The Race to Create the "iPhone of AI" is Heating Up!
HTML-код
- Опубликовано: 8 май 2024
- Get Nebula using my link for 40% off an annual subscription: go.nebula.tv/coldfusion
It seems that we're entering the age of AI hardware, will it flourish? or will this just be a short-lived footnote in technology history. Our main focus will be Rabbit's R1, but we'll look at all of the questions and implications surrounding it. In a special episode featuring Jabrils, we investigate.
Thanks for @Jabrils for providing his insight. His AI companion video: • I'm lonely so I made a...
ColdFusion Music:
/ @coldfusionmusic
burnwater.bandcamp.com
Get my book:
bit.ly/NewThinkingbook
ColdFusion Socials:
/ discord
/ coldfusiontv
/ coldfusion_tv
/ coldfusiontv
Producer: Dagogo Altraide 
Наука
![Stray Kids [INTRO "Lose My Breath (Feat. Charlie Puth)"]](http://i.ytimg.com/vi/udz0a5Wecxk/mqdefault.jpg)
These aren't products that any of us really want, these are all just "founders" trying to grab VC money by throing the AI buzzword around
If we waited for people to want shit we would still be riding horses
@@willinton06 I think you mistake life-changing technology vs. gimmicks. I have the income, would never spend a nickel on those features. I like to research and book my own tickets and hotels. I like the mental challenge of doing those tasks. You would be surprised how many people are not interested in being a passive guy yelling at his phone because the stupid gimmicky AI booked him the wrong ticket.
yes, New tech can only work only if they can do everything that Smartphones can do, and offer a new thing that smartphone can't do.
this is something that can be done with a smartphone app, or a bluetooth paired hardware.
those tech founder are Delusional, idk why they believe this kind of tech could work
that's why PC and Smartphone is dominating, because it's a Generalized computing, and can do pretty much everything,
it didn't need a specialized hardware to do stuff.
the only thing that could dethrone Smartphone and pc is Hologram AR technology, especially if it was good enough to replace a
traditional screen, but i guess we won't get that for a very long time,
because, people are jumping into the next fad which is AI, and rob as many VC as Possible, while the Buzzword "AI" is still Effective
100% agree. Its just insane how much money gets poured into this thing... if it succeeds it will be incredibly expensive because of the computational power required...
Couldn't agree more. It's just more shenanigans to pass money around to the usual suspects. It's pathetic.
AI race heating up, and here I am still trying to get Siri to understand my simple request
Yeah, once they figure out how to do it safely I’m sure apple will have a true LLM trained Ai like GPT hardware installed. It’s only a matter of time. I would imagine the real difficulty of that isn’t that they can’t already do it, it’s just that it will have to be perfectly executed or it could be problematic in various ways. They’re are already open source LLMs out there that you can install locally on your machine that supposedly competes with GPT 3.5 so no doubt Apple could make this happen. And for stuff that is crucial to the use of a phone: making calls, getting directions, setting reminders, etc etc, Siri is ok and it’s not crucial for it to know how to help you build a webpage, etc. That said, I agree that Siri is a joke in comparison. Will be interesting to see what they end up doing.
Truth!
Tell me about it!
why are you top comment with 9 likes (nvm bots)
Yeah, Siri still won't do what I want either! She'll order a pizza for me, but she won't ever pay, even on my birthday!
Dude at the beginning was dodging like a politician
I was just sitting here like "answer the goddamn question man! Wtf does your device even do!?!?" 🤣😂🤣
you didn't hear him? he said it's AI! that's all the info I need to blindly throw money at it
@@hanifh6891 exactly
"You know... it does the thing, that A.I does, A.I stuff, really complicated A.I processing...so... are you going to still write that check or should I look for the direct deposit" 🤔
Dude clearly wasn’t media ready and was just focussed on the paycheck.
The man at the beginning had diversity cachet
he literally is the co founder, not some random guy they found on the street@@virtualpilgrim8645
This is the classic “get good talent and innovate a bit so a big company buys you out” model. I think the entire point of Rabbit was to be bought out by Apple to be made into their Siri, they make a few 10s of millions and Apple gets to be at the forefront of AI.
I agree
Rabbit has what I like to call a “one update away” business model. They are one Apple update away from being out of business.
exactly what I was thinking
Well, at least with rabbit r1, i won't need to pay a subscription.
It reminds me of that weird square smartphone that Microsoft came out with .. And lasted maybe 2 months before being discontinued
Something that is holding Google and Apple might be politics, so when your assistant is able to interface only major apps (probably via special APIs), you are closing your app evosystem down, when you want the user to have undisrrupted experiences
@ninjaundermyskin Funnily enough that thing could have worked as a Smart Watch.
Everything that rabbit can do can be done as an app in a smartphone. This is basically chatgpt with access to a library of custom actions. I don't want a separate device, work on a software I can use on my phone.
I mean Id imagine most people buying it are people trying to decrease screen time
Don’t worry my friends it will come
The landscape is changing really fast and it easy to see that most of the gap where start up try to fit in will be filled next years
Quite risky to start playing in this game as a small business
Exactly, it’s just the API with a scammy hardware layer
You're saying you don't want a Palm Pilot in 2024...?
exactly , and ChatGPT for example just a python program surfing the net for you , fast
The Humane guy at 2:30 and the Rabbit guy at 5:00 shows the importance of communications professionals in the world of tech. Their expertise as engineers might be off the charts, but the ability to actually put these products into words without sounding like a charlatan (Humane) or sounding mush-mouthed (Rabbit) is an entirely different skill set.
Also it’s the classic stereotype of someone focused on specs or what’s there vs benefits or use cases.
to me, the humane seemed more impressive. Not the R1.
R1 feels like an app I could have on my phone. Not a separate device.
This feels like a pager back in the old days
Don't see the reason for two devices, just have it integrated into phones. If some new processing hardware is required then phones will probably just include it.
Samsung S24
Exactly
Well small companies cant enter the mobilephone market … and you cant call an phone with voice assistant innovating.. so you cant collect VC money and call yourself the new apple or google
phone hardware isnt there yet . also this is meant to replace phones . just like ar glasses will
@@myname-mz3loNot entirely true. I bet most phones are way more powerful than this. This could've easily been an app.
Even if it might not offer much additional value to regular people’s life, it could be a game changer for blind/vision impaired people
Or just an app instead of a new device
You are completely right. I'm wondering things like:
- How fast it can recognize things?
- Would it be possible to perform "external" actions?
I'm a programmer and purchased this product as soon as I saw the training capabilities. I can't wait to play with it.
@@pantherdddjvdgx For the impaired person, the challenge to navigate and open the app still remains while this device can has much better usability than a phone as it is voice based and has physical buttons for input.
Also, the recognition features are just next level. On an app, you'd have to press different buttons to initiate different actions unless they develop some next level UI
@@pantherdddjvdgx It's hard to navigate an ap. With this device is just a press of a button. Simple yet functional.
@@Yaqins Apple will just expand siri.
I’ve seen several demos of the Rabbit R1 but nothing really shows me what an agent should really do and that is guide me through the process, by asking me questions along the way making sure all of the relevant information is complete before processing it.
Siri and Google's voice assistant could easily make Rabbit obsolete once their LLMs and transformers mature and can do the same thing
I want an AI that actively protects my information.
Open source is the way.
Going offline the data u want to protect ?
AI is based on collecting data to improve itself, you're dreaming if you think a company will not only offer that to you but resist the urge to sell your data.
most AI required connection to the internet since most devices arent powerful enough to run the AI models, so the models run in the cloud.
unless you have a 100% offline AI your data will always be shared
As in open source software without the (in my opinion scary) AI.
Or maybe simple AI.
I'm pretty sure that simple voice assistants like Siri can run offline In a lot of cases.
Of course it still needs to connect to the internet to fetch weather data for example but actions that require your contacts list can run fully offline for example.
As an engineer, I wonder how much of the demo was "canned" or staged for the demo and how well it will work in the real world.
It's no doubt fake or scripted with limited use specific to what he asked. Go off topic and it will likely crash. You can just tell the type of people they are. They are sales people trying to convince investors, not tech people who care about the end user experience. So much of this about and it's sad when people fall for it.
Faker than project milo
@@talibong9518 lol, yeah. And that was Microsoft. Can't trust any of them
You didn't notice all the cuts when it shows the device in his hands?
lol guess you are never in any investor presentation
A big factor jumping out to me here is how crucial that interface system is, being able to have one interface for every other app, and still have that experience feel natural is going to be a leap and they have taken that leap. I've seen some refer to this capability as a 'super-app', but it also could turnout to be received as an unnecessary extra step by the market and then fail to become the aforementioned 'super-app'.
Nice video as always, dagogo! Informative. 👌🏽
"You have to log into the apps that you use, so that Rabbit can access the data within." I see no possible problems here.
Yep, it's basically putting an AI layer over the same data-gathering apps that we already have privacy issues with.
@@LeftoverBeefcake so more privacy issues with a product that looks very sus.
Indeed, I dont quite see any security issue with that, aint it.
do you expect it to just know and input all your credentials?
@@elvisvan way to miss the point
As others have said, smartphones can already do this stuff. And those that can't, will be able to in the releases this year.
exactly.
Not just that, but for when you don't want to use your phone, we already have smartwatches. I don't know any person with more than 6 functional brain cells that will go for a humane pin to use their phone less or do a function their phone does without taking it out. The rabbit thing sounds fun to use, though.
And one of them was essentially talking about no longer needing multiple apps. Just one for everything.
Most people don't want that. (And you can kind of see a primitive version of this with all apps trying to be shopping services. Because it works in china, they want it to work in the west too. But most don't want to buy directly from Instagram or tiktok)
Of course force it enough and younger ppl who grow up with it will be fine.
I honestly kind of prefer AI on a separate device as an option. Allowing an unknown AI access to your entire phone and personal info might be a little more intrusive than it needs to be.
But why pay another device that can. Just add that money to the phone and get the phone that can.
I can see where this would actually be a great training and responsitory unit for athletes and or coaches. Something to the effect that has a lot of technique or trainee applications also in perhaps technical field working on different types of repairs on modules and layouts to where it can see and then be able to combined the visual with information of what is being viewed. For example watching a batter swing at a pitch, comparing the movement and then being able to compare to successful examples and then translate that information back to the participants.
Another example could be repairing a motherboard on a PC that has a multitude of different chips and diode’s resistors etc. By analyzing the board and then knowing what the issues are could perhaps be able to analyze which particular piece would need to be replaced without having to replace the complete component. Also making a pc board for example and then accessories to go with it depending on functional needs. Just a couple of quick thoughts from watching. Good luck with it, it is exciting!
I have a neurodevelopmental disorder, ADHD, and I struggle tremendously with organisation and doing repetitive or "boring" tasks. I want a virtual personal assistant, which can manage my time, deal with mundane tasks like preparing quotes and invoices, help me log my life, so I can remember what I've seen and done, and basically carry the mental load that weighs me down so much. Then I'll be free to put all of my attention on the things that I'm passionate about, where I can actually make a difference in the world. Rather than suffering anxiety about failing, guilt about when I screw up and being paralysed by all the things I need to do, but simply can't get started on. Something like this, a task-based AI agent, is exactly the kind of path to get us to that kind of place.
As an autistic person I also hope it will also be able to summarise, translate and send my thoughts to the right people, talk to them and bring the summary to me. Talking takes TOO MUCH time and is ineffective af.
This reminds me of about 20 years ago, when refrigerators were supposed to be able to automatically detect when one of it's contents was getting low and automatically order more online. Twenty years later, almost no one has a frig that does that.
Or would really want one, which is probably the main reason they haven’t taken off. Who wants their food automatically re-ordered? How’s it going to know that I don’t want cheddar or whatever else this week?
The internet of everything is a terrible idea. An internet ready fridge is just foolish and a waste of money...
@@vooveksit could just add it to a list.
@@TheMysteryDriver How would it know what to add to its list?
@@fonkbadonk5370 by seeing what's no longer there. There are fridges that already do that.
As much as I'm always rooting for the underdog, I think the R1 is the type of device people will buy, play around with for a bit, then put aside.
Bingo
I don’t even think they will buy it knowing their smartphone already does this stuff. It’s being powered by a 3rd party cloud service so there’s not even really AI involved from the front end device.
This is clearly an investor trap
same thoughts, I don't know why this R1 is making a buzz online, I just don't see it replacing phones that can do more. Average people buy phones because of how good the camera is, or because they want to play the latest games on phones. As I've said in other comment on other videos, this won't be remembered in a few months. Hope I'm wrong. First time I didn't finish a ColdFusion video.
Reminds me of the Pebble smartwatches. Cool, but the big boys will take over the market segment and integrate it pretty quickly.
It reminds me of the Zune!
You might want to do an episode on AI hallucinations. I think most of us think AI brings superior logic to the table when in reality it does not use logic to form its answers.
Thank you for your insights!
These are just toys that current smartphone could do and beyond.
Exactly, this doesn't have to be a hardware piece, it can just be an app
Toys that are gonna make certain individuals filthy rich soon
you think noone tried to do this on a phone ? if they did they would be so rich ... the phone hardware cannot support ai yet ...
@@kevingp12 I think that would be a long term plan. They just wanted to make a proof of concept. If they the android or apple is convinced , they will buy it. I guess they couldn't have mode it in the first place if they relied on apple or google
@@amritprusty6097 you don't have to convince Apple or Google is just another app in the app store
my main problem with conversing with phones in general is that there are times i need to think before finishing the rest of my sentence, or if i stutter, or if i want to add something more to my sentence.... and phones generally just assume u're done speaking rather than waiting for u to continue. if there was a way to make them pause listening (like a button or smth) then it'd be great
I use Google Assistant a lot and that is a problem. Implementing Push-to-talk here is such an ingenious move, I can just hold down the button until I'm satisfied.
I now wish Assistants on phones at least have a Push-to-talk option, I already long press the Home button to open Google Assistant, can't be that hard to make it Push-to-talk.
Exactly,
It could have a setting/mode where it would only reply when you said "stop" when you were done. Like the old school telegraph messages.
@@middle-agedmacdonald2965 yeah basically smth like that, minus the necessity of actually saying stop.
i was thinking more like two buttons: a "start/stop listening" button, and a "pause/continue listening" button
we already have that for audio/voice recording, so why not implement it here too?
Use PI, there's a button to keep talking. Also, often I've made mistakes when talking and it's auto corrected them.
Wow, you’re on a roll this week ! Well done!
Nice video and nice track! When should we expect the full version?
One thing that wasn't mentioned in the video is that Rabbit is partnering with Perplexity, which allows the R1 to access live-time information. The catch is that this is a subscription, which costs $20 a month. Only the first 100,000 devices will include one free year of access.
Sounds like a fair price if it works as stated
My goal is to become fluent in interfacing with AI so even if the device is made obsolete in a year I see this as an opportunity to experiment, and with the Perplexity bonus it's an easy financial decision since I get to learn both systems.
@@CedarSpringWolf Perplexity is not even an AI it's just an AI skin. That functionality can be done for free. They are basically reselling GPT4 with search function included, which is easily done with other products for free.
Actually, Perplexity and Rabbit have confirmed that the Perplexity Pro service will be free on all Rabbit devices forever, but they are also gifting the first 100,000 purchases with a SEPARATE free one year subscription to Perplexity Pro for all other devices such as Android and PC.
@@Slav4o911But they are offering 600 usage per day while in Chatgpt Plus you get around 300 usage per month
That’s what we call “ solution searching for problem”
One thing that pops out to me with booking flights. For those of us who use loyalty programs, I can imagine these assistants absolutely smashing say Qantas Frequent Flyer with a million different requests, crashing it, as a thousand customers try to find hidden away reward flights.
finally another episode.... They're so rare from ColdFusion, but I love so much...
There is no way people will prefer this over an iPhone; also, one software update from Apple, and their business model could be gone.
What if they don't charge 30% on every transaction charged within their ecosystem? Or charge $200 USD for 8GB of ram?
@@SupraSavwhats ur point?
Cant wait for them to call it "evolu-ai-chromotech" or some funny word so that people can think its something new and revolutionary and charge $2000 fo rit.
@@2COLD118apple boohoo😭
Hardly
This could EASILY be an app. NO NEED for TWO devices that will get tossed ina couple of years DOUBLING our E WASTE!
This, also this rabbit thing has strong scam vibes. Like does anyone believe the response time here, or the size of the form factor? Its obvious bs.
@@banme2784 its common for scammers to arise when disrupting technology is invented.
I can sort of imagine a use case for blind people, especially young kids... Tho every blind person I know would probably prefer to just use their phone 😂 Be My Eyes has already started using AI for this exact purpose
@@banme2784 Naa, the company teenage engineering already makes headphones, music synthesizers, keyboard etc, but this is their first software heavy product.
It can’t. Apple or Google won’t give you enough permissions to launch such app.
Put it in your pocket and add a bluetooth earphone-mic-cam for the visually challenged. It should also have a phone-style vibration motor. This would be so useful for people who need to read labels on pill bottles, canned food, or need to know the color and shape of a pill that fell on the floor, etc. Tell you the best placement for a plant near a window. Tell you if your plants are suffering. Tell you whether your socks match. Tell you if you need to take your insulin. It could give more independence to people with a number of physical challenges.
for most use cases i think language is perfectly fine. most of the time im just googling a question. the camera integration with being able to ask questionsis extremely powerful, i could see old people loving this as their daily driver if nothing else, or people with impaired vision. children could learn with this without being dropped into the internet with all the good and bad it has to offer. this is the next iteration in interface for sure.
I think this kind of user interface will just be an optional feature on every smartphone. You'll be able to do anything the Rabbit can do just by saying "Hey Siri" or "Hey Google". But then you'll be able to interface with your smartphone in the traditional way if you want to as well. It'll just be optional. No need for a separate device.
Many people dont trust google or apple like that
@justas9541 those people are using a Google or Apple products either way, what difference does it make how they interface with it?
Wouldn't an 'iPhone of AI' just be an iPhone with an AI app installed (or part of the OS) rather than another device ?
Dedicated hardware
He mentions that in the video.
It's just a mobile device with an app
Absolutely agree. Really dumb devices. What iPhone and the App Store has shown for over 13 years is that with enough powered hardware, you can emulate any software from games, compass, mirrors, whips (remember?), translators word processors, spreadsheets, voice recorders, MP3 players, video screens, movie theaters etc to AI assistants (from Siri to ChatGPT) so why oh why you would need an additional ugly device with low res screen, rotatory camera (a mechanical problem waiting to happen) of low resolution? why?
Not if Apple has proven with its data collection model that it can't improve Siri because its phone users are unwilling to train their data models. I sell these phones, and every day I watch iPhone users skip sharing data with Apple, while Android users are like "Eh they probably collect this data anyways, might as well get the benefit of building a better product "
Realy interesting video. I realy like how balanced it is between interesting facts, advantages of the device and also disadvantages.
Everything you put out is perfection!
I think the most important aspect of all these devices is that they're essentially "personal information sponges", so we need to have some sort of personal data ownership in place, say if you wanted to switch from Apple Siri Plus to Google Assistant AI.
Apple and Google are already personal information sponges. No change there.
we just need offline such AI llm's. It is already possible with Phi 2 3b models...or even much better like mistral 7b currently.
Except you have to consider the value of a $200 product without a subscription that has already access to better AI and which will be upgradeable in the future at no cost.
Enjoy your vacation that is 3x more expensive compared to booking it yourself and takes you to the most ridiculous and clichee travel plan. This example gave me literal chills down my spine, what a dystopic idea.
Was thinking the same. What kind of dork would want this? The sort of dork you see queuing up outside the Hard Rock Cafe I guess.
an improvement for boomers who would've paid 5x as much and just visited the beach 😂
It's supposedly doing exactly what you would, go to kayak or some website and search cheap prices.
And oh no, you saw the jewels in the tower of London, how cliche. I looked at the local pissing well.
You sound like a hipster
man I would be so mad on this device if it would make me miss something important to visit 😂😂 I would blame it and shame it
People are pretty smart. Have more faith. If the price is too high, most of us won't use it. Plus, a good AI knows how to search for the best price, especially if you include that in the instructions.
One example I see could be a photographer in a studio taking pictures of products with a camera tethered to a computer. A one man show could do the setup and then tell the R1, take a picture with x variable changed. Take a picture with more exposure, or turn on the flash, or whatever variant. Then automatically scrub the pictures that look good and resize them.
It needs to be built into your head, use your inner voice for control, use your eyes for visual input, use all areas of your brain for output, plus the internet/cell network.
As a developer, what I think about is the potential for generative ui. as an example, in the Music Player being able to tell the AI i want to have method to quickly add this song to a specific playlist. Can you create a button for that and from then on, in the Music Player, you like, you have this button that you told the the ai to create for you. That's really cool to me as a concept.
Amazing idea
sounds good doesn't work meme
This is exactly what AI hardware needs if they want to be successful
"Play my favourite song" seem so dystopian to me. I feel AI will interrupt the process of self reflection and growth, keeping people passively moving through life
Or a way to make help more excesible. Therapy is super expensive in America and some people might find machine therapists more reliable. The accessibility of counsoling anywhere you go sounds like it can be a great novelty too.
killing independent thought and the entire thing about being a human being
@@TheJanny The point of consoling is that you have a real human empathize with you and share your pain. An AI can't do that. Instead of spending all this money on actually fixing the problem, we're going off after a solution, that doesn't fix the problem.
Most of the human population doesnt do much thinking anyway.
@@SaliferousAnimationyoure assuming it wont work
This is exactly what I would want, some people may want do their own shopping and picking each detail themselves, others like me want the most convenient option. The cheapest non stop flight, the most visited sites to see? Sounds like a dream vacation for me!
Sure, but it might suggest you, here the cheapest flight from LA to NY: $5000 or $300, 5 stops 😂
I love this channel! Great job!
Watching you for so many years. Your research is just different level. Thank you ♥️
Appreciate it, glad you found it useful!
@ColdFusion, so is your music playlist.
The initial Humane "commercial" seemed cool. Thanks for busting their chops, that is interesting.
Looks amazing
Bought the R1 today. Looking forward to using it.
The menu one could be fixed by having the call to action "I wanna order food from [restaurant] but I want to check out it's menu first" with the response of "Sure thing, just let me know when you're ready to place the order" and displays the menu of the place (restaurant) you want to order from.
The fact that the Rabbit didn't even try to find a coupon for Pizza Hut tells me there's still work to be done. I need full integration with miles programs for airline tickets too.
Best I can do is top 3 Google results
Very interesting! Thanks for sharing it. ✌️
you had me at "talk button"
I remember showing this to my dad, and he responded to this saying that this would never succeed, because this would be only good if it was usable on an app.
If they don’t make an app having all this stuff they did for the Rabbit R1 then there’s going to be someone that will attempt to replicate that or improve it beyond what they’ve did into app form, and get better songs of praises than the Rabbit team.
Remember when folks didn't think of the app paradigm until Apple popularized it? This will bring a social paradigm and user interaction shift as well
@1anre That's a hell of a task to achieve with new device. Yes IPhone kind of did it but it's pretty much been one of a kind tech product launch so you would need to bring some crazy benefits for the end user to replicate it. What are those benefits? Why can't you achieve those benefits with current paradigm?
I don't think I've ever commented this before, but you have such a soothing voice! I had a really rough week and listening to you talk tech while doing the dishes just helps make things feel alright... So thanks for making my day a bit better by doing what it is you do 😊
ok ok
i agree he has a sexy voice
Love your work ❤ Can you do an episode on Optimus and other AI robotics including FSD?
i refuse to carry one more one device, and that's my phone
I think this device might be useful for a person who isn’t very good at navigating multiple apps/UI. I think of my mom who has a smartphone but does not know how to navigate all these apps as some don’t have native support for her language which is Spanish. She loves using Google Assistant but it’s not good with tasks.
I can see her using something like the rabbit for an Uber ride, trying to find something on Amazon, etc.
This is a good point! When I was working in China, I've often had to order products on Taobao, or use apps that were not user friendly to non-Chinese speakers. If I could have asked an AI four years ago to find something I'm looking for and parse the product information for me in plain English, it would save me much hassle. Hell, even here in Japan and using Amazon I've had to utilize a lot of guesswork and now and then I still mistakenly buy the wrong product, for example I've accidentally bought toilet cleaner when I needed liquid plumber to unclog a shower drain. I'm looking forward to when AI can translate something I've said and put it in another language while keeping to the spirit of my meaning. Mind you, an assistant like rabbit could just run off a smart phone, the devs probably should have made it as an overlay to a phone OS, maybe with some custom hardware.
19:07 I imagine Rabbit R1 being used similarly in the Syfy movie Her from ten years ago
Just not as dramatic as an AI love affair between Theodore and Samantha (played by Joaquin Phoenix and Scarlett Johansson) but as a personal AI OS assistant.
Something for all of your smartdevices, sorting and eliminating countless emails, you know... a true IOT assistant... something even greater than what Steve Wozniak would have ever imagined 🤔
19:16 This concept is on the right trajectory but I wonder if this rabbit device are to be use in conjunction with other devices. I'll love to see it in action to know the feasibility of it functions. If rabbit 1 can execute voice commands in a practical way to get tangible results then this is a solid device.
great contet thanks, I ordered a rabbit just because it is compelling to try it out and be part of the experience.
I remember your old video proving how nice you could shoot 4K video on a Galaxy Note 3 and many other amazingly edited videos of yours, "ColdFusTion" youve come a long way brother, keep going!
Those Galaxy Note related videos were the reason why I subscribed his channel
One thing Rabbit got better than any smartphone is that rotating camera. Imagine it could just turn upside and there's no privacy concern because the camera can't see you or your surroundings.
"Don't think... Just consume product and get excited for more product" has never been so literal
I loved the "Is AI powerd you know, AI powered" like that mean something in a gadget that in the best case is as good as my phone
Still the biggest limiting factor is the need to voice everything. Imagine doing all that on a busy train commute to work. Not only will the device likely pickup other people talking, it won’t hear you clearly, but all your personal information or discussions can be heard by other people. That’s the R1.
The other humble one that projects into your hand is basically the r1 but done really badly.
My guy have you heard of generative ui
This is an ai device, you could tell it you are in a busy train and you would want it to display a keyboard from now on, it understands you and displays a keyboard on the screen that you could type with
This is not a traditional smartphone, interaction limited by only what it was programmed to support. This is ai hardware, you teach it to adapt your needs. The whole point of ai is that you can teach it
A lot of people in the comments here misunderstand how ai works
@@bluetoothspeakergamingHAHAHAHAHA he actuälly things AI does everything LMAO imagine defending this product lol so embarrassing
Someone should make an AI device that's packed into an emergency survival backpack. It could basically save anyone that isn't familiar with getting clean water and firemaking or identifying plants and mushrooms.
Also I kind of hate that this might become a thing (because media uses this trope alot), but AI in space travel might be incredible. A good AI could be better that asking Houston for help, saving time in complex problem solving and when space travel becomes more than manned craft orbiting the Earth, astronauts wouldn't have to wait potentially days to get an answer to their problem.
There is a lot of potential with AI, if we are creative enough.
yea, TARS would be amazing ;)
The survival kit idea is a solid use case, but whoever builds this would need to figure out how to fit a GPU or custom made hardware that will do the processing locally, because of inevitable connectivity issues.
I'm sorry, Dave. I'm afraid I can't do that. 🤖
Lol good luck making that waterproof, durable, and solar powered.
AI in space travel is already here, it's just not humanity using it
I preordered one. I'm actually keen on seeing if I can use it in my electronic music making process. Not sure how yet, but oftentimes such broadly applicable devices yield unexpected uses.
I love it personally. On the fly language translation between two people conversing would be great. When I was in Shanghai using the current bunch of apps available on my phone they were very helpful, but clunky.
It's a matter of time before Google updates (and maybe renames) Google Assistant to be able to do these same things. This might just kill the R1 in it's cradle. It's a bit more complicated than that though, because Google & co. rely on us using their services. I think a few months at most before we see a version of that from Google
Congratulations to Google when they do that. Thank you for your future-Google analysis. We're all better off for your prediction.
Get real, google assistant sucks ass. When im driving on android auto, i cant even have it add a song i like. Or adding a speed trap. Literally any simple task it cannot do. Its not an ai and not even close.
Google AI is very weak for some reason... I mean their Bard chatbot (*not their overall AI capabilities).
I am pretty sure once you train Rabbit on a certain app, you can share that with other users. So eventually it will reach the point where it just works with every app.
Don't give them more Data Collection evil ideas.
Not only you'll give them your voice, your order, your habit,
We also will be drowned in Captcha in no time
rabit will be the biggest AI scam.
Which will go completely against their objective which was to reduce app usage😂 this is a trash product, end of story.
We probably already did with many other applications @@jensenraylight8011
You're right. And learning the architecture of an app doesn't rely on information of the user.
And after 1 user had to sacrifice a couple more seconds than usual, the app blueprint can be send to all other devices so the other 99,999% of users have a seamless experience.
I don't know what the guy above me is talking about as "app usage" isn't the problem, its app usage by people which they want to reduce. Which definitely happens if only 1 person's device has to encounter the bluebrint in order to make the experience seamless again for 99.999% of other users.
Thanks for that excellent review. . I think AI glasses with a bone to skull microphone 'earbud' (?) [ for flawless voice translation] and a downsized phone with the main AI processor and screen , for when we don't want to wear glasses will be the natural outcome. Its the earbud adaption that will create the real breakthrough.
The only way i'd use something like ai for would be as an assistant in trying to design a highly efficient yet easy to build generator and other fun electronics and motor hobbies. Also having an actual editing tool for my books because everytime i use google or libre office's spell/grammar checks i always miss something because they don't catch everything which is wild to me.
The playdate which teenage engineering also built is starting to look alot more appealing now
Even as a kid I knew Pokedex was just a made up device that can't exist in the real world... until now. The future is now Old Man
This old man doesn't see a good use for the R1. Unless you are new in town, it's likely you know where everything is.
@@toonman361 that's why Google maps is so unpopular and the many when asked, they don't say they want a fully capable assistant in their phone which they can speak normally too.... Everyone already knows everything about their town and they never leave.
Your life must be boring @@Fastlan3
Hell yeah. It's a Pokedex but the Digimon pilled brain of mines want to think it's a Digivice. I can finally live out my Digimon Tamer dreams I had since 2000
Aww this is a sad story, thats not how life should be but each too their own@@Fastlan3
This is actually a very utilitarian product. It's the combination or mish mash of hardware design, AI and pricing that makes it useful. Considering that the only real limitation is the AI model, this thing could have significant legs as long as it can be upgraded over time.
It could be analogous to the beginning of the PC market in the 80s, where you have the hardware in place but the software is in its first iteration.
In my AI class last semester one of my classmates had made some models that could order food and browse the internet for you. You could see the mouse moving on its own. If it didn't know how to do something then it would learn from him. Not sure if it sounds cool to y'all but I thought it was fascinating. It might have been some simple automation or some AI thing I forgot but it's related to this video since it would interact with applications for you. Cool af 🔥
I guess this is why Apple let the pin guys go, it’s such a terrible device to the point that I feel embarrassed for them.
When I first saw it I was honestly surprised they think that anyone would be interested in such device.
Watching them try to explain why their device was needed/better/useful just felt like the guy was trying to regurgitate as many buzzwords as possible without saying anything. It reminded me of the crypto bubble tbh, no actual explanation just buzzwords.
we dont even know if its a terrible device because they have no idea how to explain what the hell it does
@@craggolly well judging from what they’ve presented. Will I am years ago had a watch that was similar. Just a bundle of buzzwords that was truly awful.
delusional science fiction fans making useless childrens toys cosplaying as tech from blade runner…. At the cost of their careers
The best part is the ability to learn to do the tasks exactly how you want. Imagine you train AI to "make an invoice" or something.
You can't train the AI to do anything unless you have a ton of hardware.... you can "fake" train it through conversation, but that's not the same.
Lmao. True
I can't get enough of this. I recently enjoyed a similar book, and I couldn't get enough of it. "The Art of Meaningful Relationships in the 21st Century" by Leo Flint
I had been a skeptic of R1 for a long time until this video. R1 will be useful as a voice-controlled general-purpose shortcut device. Just think about all the things you do repetitively with your phone that require no real thought just going through a lot of menus. Looking up bus schedules, ubering from known places, ordering food, checking emails from X sender, etc. Imagine being able to create shortcuts to things you do daily that can be used by voice! I'm still a skeptic about asking to book flights or making decisions you do only occasionally and require careful judgement.
Ultimately, the conversation must be internal for these ai agents to work. Otherwise, you only use them at home or when you are alone. Imagine a cafe full of people using them withou headphones.They have to consider the social context for the everyday person to get widespread uptake.
The same argument was made back when people started having cellphones. Having to hear someone's one-sided conversation (or two-way if speakerphone) was seen as intrusive 25-30 years ago. Today you wouldn't even think about it.
@@oldmandoinghighkicksonlyin1368Voice assistants has never really taken off, speaking on the phone is different from talking to an ai when you can use your phone discreetly
@@oldmandoinghighkicksonlyin1368 Today I *do* think about it. When someone's using their phone in a public place, it's hecking annoying. I certainly have no intention of subjecting other people to that annoyance either.
Within 5 minutes you can clearly say it falls under the gimmick tools category
I wouldnt be too sure
If the hardware enables this then how big is that hardware physically?
The device seems quite small and if you would add that key hardware to top tier phone it would eliminate the need for screen/camera/battery/mic/speakers/body since the phone already has it. I doubt it would take that much of physical space to just add that hardware to a top tier phone and an extra button to activate it directly with access to all these parts of the phone.
I believe this can be implemented as an app on a phone or inbuilt into a phone. What is strange is that it has a scroll wheel when he uses a touch screen, what's the point?
Whatever progress is made in the space privacy concerns should always be at the fore front ,AI businesses should not be allowed to run rough shot by treating our daily lives ,as opportunities to gather more data.
You summarized that risk well
That battle was fought and lost 15 years ago.
Offline a.i would be nice
6:15 lets talk about th3 fact that the he said many things in his voice prompt but no verification of those requests were provided by the product. He also just couldve said, can you help book a trip, i need a flight, car and hotel. And the same exact results couldve shown up and we wouldnt know it. That prompt was just an effort to make the ai seem advanced by actually listening to what specifically you want, when in reality its just giving a generic trip booking result.
It's probably using Perplexity with GPT4. So it's totally possible, but also any Smartphone could be able to do it with an app, so no additional device needed.
I think no device can replace a screen with HMI, best possible way is to integrate this sophisticated AI model into a phone. Humans love to look at things more than just voice control, I would have liked more info on the screen specially for the flight booking use case.
I know Google Assistant is rudimentary compared to MMLLMS, but you've been able to integrate app control with the Google Home app for years. And you can assign complex tasks to custom commands. The R1 seems to combine the Google routine functionality with an LLM, which Google should've pioneered months ago. They certainly have the resources. But I guess they need to get their LLM sorted first.
Actually devices like Rabbit would take the effort out of older generation Gen X and Gen Y who find the clunky UI/UX difficult to comprehend on board with the smart app capabilities
The way they're doing it is actually genius. Sell 50k units to enthusiasts who like the weird form factor, this also raises some money and gets your name out there. Have them train the device on every popular app and website. The release a more normal mainstream model next
What you're actually wanting to tell us is "I'm a genius". You're not
@@dariusus9870 having a bad day?
Last week we received an e-mail mentioning we are already 100k enthusiasts 👍
Or... Sell the company to a major company for 10000000 times more than it's worth.
@@gofres This is prob their main goal
In theory, natural language input and actionable output is a winner. In case of the R1, I think they made 2 mistakes. 1: Natural language input shouldn't be limited to just voice input (it needs text input for both accessibility and silent operation). 2: If we need (almost) no GUI, that doesn't mean you can get away with a tiny screen. Because we still want to experience the content output (video, photos, text and more) to its fullest. Still ordered one, because I admire the thought process.
This reminds of a pager which created a hype in the initial days. As time passed hype faded away. So i do feel the same thing will happen with this device
The race to skynet is impressive!!!
That's insane, I just finished watching the video and back of the mind thinking that what apple could do with their already capable hardware and their apple way of doing things. Then Steve Wozniak's 2013 part dropped.
Yeah, I don't see why we need a separate device specifically for AI integration. It's cool that people are looking into it, there may be a use case we can't predict where a separate device is ideal.
Honestly I loved the idea, nobody presented anything similar so far. Nadella knows it, and probably, like many other companies, Microsoft is already working on implementing the same model in Copilot.