Okay Rabbit R1, You Have Our Attention!
Вставка
- Опубліковано 18 січ 2024
- In this clip, Marques, Andrew, and David discuss the Rabbit R1 AI hardware device that was announced at CES.
Watch full episode: • The Rabbit R1 Is the W...
Shop the merch:
shop.mkbhd.com
Twitters/X's:
/ wvfrm
/ mkbhd
/ durvidimel
/ andymanganelli
/ adamlukas17
/ ellisrovin
Threads:
Waveform: www.threads.net/@waveformpodcast
Marques: www.threads.net/@mkbhd
Andrew: www.threads.net/@andrew_manga...
David Imel: www.threads.net/@davidimel
Adam: www.threads.net/@parmesanpapi17
Ellis: / ellisrovin
Instagram:
/ wvfrmpodcast
Shop the merch:
shop.mkbhd.com
Join the Discord:
/ discord
Music by 20syl:
bit.ly/2S53xlC
Waveform is part of the Vox Media Podcast Network. - Наука та технологія
Me: "Call me an Uber home."
Rabbit: "You're an Uber home."
😆🎅🤣
Honestly the best use case I can think of for this actual device is as an accessibility device. For someone wheelchair-bound, blind, even dyslexic, etc, this could be a life saver.
then why not use smartphone with this capability (or app in the phone) ?
To get rid of VoiceOver or TalkBack. The whole idea is to not rely on visual UIs and do it all with natural language (if accessibility is done right).
@@jameslin7457 Because there are people, older people usually, who are not up with the tech, for whom the learning curve to use a smartphone or tablet is too high. Gadgets like Rabbit R1 are perfect for them.
@@Corteumyou must be joking
@@jameslin7457 Joking, how?
I think the point is: they annouce it. EVERY Tech bro buys one because it's ONLY $200. They sell out. They get bought by Google or Apple and they make a mint.
Modern day “innovation”
Probably not. All Apple acquisitions are always for the tech, never for the brand (only exception is Beats). This tech is super easy to replicate with a few AI engineers. Google could buy the brand, but since people will realize the whole thing is pretty much useless, I dont what would be in it for them.
@@bahroum69 this way apple can just acquire a LaM workout having to develop it from the ground up. Like Siri.
@@bahroum69 Beats brand aligned perfectly with Apple so it's not much of an exception. Apple rarely buy brands because they specialize in ripoff product that exudes a fake quality through herd mentality. Few brands do so.
You have got it in one!
"I will call you by the name 'Uber to New York Airport'"
Amazing
Really summed up my feelings about Assistant after all these years 😂
Call me an Uber. "You're an Uber!" Google Assistant almost has the comedy shtick down! Can't wait until it gets good enough to do the "Who's on First?" routine.
Trillion dollar company.
“take me to new york”
He said Newark (New Jersey) btw…there’s no such thing as New York airport
There’s no way the big tech companies aren’t going to be able to offer everything this thing offers and more within a year or two.
Siri has sucked for so long though, so that’s why I think they are hopeful
I think this entire company exists to be bought out. And it will be in less than 5 years I pretty much guarantee it. I think the actual product is just proof of concept.
@@ruekurei88facts
Except the they won't. Siri is still like 10 years behind. Alexa is behind. Chat GPT is a revolutionary tool that was given to us in 2023 but it can't do what Rabbit does.
They can but they choose not to. They cant sell you dumb accessories if you dont need anything else
The main reason I’ve never found use for any of the assistants is because it requires me to speak aloud. If I’m in public, I am far less likely to speak aloud into Siri/Alexa/Google/etc for anything because why would I want strangers to know my business…or even at work. I don’t want to disturb or broadcast to the entire workplace.
Exactly, beside, and that's something that quite suprised me during the vid, is that they absolutely did not talked about the realism of the device. Like an unknown companie, pitching a totally new device type ? Sounds more like a scam than anything anyway.
Same. I feel like an asshole talking to Siri even if I’m alone!
This
25 years back normal people would have scoffed at smart phones the same way. I mean how many of us would have said that looking into a screen and not paying attention to fellow humans extremely rude. But look where we are today. We talk in public into a device today and our fellow travellers overhear all that right?
@@gamernerd7139 sure I get that. But I still don’t see people using virtual assistants in public frequently enough for me to feel comfortable doing it.
I saw it somewhere that the founder Jesse said the biggest potential of Rabbit is the ability for everyone to train it to be what they want it to be. Then Rabbit will be their 'product' instead of merely an app that they use. This is very interesting and makes it so fundamentally different from any digital assistant.
Yeh to me that’s the killer feature.
I can train it to get / do things specific to me that Apple or Google are never going to build.
Like go and get my account balance from my bank. Under Google / Apple that would be via APIs which would require my bank app to enable it…. Never going to happen.
So think the idea that Google / Apple will just do this is not as sound as some think.
Also they was kinda wrong about the funding. That was just to build a hardware device. They had been working on this for a lot longer than that.
But how are you going to train it if it says 'sorry I did not understand' half of the time and copy-paste a wikipedia page the other half of the time.
@@DJ_POOP_IT_OUT_FEAT_LIL_WiiWii Show us a demo with it FAILING THAT OFTEN, otherwise it's just a claim, when the demos DO show it working.
@@David-xt5ob Very similar to Open Interpreter as far as automating between apps... Good to have competition.
@@brianmi40 I'm not buying it to see just how bad it is. Apple claimed Siri can recognize your voice. Newsflash it doesn't, it worked one time out of three at best and in ideal condition. That's bad and make it useless in my book.
Same goes for Chat GPT. It can't provide help for any non trivial task. What it does all the time I used it was copy-paste Wikipedia article. Now for Rabbit MQ or whatever. They have shown zero insight that convince me it is any better. Same false claims Apple made.
15:40 I love that Google basically did a dad joke and essentially said, 'Okay, you're an Uber' 🤣🤣🤣
Someone tell Marquis he has dad mode turned on in his google assistant.
Looks great but I see no reason why I would want this to be a standalone device to my phone
Meant to be a device that works alongside your phone. Its basically a personal assistant that can actually do things and stops you from swiching from one app to another on your phone. That being said, whether a seperate device is needed for this is function is questionable...
@@2ueYeah I get that bit. But I guess I'm just like, with all the major OEMs already working on this in the background it has a relatively short lifespan before it gets replaced by your phone, without the need even for a Rabbit app.
Because Apple won’t let you replace Siri with it. So anytime you wanted to use it you’d have to open the app.
@@LucidSquirtle Apple are most likely adding their own LLM into iOS 18 anyway, so will we even need it as an app? Cool if people want to be an early adopter, but I don't think it's bringing the timeline forward much in this scenario.
Happy you and so many others feel that way so this company has time to scale these out to deliver to interested users like me.
Mom says she already has a rabbit in her drawer.
Aw, bruh.. don’t go looking for that rabbit, you hear now?
Imagine that one with AI capabilities 😅😂
@@angiealphabet5259 Actually, you might be on to something...
@eldritchbeauty 😂😂😂
And I'd totally fund that too ngl 😂😅
what kind of relationship do you have with your mom?? so creepy
Been using aitranslations myself lately, saves a ton on translations.
ChatGPT gets basic factual things wrong all the time, so I hope you have someone looking over the results
@@blinkx1070yeah and if you know both languages you can even see where the mistakes happened. I had it tell me a frog flew with a snake on it's back, because it was clearly going back and forth between Japanese and English when I was only trying to use it in Japanese. If you were translating the words directly without understanding, jump and fly in Japanese (飛ぶ), it makes sense that it might have thought a frog (known for jumping) could fly.
Anyway yeah, it's kinda wild when you know both languages, 😅
If accessibility is worked out correctly, this could be a huge step forward for everyone, but especially for disabled people. Moving towards a future where natural language is pretty much the only thing you might need to perform actions online instead of relying on UIs such as browsers, websites or apps with their own limited accessibility tools such as VoiceOver or TalkBack is the way to go. But it has to be built with accessibility in mind from the very beginning in order to be a real breakthrough. Total "Her" movie vibes.
AI acceleration is being built into cell phone chipsets, as you know, so Rabbit is likely going to be integrated into them. Apple or Google may gobble Rabbit up very soon.
This is 100% an acquisition play for Google or Apple. No way this product makes it to market
@@Chris-qy3dp
IDK, cloning their features seems extremely easy.
There are already open models who can work together to do the same thing as the LAM and training one from scratch doesn't sound that hard for companies who can create that exact same data automatically at scale
This is assuming Rabbit AI is anywhere close to being so context aware that it can take in these vague prompts and actually execute what you want it to do. Highly skeptical of this feature, which tbh is the only real game changing feature of their AI if its true. If not, then its not much better than current phone assistants which companies like google, apple etc. can improve upon pretty easily if they invest a little more resources into it
My guess is that they are nowhere near having a product ready for market. So any potential acquiring company will see that and Rabbit will die.
If they can implement it into a smartwatch I can see this type of concept working.
Hey new subscriber and long-time listener. It's great to hear you all reviewing this product and going over possible pros and cons. I look forward to more content from the group.
I see HUGE potential in the accessibility market.
Everytime a sketchy startup launches a product no one needs, people say "'it aktchually would be great for the disabled or elderly people". And then, it flops.
And in the market of people who are not really phone-dependent, but just keep it to be able to get stuff done on the go
You've convinced me I *don't* need one.
I truly don't see how useful this is going to be. I don't take Ubders/Lyfts;I rarely fly; I'm a former chef so I'm quicker at figuring out what ingredients I can put together for a meal----I just don't see the need for it and yes, """one more thing to plug in and charge""", thank you for pointing that out.
The physical product just feels like a novelty. This should absolutely be either an app, launcher or an actual OS option. Given what access it requires though, there's a lot of privacy concerns. Not to mention having to talk your life out in public.
Not having to wait for the app to load, not having to look at the device or wait for it to turn on is the selling point for me
Well isn’t that the point though? You don’t have to be specific with the device. They’ll just hear general things. For example if you say “Schedule my next doctor’s appointment.” People won’t know who the actual doctor is along with the type of doctor (primary care, cardiovascular, OBGYN, etc).
@@woaiiw8735 They mention it displaying stuff on screen so presumably you will need to look at it for feedback at some points. As for loading/turning on, those are hardly issues with current day phones, especially if this was made to be a phones launcher or OS. Then it would be identical just more powerful/convenient.
@@woaiiw8735 wait till it gets things wrong and orders or books events incorrectly and then you have to deal with calling chat bots around to fix the mistakes. Not worth all that trouble in my view. Simple things like ordering a pizza is not simple as it involves customizing a lot to your needs. My kids go over so many things, half pizza peperoni & other half with few diff veggies, pasta of certain type, quesadilla with variety of fillings etc. When was the last time I picked up a call and said one 12" cheese pizza. This device is gimmicky and will work best as a personal companion for conversations, which ChatGPT already does with dataset up to Apr 2023.
@@thaBootyologist Sure, if it does indeed work that well, to a degree that’s true. But that’s still exposing more than via current methods.
If a young woman is sat in a bar right now ordering an Uber, unless you can see their screen you’ve no idea what they’re doing. If they say out loud “Order me an Uber home”, anyone around them now knows that within a short period they are going to be hanging out outside, potentially alone waiting on a ride. Ample information to potentially follow or intercept.
🎯 Key Takeaways for quick navigation:
00:02 🐇 *Introduction to Rabbit R1 AI Device*
- Discussion about the Rabbit R1, an AI hardware device.
- Initial intrigue sparked by social media teasers.
- Identified features: microphone, speaker, camera, and a screen for natural language input tasks.
01:28 🗣️ *Natural Language Input and Use Cases*
- Exploration of natural language input through a video demonstration.
- Examples of commands: ordering an Uber, finding a podcast, checking the fridge, and more.
- Emphasis on the device's focus on understanding and executing natural language tasks efficiently.
02:53 🤔 *Challenges and Expectations*
- Reflection on the challenges faced by current AI assistants in understanding human-like commands.
- Desire for AI to be as helpful as a human assistant, eliminating the need for specific prompts.
- Skepticism about how well the Rabbit R1 can perform in real-life scenarios and concerns about its effectiveness.
03:35 📱 *Comparison with Existing Solutions*
- Comparison with existing AI assistants like Google Assistant, Siri, and Alexa.
- Presentation of the Rabbit R1 as a solution to reduce the complexity of app interactions.
- Debate on whether a dedicated hardware device is necessary or if improvements can be made to existing platforms.
05:24 🔄 *Large Action Model and Context Awareness*
- Discussion on the significance of large language models, specifically mentioning Transformers.
- Importance of understanding context in voice interactions with virtual assistants.
- Theoretical advantages of a large action model for providing a more natural and context-aware AI experience.
06:47 📐 *Design and Form Factor*
- Commentary on the design and physical attributes of the Rabbit R1.
- Mention of its size, appearance resembling a phone, and the involvement of teenage engineering.
- The comparison of size with existing devices and considerations about practicality.
08:36 💡 *App vs. Dedicated Hardware Debate*
- Exploration of whether the Rabbit R1's functionalities could have been implemented as a mobile app.
- Raised questions about the necessity of a dedicated hardware device.
- Mention of potential updates to existing assistants rendering dedicated hardware less relevant.
10:29 📉 *Long-Term Viability and Market Dynamics*
- Consideration of the company's recent funding rounds and rapid entry into the market.
- Speculation on the long-term viability of the Rabbit R1 in a competitive landscape.
- Reference to the different approaches between Rabbit and other companies like Humane.
11:38 🌐 *Multimodal AI and Prompt Engineering*
- Acknowledgment of the impressive multimodal AI capabilities demonstrated by Rabbit.
- Discussion on the evolving trend of reducing the need for specific trigger words in voice interactions.
- Insight into the potential ease of use through multimodal AI advancements.
13:30 🔗 *Integration with Existing Platforms*
- Emphasis on the integration of Rabbit R1 with various accounts and apps for streamlined tasks.
- Suggested alternatives where the Rabbit features could be added to Google Assistant or Siri.
- Debate on the likelihood of major players incorporating similar features in the near future.
16:19 🌈 *Aesthetic and Trust Considerations*
- Brief mention of the device's bright orange color as part of the trend of colorful tech.
- Expression of trust concerns regarding AI, comparing trust levels with current smartphone usage.
- Reflecting on the assumption of trust required for effective AI interactions.
16:33 🚗 *Rabbit R1 Use Cases and Trust in Technology*
- Discussion on the execution and trust in privacy when using AI devices.
- People's tendency to shop around for options, even in tasks like calling an Uber or planning a trip.
- Balancing trust and preference in utilizing AI for various tasks.
18:11 🔄 *Training Mode and Universal Access*
- Introduction to Rabbit R1's training mode for interacting with different services.
- Theoretical capability to access any app through training and using prompts.
- Comparison to Google Assistant for developers and the potential of universal access.
19:36 🎶 *Customizing Actions and Potential Limitations*
- Exploring the concept of training Rabbit R1 for specific tasks not initially supported.
- The potential of user-generated actions and customization.
- Acknowledgment of challenges, such as changes in UI or website tweaks affecting functionality.
21:55 💡 *Potential Use Cases and Market Adoption*
- Speculation on the two potential directions for market adoption.
- The possibility of users finding specific tasks to enhance their workflow.
- Comparison to other niche devices like the Stream Deck and potential modding community.
23:17 📱 *Comparison to Palm Phone and Social Media Integration*
- Discussing Rabbit R1's role compared to a phone and potential limitations.
- Addressing concerns about social media integration and messaging.
- Consideration of scenarios where Rabbit R1 might be useful as a secondary device.
25:14 ⚙️ *Device Material, OLED Screen, and Charging*
- Thoughts on the material quality, suggesting a desire for a premium version.
- Speculation about the potential for an OLED screen and concerns about charging.
- Personal preferences for a more premium build of the device.
27:33 🌐 *Connectivity, Siri, and Google Assistant*
- Discussing the connectivity and potential competition with Siri and Google Assistant.
- Speculation on the future development and potential integration of Rabbit R1 features.
- Comparing Rabbit R1 to Clubhouse in terms of being a hardware or feature-centric product.
28:45 🤖 *Modding Community and Product Strategy*
- Predicting the emergence of a modding community and unique uses for Rabbit R1.
- Considering the product strategy of Rabbit, focusing on affordability and widespread adoption.
- Acknowledging the potential challenges and uncertainties surrounding the product's success.
Made with HARPA AI
What is Harpa AI?
@@prithviraj1080 it is a chrome browser plugin that summarizes web pages for you. Very helpful!
What is Google
I bought it. My selling point was that they included one year of free perspective ai pro. Just that subscription alone would be more expensive. Also, being involved with the music I can tell you teenage engineering is a revolutionary company.
The product is really just a catalyst for the big two to get better. Which is not a bad thing.
People in the US using a Voice based assistant as if the wall in their house wasn't made using Cardboards
@@jensenraylight8011Crazy how my cardboard house is over two hundred years old with no issues. Almost like there’s no reason to build a house in the US out of concrete.
@@asphaltannihilator157 it's not a durability issue, it's a Noise issue if people talk to their AI assistant loudly all the time
Your Cardboard wall basically just let all the noises pass through
Just imagine if your roommate, kids or wife talk to the AI assistant and the Assistant talk back, all the time
Even with brick and concrete wall, you can still hear a conversation, even if it was just a mumble.
With Cardboard wall, RIP sleep
@@jensenraylight8011 Are you under the impression Americans don’t currently talk or make noise inside of buildings? Why are you projecting a concern none of us have?
@@jensenraylight8011lol what? You think this device will suddenly cause a huge surge in noise pollution? I mean it’s not like people have watched tv or listened to music (often at much higher volumes than speaking) in their homes in the US for literally decades
I think if you think about what it is trying to be, it actually makes sense.
If you have a human personal assistant, that is with you most of the time, your communication is mostly outloud. When that person is first hired, they will take some time to learn how you like to travel, what restaurants you like, where you shop, etc. At first they will ask a lot of clarifying questions that will diminish over time. After they learn, simple commands result in the desired outcome.
You tell the assistant what you want and they take care of most tasks, and while you still have a phone, it is mainly for personal communication and entertainment.
The issue with this, and other new technology, is that it is meant to do what a person does, but people expect it to work better, cheaper, and more seamless than someone hired to do the job.
Phone assistants are absolutely going to catch up to this very soon. We are already seeing assistant with bard
We home assistant users already have LLM’s as our in home assistants replacing Alexa , etc. it does much of what rabbit does.
Can Siri actually do voice to text properly? Last time I tried it failed miserably to understand 'Turn on the lights'
@@ChrisS-oo6fl Alexa sucks most of the time beyond the most basic tasks.
I use bard daily and you have to be very specific and give it guidelines or it will start lecturing you about the risks and disadvantages of eating your salad with a plastic fork instead of just telling you "yes, ranch usually goes on salads".
12:35 you can take the stuff off your fridge, put them on a plate and ask the rabbit... the fridge is there for marketing value but you can be smart and adapt to it's limitations
I pre-ordered one, one of the 5th batch. Yes, I agree I'm paying $200 for an app. But if it can deliver just 70% of what the keynote promised, it'd be one of my main apps if not the app I use the most to interact with the other main apps. Netflix costs more every year. And why shouldn't it come with a beautifully designed physical UI?
i know an african prince who just told me he has a fortune he wants to give you
@@RoZaxTheGreat Is he Sundar Pichai in disguise?
Meanwhile people pay thusends to use an app for a month... And the switch still exists despite being another thing people have to charge and this could honestly replace a "phone" for some people who hate touchscreens and just want an phone with an AI... Which today is like 6 phones on the marky the Pixel 8s and Galaxy S24s... Which each cost more that the price of the R1 as they are more then just a phone and AI chatterbox. So yeah perosanlly it makes me want to use Edge on my phone more to talk to Chat GPT... But I kinda do that on my pC any way just in text form... But yeah if Rabbit is litterlly a heavly modded Chat GPT I don't see why it won't work as advertised.
yeah, but, what would you actually do with it? Order pizza?
@@bahroum69 Probably not pizza, but maybe the more complicated Chinese and Korean food combos I order frequently. Besides that, I'll try use this as a remote control for UA-cam, to create the same chart I make for every survey I analyze at work, to create temporary meeting invites just for checking a coworker's availability, to monitor content sites I'm interested in and summarize the updates... and any other functions devs/moders may come up with for this system. You know, there are a lot of things I do regularly across multiple devices that take too many clicks. It's too early to say which ones of these can be done well. But I feel confident some of them can. For $199, I'm willing to give it a go.
I see a lot of potential for this product. For the longest time, I have been wanting a device where I could take a walk, listen to my music, but not be distracted by everything else the phone has. Like one of the host were saying if he was going out to meet up with friends and he didn’t need an entire device, this would make a lot of sense also in situations like music festivals and concerts where I don’t necessarily want to be worrying about my valuables having this cheaper alternative makes a lot of sense as well.
They made these things called MP3 players for a really long time. Brilliant. Grab yourself a Diamond Rio PMP-500. Also, snag a 64-128MB SmartMedia card if you can find one!
What I want is an AI assistant that talks to ME, not the other way around. I get a buzz on my phone, I tap the notification, the AI says "hey, you have an unanswered email from this company, did you want me to write something up for it? Okay, I'm sending it to you, I'll go ahead and send it after you approve it."
When the guy mentioned no social media, I think it hit on the reason I'm excited for this as a standalone physical device instead of an app. I don't actually want an AI function like this to have direct and uncompromised access to everything my phone has access to, and sometimes I don't want to use my phone (or, as with social media, I want a barrier for when my addicted brain wants to default to doom scrolling). A device that is just there to let me get stuff done, that is limited in some ways and open in others, that's perfect. The, "It just needs to be an app," thing is like, "Why would you buy a standalone screwdriver? Just put a hole for interchangeable bits on the handle end of a hammer!"
I want that hammer! 🔨
This literally like the two-way pager back in the early 2000s. Great idea but then phone co-opted the tech and it went away. Rabbit it is a great proof of concept for exciting future of tech potential.
I wouldn't call it a proof of concept after the miracle that is chatgpt. Even the foremost experts on AI were convinced that something like chatgpt would only be possible in no less than 10 years and yet here we are.
@zedetach you do realize what "proof of concept" means right? Are you saying that rabbit and chatgpt are the same? They are the same in that they are generative a.i. acting on large language models but rabbit takes it further. It's a large ACTION model. So YES it's a proof of concept because no on else has brought this capability to forefront. You must have been on the short bus to school I see. It's a cool idea but in its current iteration it doesn't have enough legs to be a real thing. This is just so show what capabilites are out there now BEYOND chatgpt. Again...proof of concept. Class dismissed.
Buddy a proof of concept is a novel idea that is still far from being realized commercially. R1 has gone way past that stage. This is a product that is scheduled to be released this year. Class dismissed.@@thereal415er
it's going to be a paper weight in no time
@@zedetach the AI experts are right though. Chat GPT is only an illusion of intelligence. In practice it barely does more than copy paste wikipedia pages.
They are trying to have a paradigm shift away from apps to AI as the primary OS. This could be really good for attention and healthy human interaction. Def some challenges there but I see what they are going for and I’m here for it. It’s serving the purpose of a dumb phone by approaching it from the opposite direction. Point being to give you your life back.
apps are the primary OS and they'll never change it!!!!
Yes I agree I think the mvp behind the project was when they said “our smart phones were supposed to give us time back instead of taking up all our time” this would be a great product as it has all the essentials of a phone without all the addictive things such as social media or apps in general
@@jakealering5711 Agreed. Apps are literally optimized to take our attention. This could be a breath of fresh air.
Thanks for the great summary of Rabbit intention according to you which I totally agree. Hope they will succeed or influence the way we interact in near future as the current one is clearly not the way to go anymore
Well said! I think it will be the next face of phone for those who want to be done with distractive apps and unnecessary social media … think of it this way; those who need their peace of mind will opt for this… I actually can’t wait to buy one … I’m tired of having to pay for every app service 🤦🏿♂️
“Tell everyone I’m running late” is scary.
@everyone in discord lol
I'm sorry I did not understand.
15:32 - The interaction with Google Assistant....perfect.
This device reminds me of Tamagotchi pet.
You guys discussed that the r1 should offer a store where people can upload their best trainings to get it to take cool actions for them. I believe that rabbit will make that available to others by default so that their LAM gets very powerful very quickly. What is smarter than having users train your model?
A lot of this I believe depends on a) solving input information and b) people using calendars effectively…
Actually, because I have no idea where on my home screen most apps are (except for the most commonly used ones), calling an Uber on my iPhone is: open phone, gesture down to bring up search, type 'ub', select Uber app...
Not saying it's complicated, but I think the rabbit guy has a bit of a point, in that AI could make that easier. Of course, I'm sure Google and Apple are working on this too, so I don't know what the longevity of this company is going to be.
Edit: Of course, I can already open the Uber app (or any other app) with Siri. I usually don't because I don't necessarily want to bother other people by broadcasting out loud what I'm doing. That's a problem in general with speech recognition AI. It's bad enough people are walking around with their head down staring at a little screen, we have to listen to them too, now?
You could ask google assistant to order an uber. Then all you have to do is click a button to confirm it. Google assistant has nearly all the stuff it advertises, or the ability to do it very quickly
I think it could be a better smart home interface than Alexa or Google but Google will probably catch up soon. I think it could also find a niche as a non-app based smartphone UI. I think if they figured out how to handle pictures, messages, pay and calls in their UI I think it would be an interesting play in the android vs apple phone ecosystem. It would also be cool if it was a single OS that did your phone and smart home. Let’s see how ambitious they will be. I fear if they are not super ambitious they will run the risk of being made obsolete by Google, ChatGPT, Microsoft, etc soon
Re: The R1 could just be an app.
The R1 being a device that can do "everything" a phone can do without it being able to suck people into a social media feed is its own superpower. Don't discount it.
My cell is a full size desktop the size of a screen. It's mostly an app storage device. Apps that connect to internet connected things. The r1 might be a utility that is voice programmable, not with keyword or search phrases, but just normal speech. It's a voice remote for the tv, lights. The phone will have every app open through the AI, at will. Refrigerators have AI now. We won't have to remember to click on an app or make a quick reminder while doing something else. Eventually, our phones might be like desktops, but not go away. I think I still have my brick phone.
Rabbit R1 makes alot of sense for people that consider money as no issue, and don't want the hassle of clicking on apps to order taxis, book holidays or ordering food and maybe do your taxes for you too. It automates heck out of alot of tasks, so you can focus on other things in life, like enjoying it.
Actually, I think people are 100% wrong in their assumptions that you don’t need a separate device, and eventually your phone will be able to do this. The problem is phone manufactures have their own App Stores along with terms and conditions or other legal frameworks that have been established along with copyright restrictions. Phone companies will probably never allow for cross integration between apps through the phone. You can’t even access a lot of third-party apps through Siri because of this reason. I just think the phone companies have built themselves into a legal corner that will be really hard for them to pivot from. Although rabbit may have to worry about the third-party applications and websites coming after them.
thats no the problem. All you need to do is give an app a permission to do so. Don't kow how does it work on iOS, but on android it is fairly easy. You already got apps to do macros in other apps. With AI you just add another layer that do this for you.
@@megustaav Apple's walled garden stifles innovation
google already does third party integration. and with competition they would definitely do that more
Why couldn't Rabbit be fitted into a watch design? It connects to the phone and to the internet, so it does not do all the processing.
delay via cloud and the warehouses needed to process and store all that info
Agree, this as a watch would be sick
I need it in a tamagotchi
Great breakdown 👌
Im waiting to receive my R1 soon, cant wait!
8:05 that is a thought that my friends and I immediately had: when starting a business you want to choose a market that is not so small that it can esasily be engulfed by larger companies.
Maybe they’re just trying to get their foot-in-the-door with hardware and will switch to a cheaper app later on?
Awesome finally we have real world Jarvis now
Is this a reupload, or an excerpt? Edit: oh it’s the full unedited recording?
Huge market potential for this. Imagine in places you don't want to take your phone, e.g. music festivals, exercising/gym, day parties, beach trips/nature experiences. Yet still want to have some connectivity and ability to access the use of your daily apps. As long as it has a great build, waterproof, this has great potential.
I'm surprised no one seems to have mentioned this: Imagine if the R1 could serve as a virtual proxy on social platforms like Facebook and Instagram. This could revolutionize the way we interact with ads and updates. This is exactly what their user interface training promises. The R1, with its user-trained interface capabilities, could filter through ads, letting our virtual presence engage with them. If I only want to know when a particular set of friends post, I don't have to scroll pass sponsored content. Meanwhile, it could summarize important updates from friends, triggering only the notifications we care about. This concept not only challenges traditional social media ad models but also enhances our digital experience by focusing on what truly matters to us.This is why Google Assistant and Siri may not be ready to offer this full functionality. It is a can of worms, actually. I ordered mine exactly because if this one feature. I'm an old guy and don't do much instagram, but I still look at Facebook.
Great point!
I suppose we utilize the training feature to filter our news feed. But how exactly do you plan on consuming it? We can assume you wont go back to your phone to do so. Are you using rabbits small screen to see the post? Or listening to written posts/tweets in a text-to-speech sort of way through the rabbit device? What if tthe post has a picture or video?
You bring up what seems like a very specific situation that i can respect however id like to hear your rebuttal i do want to understand your pov
@ItsTheMajorScalebro I don't have the R1 yet, but I have thought about how to filter the news feed before, and making the R1 able to operate as a virtual person makes everything possible. I'd be fine with asking R1 about friends and it tells me anything in the newsfeed that I'm interested in, and NOT bother telling me about sponsored content or other updates I don't care about. The reason Apple, Amazon and Google assistants wouldn't do this if they didn't have to is that they are upsetting the Advertising revenue. That's why I think this sort of product would NOT come from any of the existing major players. It had to be Rabbit.
I think after the initial setup. It might be good. Also, could you use this as a Bluetooth device to your phone instead of getting a plan on a carrier?
It might just be me, but I could listen to David explain these types of "under the hood" processes weekly.
Thanks for this talk...interesting how this rabbit will evolve.
I am in the fourth batch.....fifth batch sold out. I am cautiouslessly excited for this. I think it's neat tech and I hope it works as advertised. I am sure in a year this will be an app on a phone.
It won’t work. It’s a scam. Good luck.
@@alfreedom7750 lol....oh ok.
why do you say so? elaborate your accusations @@alfreedom7750
The clip turned into the whole podcast 😂
So I was already looking at getting a Perplexity Pro subscription because I have using it a lot more lately, and the fact the R1 costs the same price and came with Perplexity Pro for a year, made it an easy choice for me to pick up.
So as you guys explained the use cases it quickly fell apart for me UNTIL I remembered my father, who is 80, never texts, never emails, has a flip phone and is constantly having challenges like not being able to summon an Uber… For someone like him this would be a huge game changer! For example he loves UA-cam but can only access it thanks to voice commands!
If it feels to good to be true, it probably is.
Why isn't this an app?
Companies like apple an google won't give them full integration/data access to their phones for it to work smoothly.
15:33 perfectly summarizes the current state of personal assistants 😂
The amount of setting up this will require is going to be crazy. And i have to constantly tell it about appointments like going home? No lol
So thinking about this from a cybersecurity perspective the issue that I see is that I think that the rabbit r1 is going to get shut down by the app owners. How this interacts with applications is similar to how spam bots or bots would be interacting with an application unless the rabbit r1 has relationships with these app owners or companies, those customers will be viewed as bots and I think the potential for fraud using the rabbit r1 language action model is too high to where they're probably going to get shut down by a lot of app owners from allowing the applications to do the actions.
Some might still let it and Rabbit might constantly have to adjust the LAM for that but I feel like they're going to get shot down and because the app owners also have an incentive to have users in their application because they want them to see ads and everything like that so I think they're probably dead in the water once they come out.
Finally I think a lot of businesses also benefit from trying to tack on extra items for example if you're going through Amazon they try to get you to buy something else or if you're in the Walmart app ordering groceries they're going to try and get you to buy other things with your purchase to increase your overall sale so the r1 preventing this is something that they're not going to like and they're going to again try and stop.
Anything that reduces a user's screen time on an application is a direct threat to the business model so Rabbit R1 is going to be a threat to them and allowing them to interface with the application like that will not be beneficial and so I think they're going to get a lot of impediments put in place by the app owners.
100% true,I was thinking the same! When its too good for customer, its not good for owner
It doesn’t actually go to webpages like a bot, it interacts similar to how IFTTT does with APIs. For example, Google Assisrant can already order an uber
Here is my take on this: if we get to the point where this device can actually do anything your apps can, the app developers will quickly find ways to stop it. If you are not entering apps, they won't be able to push ads/subscriptions/micro transactions etc.
Yea I was thinking the same, app Owner will maybe not accept their acces from R1
They can, and many already do, charge a fee for extra features like assistant integration
The potential on this is insane, like for starters they right about the fridge thing but you could also take put what you have or want to use, put it in the table and ask it what can i make with this, i just see so many possibilities to ask it anything i would ask someone else that knows something i dont
I think it’ll be cool if you can record a transcript and then print out or email a summary of what notes you have, and that what was recorded on the transcript
No matter how good AI gets it will never be able to read your mind. If I was a personal assistant and my boss told me to call an uber I would still double check to make sure I was thinking the same place as them.
Never? Why so confident? Have you see how far time stretches into the future? Infinity! We've gone from Alex Kidd in Miracle World to VR pron in 35 years. It's an exponential curve.
This is going to be baked into phones 100% - Rabbit option one, It gets bought and baked into a phone - Option 2, It dies and goes away - Noone wants another devise
Option3 - next product is a cheap home device. People still buy Google Home and Echo speakers and smart clocks
I kinda dig the design. And the fact that they went with the Helio P35 and not some cheap manufacturers, inspires confidence in me. I have used Mediatek devices and I like them.
Very helpful, might have bought this thing , until I saw this. I don't need it, I know that now! I've also learned what it cannot do , which the big sell glossed over. You are right, Big Tech will buy them out and you will be left with a cool orange ornament for your mantlepiece. A bit expensive for that!
PS. I like the presenter and the contributors, low key, soft spoken , to the point , unabrasive. That's refreshing.
This would be good for young children because it has no games.
I had no idea they were SUCH a new company! That is pretty incredible. I like the idea, I still feel like it has to be a phone replacement. Interacting with apps isn't difficult enough. Even my tech averse friends know how to order an uber from their phone. I feel like nobody is going to want to have ANOTHER device, plus as noted other AI models will get this functionality via the google assistant etc.
Rabbit Inc. was formed in September 2020, so it's 3+ years old. It just got it first round of funding (Series A) in Oct 2023. Doesn't mean it's a new company
Maybe their intention is to showcase the technology in their physical form factor with the end game to be bought out by Google or Apple who then would integrate within their OS or a separate app.
It could be life changing for people with learning disabilities death (if it learns sign language), or blind or even with physical disabilitisles....
I pre ordered and in first batch. They giving early adopters 200 dollars worth of Perplexity. So that’s a years subscription for the price of the pre order. Win win
This will be an app very soon, don't quote me.
“This will be an app very soon, don’t quote me.”
-Ramonariasdisla 2024
@@garrettjohnsen
As @ramonariasdisla once said: “this will be an app very soon”.
@@trolley4388
Can it be a physical button very soon though?
Lol. Your videos are so informative and fun to listen to. I am curious about this device, too. From a cybersecurity perspective, I am a little hesitant, so I will do my research on that as well, but I do want to try it. 😅
Remember the adverts for Google Glasses before they came out versus what was delivered?
The fact that you are going to train the rabbit AI to perform better improved customized actions is a plus for me. This even reduces the burden of devs to teach their ai more customised improved actions
Puts the burden of UAT on the actual user and not on QAE's.
I can build an app with all of the R1 features in a day. It's basically an app with an OpenAI subscription.
Ohh wow we got a badass here ladies and gentlemen…
They trained a custom visual AI to navigate any apps GUI visually rather than connecting to each API.
@@femiairboy94 This isn't about me. This isnt a brag. Its just a simple fact. This product is just trying to circumvent the camera and mic access that ios and android wont readily give. What am i missing here? Everything they talk about can be done using the paid ChatGPT subscription.
I think the analysis here is great, I'd just define it more as "an assistant device" as opposed to a phone. The point, it seems, is like you all said earlier, it's like having a physical assistant for those who don't have a personal assistant.
Sidenote: Kinda reminds me of Pebble when they first started.
The problem I see is, first my pockets are already full. Two, I still need a phone. Three, soon phones will do this too.
What about an Enterprise version for industry?
Consider if schools had enterprise models as teacher’s assistants:
“Check attendance” teacher to Rabbit assistant
“All accounted for, Ted & Cruz have traded seats, Ted is now next to Tina.” Rabbit to teacher
“Ted, Cruz, move back to your assigned seats please.” Teacher to class
I could see the PRC using an enterprise version of this thing in Red.
"Comrade, what does Xi Thought say about eating ice cream?"
"Xi Thought says eating Ice Cream is a frivolity of the West that should be avoided."
Hot tip: no one buys the hardware and they pivot to making an App. 👍
The problem with R1 is that society doesn't want to speak anymore
You can already tell the Google Assistant to "book an Uber to my office" and it instantly loads the app ready for you on the confirm button. It's lightning fast, too. "Call me an Uber" is a ball drop on Google's part, though.
The other issue with this replacing my phone is the camera. Phone hardware has iterated toward where it is now for a reason. Sure if we don't need as much screen real estate because we make apps redundant that could be an area for change in the hardware, but other than that I would much rather see this just as a well integrated feature of my phone.
Not to mention the fact that just because many interactions would benefit from having natural voice interactions rather, there are apps that people enjoy and would not want to loose entirely. Like UA-cam and Instagram etc, they are fundamentally visual apps
I still don't trust any software or AI that does something for me, such as ordering or getting a ride or making a reservation. I want to physically see and make sure everything is correct.
This is the number one issue that i saw immediately. "Get me i home" turns a $25 uber ride into a $100 Uber Black SUV trip without telling you. There is a reason for the additional inputs and most tech companies are trying they're hardest to get you through the process as quickly and HAPPILY as possible.
@@GoFidoGo plus if this took off I could totally see companies making it so if you aren't specific with what you say it WILL give you the most expensive option without telling you of the other options!
This could be an app and not have to purchase TWO devices that people will toss in a year! More E Waste for NO REASON at all!
For $200 this is a horrible phone replacement/alternative for average users, but could be life changing in the disability healthcare sector - ESPECIALLY with the individualized training potential.
If Rabbit or the company that buys them focuses on that, and on building a seamless and robust ecosystem of healthcare devices - that’s where this becomes revolutionary
What I like about rabbit R1 and other devices such as the humane pin is that it challenges the notion of having a phone; what the utility of having a phone is and a deeper question of what the purpose of a phone is
Like if you would adopt MKBHD (in the right circumstances)
wtf
Already preordered. Time to see what else it can do.
I'd def get this for anyone who can tell it a series of steps and a goal instead of app, app, app...
This convo is funny. We're SO spoiled nowadays
Like all of your videos ❤
My favorite podcast
If the model learns on its own at some point how to perform click/tap actions on websites/apps, it could also learn how to create new accounts for any of the services you might be interested in. Your rabbit account linked to your rabbit assistant could even create the rest of the new accounts needed for you (gmail, spotify, etc...). You just need a rabbit account with some personal data that might be updated every time you subscribe to a new service (because of the specifics of the service), an account/password manager to handle all the new accounts generated on all the services or to manage the connections to those services, a secure authentication method for confirmations and payments (fingerprint reader on the push to talk button) and that's pretty much it. Rabbit R2 sure will look even more promising if the big ones doesn't catch up.
I think they made it look like the phone from the movie “Her” on purpose. Don’t know if anyone else has noticed. The movie is about human like Ai.
But… where is the processing done? On a server? On-device? On your home computer that has to be turned on? A mix of all of that, depending on the task?
according the their website:
"When you interact with rabbit OS, you will be assigned a *dedicated and isolated environment on our cloud for your own LAM.* When our rabbits perform tasks for you, they will use your own accounts that you have securely granted us control over through our rabbit hole web portal. We do not store your passwords for these services. rabbits will ask for permission and clarification during the execution of any tasks, especially those involving sensitive actions such as payments. They will provide predictable feedback on whether the task was successfully executed or failed, along with a reason, thanks to our neuro-symbolic research."
@@pikaa-si9ie thanks for looking into it. So basically the hardware is mostly a terminal to their cloud service. I guess the language processing might be local, since they’re promising such fast response? In any case, if the company goes out of business, it’ll stop working I guess. And possibly they might be forced to a subscription model after growth flattens out down the line, even if they’re saying no subscription now.
I think the lack of shopping around feature can do a couple of things: people will get used to trusting the device out of convenience**. And it will force service providers to provide better and more reliable services at competitive prices because the AI simply won’t book service providers who have a bad track record. The efficiency of the AI will enable it to filter out the worst and select the best service providers.
**When UberEats started I was skeptical because I didn’t think people would be willing to pay a 30+% premium for their food(delivery fees&tip). Why not just order straight from the restaurant? But no Uber has caught on and people don’t mind paying those extra fees in preference for convenience.
I think southpark and the Alexa episode correctly captured frustrations with talking to tech. Even if r1 works, may still be harder than clicking.
I wonder if you would be able to train it to read a document or text book and ask it questions in relation to it such as construction standards. That would be extremely helpful in my line of work as a builder.
I have a feeling that they're hoping to get ahead of the curve just slightly, to then be bought at some point by Apple or Samsung or someone wanting to implement this functionality. Tried and tested.
My thoughts is that the idea of the device is great, love it. I think the biggest problem is this is another device. Like it was to be dependable enough that you would take it instead of a phone. I'm thinking of my elderly father who doesn't have a smart phone but can use a google home well enough.
The problem for what you pointed out is that, yeah, uber might fix their app but these companies hate the idea of that because while you're looking at the screen they can introduce more things to you. Like going into a McDonalds now and the self serve kiosk pops up dumb adds making your experience slower. They're never naturally going to want to make your experience voice only.
Looking forward to the rabbit watch :)