do understand why they went the hardware route as yes, an app can't lock people in for apple or google to replace with their own. But it's just a bad idea especially if it's that fragile it should have never been considered to be made
There is nothing stopping apple or Google from doing that now, in fact they already have their own AI assistants. Soon those AI assistants are going to get big updates...
The original iPhone didn't even have a flash that could be used as a flashlight - the original flashlight app just turned the screen white. Which you could do with the web browser.
IIRC it was also spyware.
11:37 - for the non-technical people watching, the researchers and experts and hackers have actually become much better at reading this compiled, obfuscated bundle of letters and numbers and turning it back into actual code
but still, submitting compiled code to an app store is not even a moderately significant problem in the grand scheme of things... whenever a user recieves software, whether it's through an app store or through firmware installed on hardware or otherwise, they'll receive that compiled package
I'm not a genius. I will never develop a genius device, nor a genius app.
That's why I don't seel genius promises to other people.
People nowadays should:
a) Accept their limitations
b) Take responsibility for their actions
c) Be held accountable for their mistakes
Scamming people shouldn't be so easy.
This man does not remember the days of the flashlight app setting your brightness to max and filling your screen with white because hardware flashlights didn't exist back then.
I work at a robotics automation scale-up and one of the best things we did was move away from manufacturing our own hardware product. The sheer monetary and personnel overhead required to start manufacturing even the simplest product, heavily outweighs the margins on selling the product itself. I could speak at length on those tweets from the CEO but this is a comments section and nobody actually cares.
This is a CEO that has been sold a dream by a Product Manager that doesn't own a smartphone.
LAM is actually the same functionality OpenAI has had for a long time
Honestly if what comes from these junk AI in a box devices is Apple pulling their finger out and making Siri decent then it will have been a positive in the end.
That's not why apple is working on upgrading siri, it's because Microsoft was working on upgrading cortona aka co-piolat. Microsoft has been working for a long time on this and working with partners like Intel, AMD and Qualicom...
These devices are getting more attention than they should. Didn't we learn anything from the NFT era?
I don't know what is to be learned. NFTs were relentlessly mocked as well and now they're all worthless.
remember: this community will encourage you to try and build new things and that it is okay if you fail, but if you fail, they'll make sure you regret it dearly by berating you in a 30 minute highlight reel video of all your mistakes
everyone is just being so wholesome and building their community, right? it's all for views guys ... it's all for views ... and they'll get you riled up and teach you to not try to break the mould by making an example of someone else in their videos ... toxic
@@tomatoslavnobody makes videos like this about the poke walker released for the Nintendo DS, that square smartwatch that got bought by Fitbit, the micro PC community, the raspberry Pi community, dozens of communities are propped up around actual good hardware. this wasn't people trying something and failing, this was people selling a bad version of Siri on a $200 device that has next to no function
Jesse's reply sounds like chatGPT, just that he made all of it lowercase so that it "looks" like it's not, together with the "its" and "it's" mistypes.
lmao yeah. Weird af if he types on his phone but the auto capitalize didn't turned on.
Calling it cringe hits so much harder than simply calling it a bad product.
I saw some dude hyping this up a couple weeks before it came out.
And immediately I thought this is going to fail
Google lens is a far superior product
And it's free
The thing that had me worried pretty much right from the start is when Marques mentioned there is no subscription in contrast to the Humane Pin. Given that AI has running costs, how can that ever be a sustainable business model if you're not one of the big ones and this is just a side project?
Crypto bros are now "AI" bros 💀
Lit
I want this AI thing to turn into the pin from star trek but Im guessing Ill never get that in my lifetime.
That's the other failed project...
I'm impressed they've managed to jank their way to getting GPT-4 to talk to a back end on a remote device. But due to the nature of prompt engineering with extremely large LLMs like GPT, the moment OpenAI moves on to GPT-4.5 or whatever, it's going to break possibly forever. Worse, they implied the reason the Rabbit r1 even needed a dedicated device was to make use of a Qualcomm chip with it's own tensor processing unit to run some component of their app locally, which now the specs and app are out there and working on other devices, was obviously a lie.
Worse still, the hardware there is about $80/device, after losses on shipping, tax, and etcetera, there's maybe $25 of income left on their end, how are they paying the fees for GPT-4 access for more than a year for every one of these little Android doodads? I smell an incoming rug pull again.
Canceled my preorder, I agree, this needs to be an app and not a dedicated hardware device. One in which I have to carry, and keep charged, etc. This combined with their shitty behavior, I am done.
Its called ChatGPT, it does everything Rabbit does
Jesse should have said only one thing: We made a device because we wanted to. Not only the profit margins are higher, but we also wanted to minimize the amount of time people spend on their phones. The R1 is a cute little rectangular that can do lots of things and could help with your mental health by minimizing smartphone use.
I have seen your face in my feed a few times and i expected a gruff, JJJ like voice. I did not expect your voice to be so gentle
The CEO should've said "AI will change the way humans will use computers. The rabbit is that change."
A camera that doesn't take pictures is like a phone that doesn't make phone calls.
Tbh the thing with it being android based is just funny and kinda ironic *because* everyone has been saying for a while that these AI hardware devices could just be an app. It’s not that surprising it’s built off base android. Then bundled with the fact the app they built is actually (to an extent) compatible with android phones with not too much tweaking - is also pretty funny.
Drinking game. Take a shot everytime he says cringe. I’m dead, and posting from the other side.
At 14:49 the "even if we built RabbitOS as an app..." is discussed. Theo reacts to it by giving an Uber example. To be fair to Jesse, he has a point. He's basically saying: "If RabbitR1 was an app, rather than a separate minimalist hardware product, the experience would be way worse, as in comparable to Siri". What he means is that as an app, you would have to launch it, be interrupted by the notifications from other apps on your phone, etc. He's arguing that by giving you a hardware thing without notifications the UX is better as the assistant could be immediately available and not get its volume reduced every time a notification comes in etc, distracting you from the assistant / intended UX.
All other points made by Theo are valid, just being fair to Jesse here
they didn't want to make an app because they knew how badly they would be roasted for it and likely would also have been denied from appstores for data collection policies
15:45 he created the design of a mug ? or the color of nail polish the guy is using ? I don't get it
the mug is cool thou
I think it's supposed to be the stack of colored squares. Kinda fugly ngl, but I don't like their company's design aesthetic anyway so... 🤷
3:48 Well, I have used that fact to customize my user experience on my Samsung Galaxy A5. I use Taskbar as my home screen and GBoard as my keyboard.
That Jesse Lyu thread is basically: "if we me make it an app no investor or customer will pay us an insane amount of money and we don't want that. Of course that's bad to admit outloud that it's basically a scam so here's a bunch of barely thought out excuses."
16:18 "How did he raise money?" There are , people asking the same question , and for example Cofeezilla posted video about it
The moment I saw it I said so it's just a GUI app and they did the hardware to stand out and raise funds. It has absolutely no other reason whatsoever.
I love that they are like trying to hide behind the thought of companies stealing their code to integrate into their OSs, only for all the companies to be midway through doing the job better than they ever will by doing it directly on the phones all on their own... not that anyone will really use it anyway.
I already thought rabbit was crap, but now you made me hate the founder. I don't like grifters and nft bros.
I want to rephrase the first point here: An App is hard to build but easy to copy. To make a good App you need two versions and maintaining both well is a challenging task, as stated. But your competitor can make a crappy App and claim to do everything you can do and undercut you in price to a point that users are not trying your App anymore. They would get the competitor App, have a bad experience and go, 'well, that's bad, why should I spend even more money giving this other one a shot?' Result: in the end, your great idea fails and the bad actor gets more money out of it than you.
How did the thousands of successful app companies do it then?
Obviously both of these "AI in a box" projects are a way to make a quick buck on AI hype. They don't need to be a separate hardware, but if they release as an app they would have to compete with millions of other apps and they just would get overshadowed by Siri or Google voice assistant. And they know that they don't have anything worth of value to offer with their service, because the app model is more sustainable in a long run and if they were confident in their service they would release it in as many forms as possible (like app, web, PC client program, etc) like any other big service does.
You forgot to mention that the LAM is not currently used for the 4 apps. It's a research model that they intend to use in the future.
because it doesn't exist probably ?
"for future use", yeah, sure, meanwhile lets use OpenAI chatbot version 3
that flashlight he was referring to was literally a white screen with the brightness on max. Because the first iPhone didn't have a flash led on the phone to function as a flashlight!
~ 29:00 Imagine trying to find a video to back up your point that MKBHD wouldn't have gotten here without iteration.. and choosing the video that is probably the most opposite your point possible because it highlights a thing no one talked about, it covers something that no one would've even paid attention to if he didn't. It's like the exact opposite of a valid example.
0:47 Wait, what was that pop up? It is like an extension or something?
There's a TH-cam short that makes fun of the phrase, "Think about it". I can't believe I saw it in real life.
Wow, that response from the Humane guy makes me want to learn more about their products.
A typical example is that when I make an online purchase with my credit card I get a notification from my bank requiring me to approve it, then I get a notification from the online store which arrives through the mail app, and finally I get another notification from my bank that the amount has been charged to my card.
What I find even more ridiculous is that this bespoke device of theirs is so expensive for what amounts to a very crippled phone. I can buy a decent Android phone for $200.
I think his English is so bad because he's actually from China and it appears his funding has primarily come from Chinese investors. He wouldn't be the first to bamboozle venture capitalists though.
OS/2 Warp 3 and 4 were actually really good.
ToS barely changes - it's more the user end, rather then the programmer end - and while i'm not sure if i'm CORRECT i'm. going by the fact that Google's ToS even for colab BARELY if ever changes - but the Acceptable Use Policy gets vague and unvague -- The laughable points that you could in some ways translate most of it to "You can't do anything on colab until you just pay us all of your money and still can't do anything" -- is why I laugh when people are saying "ToS keeps changing" -- A website's one could change, but a corporate one rarely changes.
the first thing i thought when i saw the Rabbit was 'this should be an app' and then 'i don't want more crap in my pocket'. You have to have a REALLY IMPORTANT use case to create more junk for my pocket. having to put more crap in my pocket and remember to charge it is just not worth it. This should have been a an app or i can live without it just to avoid the clutter.
Just to be clear: Jessie is a board member of Teenage Engineering. So when you say "its nice that they didn't try to make their own sht hardware", that doesn't actually hold much weight.
Wait what?? This is news to me, ty for sharing
@@theorants "SANTA MONICA, Calif.-(BUSINESS WIRE)-Today, Santa Monica-based AI startup rabbit inc. has announced the appointment of Jesper Kouthoofd, founder of teenage engineering, as Chief Design Officer at rabbit. Together with rabbit founder and CEO Jesse Lyu, Kouthoofd will guide the future product design for rabbit while maintaining his roles as head of design / CEO at teenage engineering.
Lyu and Kouthoofd first met more than 10 years ago and immediately connected over their shared passion for synthesizers, music and product design. Their collaboration officially began when Lyu’s previous company, Raven Tech, engaged teenage engineering to help design award-winning products like the Raven H and Raven R speakers. Lyu joined the board of directors of teenage engineering in 2020 and continues to serve in that role."
Although AI is incredibly useful, everyone is so focused around making it a product they can make money off of, rather than making a tool that can actually better society and vast amounts of work force, there is only so much an any form of AI can do, at some point down the line, a human is going to have to be overseeing everything.
Rabbit R1 & the Humane AI Pin are pre-Beta/Alpha products sold for full price.
19:48 bruh literally thought "if OS is so great why isn't there an OS2" and stopped thinking entirely after that
IBM: hehehehe
Bro has bleached his hair so hard, he thinks Swift or Java can't be discompiled. You're such a phony 😂
Rabbit OS sells itself as an OS. That's a very special type of app. Rabbit would be a good app if running on the outer screen of a foldable, the form factor is just better and every single resource needed for the app is present, but no, lets make an overpriced plastic square designed by TE just because.
I’m glad you inserted the intermission commentary, because I totally disagreed with your generous defence to the criticism of R1 being an android app.
Those critics are well aware that many devices run android, it’s common knowledge. The criticism is not for the fact that it’s an android app, but the implication that the Rabbit hardware adds nothing.
On your comparison with android-powered devices (camera, TV, vacuum cleaner, fridge): well yes they run android, but you DO still need the hardware. In fact, the hardware IS the product. I can’t just install a fridge app on my android phone and suddenly have a fridge, can I?
But I can install R1 on my phone and have everything Rabbit has, thus what the hell is the point of the hardware?
THAT… was the criticism.
The only value from the Rabbit hardware is for the company themselves (to paraphrase them: to control higher profit margin), but none for the consumers.
They really should’ve pushed this device as an educational toy for kids. It solves all the problems. Kids don’t (shouldn’t!) have phones. Kids ask a TON of questions which seems to be the only thing this thing can do. It can also scan random objects and say what they are which kids would find fun. Heck you could go on a hike and have the kid scan a bunch of plants. Will it always be right? Probably not since plant scanning isn’t even great on phones (I’m a plant addict) but it’s still fun!
Now $200 for a kids toy is insane BUT it’s the only way to solve the “it’s just a worse phone” issue
This is the first time I came your channel and by accident, I really like the way you deliver your content, unrated and straight to the point talk. Now I'm subscribing and eagerly waiting for new content from you.
I believe Rabbit r1's biggest mistake was that they didn't know how to market their product because they kept focusing on showing the benefits of the software, and it ended up in the hands of tech reviewers who started roasting it as an Android app, wondering why I would buy a device designed for an app when I could install it on your phone. That's a reasonable point, but why would I need a watch when I can see the time on my phone? They may be better off presenting their product as a luxury item that blends the beauty of Teenage Engineering designs with the power of having an AI agent right in your pocket.
I didn't realize the design was Teenage Engineering. From the very first time I saw it, the only good thing I had to say about the rabbit was that it had great industrial design. I think anyone who's ever actually used AI would have known how useless this would be as a complete product just from the concept alone.
If you dont have permission to perform cybersecurity actions. Even if you are doing it for the good of the company. You are still breaking the law. The number one rule of cybersecurity is scope. No permission. No scope of work. Anyone REing the rabbit app is breaking the law.
Basically they want to become a if this then that alternative with a voice assistant frontend. You can do the same thing with node red.
14:25 Rabbit OS doesn’t need apps but 4 on device and have 800+ coming. Sorry, Jesse, you’re not just shooting your toe, but your whole foot in the process.
So they have a public endpoint which was publicly accessible but if you interact with it you’re breaking rules.
Can I sue WebScrapers for even landing on my website? Did they accidentally find the infinite money glitch?
The business model is so clearly unsustainable, it's either a ruse to make some short term profit on the hardware, or it's a bait-and-switch exercise. The problem is that there are investors happy to look past this just to be in the AI space. Welcome to the next generation Dot Com bubble.
4:38 “rabbit’s just a crappy android phone running an app” just like all android phones
What I find especially weird that people call LAM "exciting." Did anyone really think about logistics of it though? I'm assuming it's something like gforce now where you rent a virtual environment your AI can run in. What else can it be, neither apple nor google will ever give root access to allow rabbit's AI to control your phone, and we all know they aren't releasing desktop/server software for self-hosting. So what then, you get a virtual phone where you gonna store your bank password alongside with your credit cards, addresses, IDs, social media, messaging and so on? You trust the big brains behind rabbit with all of that? Whatever, let's ignore the security nightmare for a minute, suppose they succeed and everyone adapts LAM technology. I can already see it, endless rows of GPUs spending thousands of hours comprehending fancy animations and performing screen gestures. Emitting tons of carbon just to let you do in 3 seconds what took you a single swipe before. Because somehow having machines talk to machines though a layer made exclusively for human convenience and aesthetics makes sense.
“I love teenage engineering” is the biggest red flag I’ve ever heard from what started as a level headed video game
I hate that companies rushing this kind of slop to market are directly leading to people not taking AI seriously
there is still a subdomain from their old website that is still accessible I commented it but i think youtube took it down
This video and Emily's Twitter thread will be the script for all the tech explainer channels. I can hear Logically Answered rn.
Fun Fact it had to be hardware because anyone can make that a app and they would have no moat around them so they made it hardware because you can't compete with them
The CEO is not smart. There is nothing stopping someone from making the same thing and releasing it on Android and iPhone...