7:18 "Akshually", Gates said AI was simply the second ever demonstration of "technology that struck me as revolutionary", not the "second most revolutionary". He didn't rank them.
thank you for the clarification - I can't believe I missed that in my research! I do think the broader point stands (one-dimensional UI has a hard time competing with multidimensional UI), but I may need to make a follow-up video to clarify this at some point. appreciate you!
@@uncoverage oh yes, your overall point is still solid. Your video was one of the more thoughtful critiques of the humane device I've seen so far, and I loved the way you presented it. I was just in a nitpicking mood. 😊👍
I hate screens SO much; To free myself OF them… I strapped three to my face… It’s done wonders for my screen time how incredibly and ironically beautiful is that
ha!! I definitely feel what you're getting at - I think the way I've put it before on this channel is that as screens have gotten closer to us, they more and more isolating they've gotten, right up until the point that they’re basically directly on our eyes. it definitely feels like we're at a point where it's both peak isolation when I wear a headset *and* peak connection with the real world, when it comes to MR. a fun paradox :)
@@thomasthemarstrain2141 Apple Vision Pro gives you the best of both although I am biased and it is my first headset ever I wear it all the time and I found it to be dope and awesome. I definitely recommend it and if you get a chance to try at least even the free demo go for it. The argument of all my Apple products already do that… That’s the point it’s a new way of doing everything that you already do and then some while having your own personal heads up display and your hands-free… I’ve found that I can juggle as well as drive. That’s my experience not a lawyer not a doctor just a guy who loves tech
ha! funny enough, I had such a hard time making the argument against audio because I absolutely love audio as a medium. my primary content consumption has almost always been podcasts, so for me, yes! radio is the future of television!
I agree with what you've said and I think there's more to why screens aren't going away. Specifically, the three biggest issues in my eyes are the limitations of their reference resolution, the lack of "agentic" AI, and the importance of privacy. I'll quote the feedback I personally gave to Collin from Humane, "I think the major issue the AI pin faces is reference resolution. What you are referring to is explicit on a phone or screen when you click or tap, but implicit on the pin. You have to fully trust it and cannot double check any of your inputs, nor can you easily edit them. You and the pin must work 0-shot." The benefit of the screen is direct control, immediate feedback, and precision: When you take an image, you can view it immediately. When YOU click on something, you watch what happens in reaction immediately - it reacts to you. The feedback is instantaneous, so you can make micro adjustments as needed to get the desired effect in real time. Screens let you verify accuracy immediately. You can quickly and precisely communicate your desire to other devices through screen/cursor. Imagine ordering an Uber (you can actually do that with AI pin) and it's going to the wrong address... except you have no way to verify that! Contrast that to the AI pin, where you have to say your request, then wait for it to finish parroting back what you just said to see if you have to right it. The pin has inherent ambiguity and does not know what you are explicitly referring to. I'd go further to say that inputs are more like suggestions. It's interpreting your words through an LLM that cannot intuit the world. In certain contexts it's fine, like with voice assistants, but it isn't optimal when there is too much abstraction. These models do pretty well when they have their bearings, but their internal world models fall short in novel situations or are given ambiguous context and they hallucinate. Those hallucinations make the AI not just useless, but actively unhelpful; it's extremely frustrating to be on the other end of. They kind of get it, but they also don't. It leads into the second point, the lack of "agentic" AI. On phones, we rely on apps being intuitive, but mostly ourselves getting things done. Agents are becoming more sought after and worked on in the LLM space, but they aren't nearly where they need to be for a device like this. An agent would need to be able to complete multiple chained tasks accurately and completely in a way that we trust it without supervision. After all, the pin fundamentally requires you to relinquish all control to the LLM since there is no screen. Unfortunately, even the smartest models are not ready to become full agents as they are still to sensitive to the context and prompts. Another thing I want to touch on is privacy. Our phones are the most private device we own. It's with us in many intimate moments, it contains our daily correspondences, images/video, our habits, our searches, etc.. you get the point. You cannot have anything private with the pin since it has to read it out loud. Yes, you can wear headphones, but that means your input will still have to be spoken without a screen. Going back to that Uber example from earlier, IT READS YOUR ADDRESS BACK TO YOU ALOUD TO CONFIRM AND YOU HAVE TO SPEAK THE ADDRESS YOU'RE GOING TO IN THE FIRST PLACE. It is very frustrating seeing the cope and oversight that people are having. You don't want it accidentally sending a wrong text, or equally a text to the wrong person. You have no way to prevent that directly. There's always the moment where you have to sit there and wait for it to talk back and there's panic that it will mishear you. Have you ever asked Siri to send a text and it says "Sending (insert message) to (insert name of wrong person)"? It's so frustrating and nerve racking - and you can verify on screen before it sends. For those reasons, we need to keep screens. It's not just more convenient, it's necessary until we can trust these systems to be accurate, accomplish tasks efficiently, and do nothing to harm the user. Thanks for coming to my Ted Talk
"The 5 Point is the first Seattle business to ban in advance Google Glasses. And ass kickings will be encouraged for violators." -- 2013 Sensing versus Surveillance... another market miscalculation? The example of automatically recording a concert performance is problematic because many artist ask fans not to pull out their phones to record.
love love love the concept of sensing vs. surveillance - i tend to think that our current behavior is hugely shaped by the mere presence of smartphone cameras, and that only gets more intense when more people have cameras available even quicker than they currently do, and even *more* intense when people are constantly recording
I really feel they dropped the ball on selling a viable use-case for the pin. It doesn't seem to replace anything essential yet. Smartphones weren't a thing the average phone buyer thought they needed back in the day. The name itself was a joke because why would a phone need to be smart? Phones were around forever and couldn't be easier: punch a number and start talking. Most (non-tech) people just considered them a new type of 'feature phone'. We'd had phones with mp3 players, cameras, radios etc, and most presumed smartphones were something similar. We bought them because they were phones. They worked as phones. I still know people who use a smartphone like a dumbphone. Calls and texts. The rest of us bought phones and were pleasantly surprised to find computers in our pockets. Humane AI wants to leapfrog gradual adoption and go straight to cultural icon status without earning the transition. It's a pity because that's a neat little device.
yessss, you nail it here. i talk constantly here about the smartphone is fundamentally different than other computing devices from the past - and since!! even iPad struggles with the Human Ai Pin-esque “what do i use this for” issue
29:35 I think you were struggling to remember the word signifiers possibly? There’s a line about how the device doesn’t have any affordances and that screens allow you to have them because of the UI. Technically and referenced as often confused in Don Norman’s human center design. You’re referencing the issue with the signifiers to the affordances of the pin. The things that can do is the affordance. The signifiers are the things that lead a user to discover the affordance. But your point still stands. I just wanted to make a note for anyone interested in this to checkout Human Centered Design where he defined these terms and Don’s legendary book and word. (First human to have the role with the title UX and it was for Apple!) also I enjoyed the door handle analogy but the push and pull analogy is my favorite in his infamous door analogy in his book. I got my office to change their doors for our bathrooms because of it 🤣
whoa, you've now sent me down a rabbit hole trying to distinguish between signifier and affordances - it looks like Norman added "signifiers" to his revised edition in 2013? I used the word "affordance" because that is what I was taught in college - but I can see how a distinction between the two could have implications for design. also, I didn't know about the UX/Apple thing! do you have a link for me where I can find out more?? thanks for the thoughtful comment :) hope to see you the comment section more!
@@uncoverage yah I believe in the revised version he specifically mentioned how we had to add it because it was being confused with his interpretation of affordances. There’s a few podcasts where the Apple/UX things come up. I gotta find links but one was a recent once he did where he’s now speaking for economic human design (thinking of implications on environment, broader issues with mass adoption, etc.) And my pleasure! Really enjoy ur discussions on these topics! I think the nuance is missing in most tech coverage/opinion pieces on TH-cam lately and I’m glad you are filling that gap. I feel like I wish my IRL friends cared hearing me go on like you do sometimes cause I could go off! 🤣 Oh and I’ve watched a handful of your Vision Pro videos. Since you add clips for context, there’s a really good clip of Steve Jobs that I think could be an interesting topic to explore. He talks about VR/AR/Glasses. About how when it’s like the “headphones for video” in the sense that video is better than a theater like some headphones can be better than speakers. With Vision Pro I think the best (maybe only for most) use case right now being consuming video content on huge screens better than a theater, I think it’s super interesting and relevant and speaks to how Apple would have/maybe should have marketed a bit differently. You can find it on TH-cam shorts let me get a link too. I’ll see you in the comments more!
I wouldn't get the Humane thing just based on the TED presentation alone. The melodrama of his voice with the background music was like a bad skit on SNL. Why would anyone want to be limited to a device that you have to talk to? There are so many situations you don't want to be talking to your computer.
They say it's a standalone device, but it feels like not one of Humane employees actually used it as one, because basic functionality is lacking. It should've been the "Humane Pin" not the buzzword "AI Pin," as in: it should have focused on being a wearable that doesn't get in the way, processes things locally and sometimes helps you with "intelligence" queries. Not even the device itself knows what it should be doing. It feels like a Light Phone done wrong. I do think that Apple is coming with AI stuff and I hope it's mostly in the device, not in the cloud; Humane says it's "safe" but would you give all your personal data and current interactions to the internet?
interesting - I'd love to hear more about this. to be clear, I don't think they're aiming to get rid of screens altogether, but an utter minimization of screen time seems to be their primary goal, to the point that they compromised the product to achieve this idealized vision
Really like what you had to say but found myself very underwhelmed by how far away you seemed. Ended up zooming in 7x. Please consider a closer angle next time.
I do think it's easy to go far in one direction or the other when you make claims about job replacement, but the argument that has resonated most with me about AI is that it is a very fast, very rote intern.
Humane founders watched too much Star Trek. They basically created a shody version of the Star Trek communicator with visual tech included, without considering the fashion aspects of the device. The Meta Raybans or the Apple watch seem like a better UI, at least in terms of fashion and probably in performance. If those can't do what Humane can do now, they soon will.
Let's also forget the hypocracy of going on a Ted Talk to sell this thing and not just doing an audio only podcast. It invalidates the premise that we should have a screenless experience.
@@uncoverage it just feels like an anti-technology device. My second statement is about if this is your main device and the battery is so poor it’s gonna be noticeable how inconvenient it is.
![PURPEECH - นี่ฉันเองคนที่... (It's me) [Official MV]](http://i.ytimg.com/vi/NuhHd0hOOg8/mqdefault.jpg)
7:18 "Akshually", Gates said AI was simply the second ever demonstration of "technology that struck me as revolutionary", not the "second most revolutionary". He didn't rank them.
thank you for the clarification - I can't believe I missed that in my research! I do think the broader point stands (one-dimensional UI has a hard time competing with multidimensional UI), but I may need to make a follow-up video to clarify this at some point. appreciate you!
@@uncoverage oh yes, your overall point is still solid. Your video was one of the more thoughtful critiques of the humane device I've seen so far, and I loved the way you presented it. I was just in a nitpicking mood. 😊👍
Very impressive improvisation speech skills. Just a pleasure to listen
Feels natural to listen to. Organic. 🧠♥️🫁🗣️
aw, thank you!! 😭
I hate screens SO much; To free myself OF them… I strapped three to my face… It’s done wonders for my screen time how incredibly and ironically beautiful is that
Wdym? Do you use VR?
ha!! I definitely feel what you're getting at - I think the way I've put it before on this channel is that as screens have gotten closer to us, they more and more isolating they've gotten, right up until the point that they’re basically directly on our eyes. it definitely feels like we're at a point where it's both peak isolation when I wear a headset *and* peak connection with the real world, when it comes to MR. a fun paradox :)
@@thomasthemarstrain2141 Apple Vision Pro gives you the best of both although I am biased and it is my first headset ever I wear it all the time and I found it to be dope and awesome. I definitely recommend it and if you get a chance to try at least even the free demo go for it.
The argument of all my Apple products already do that… That’s the point it’s a new way of doing everything that you already do and then some while having your own personal heads up display and your hands-free… I’ve found that I can juggle as well as drive. That’s my experience not a lawyer not a doctor just a guy who loves tech
@@uncoverage indeed a fun paradox lol
You’re right
So I guess radio is the future of television?
ha! funny enough, I had such a hard time making the argument against audio because I absolutely love audio as a medium. my primary content consumption has almost always been podcasts, so for me, yes! radio is the future of television!
Good job with your audio!!
thank you!
Wow, barefoot man in wilderness deeply discusses cutting edge tech. What a beautiful life
I see myself kinda in 5 years 🐕
haha! nature is very important to me; I'm glad you're in the same boat!
I agree with what you've said and I think there's more to why screens aren't going away. Specifically, the three biggest issues in my eyes are the limitations of their reference resolution, the lack of "agentic" AI, and the importance of privacy. I'll quote the feedback I personally gave to Collin from Humane, "I think the major issue the AI pin faces is reference resolution. What you are referring to is explicit on a phone or screen when you click or tap, but implicit on the pin. You have to fully trust it and cannot double check any of your inputs, nor can you easily edit them. You and the pin must work 0-shot."
The benefit of the screen is direct control, immediate feedback, and precision: When you take an image, you can view it immediately. When YOU click on something, you watch what happens in reaction immediately - it reacts to you. The feedback is instantaneous, so you can make micro adjustments as needed to get the desired effect in real time. Screens let you verify accuracy immediately. You can quickly and precisely communicate your desire to other devices through screen/cursor. Imagine ordering an Uber (you can actually do that with AI pin) and it's going to the wrong address... except you have no way to verify that!
Contrast that to the AI pin, where you have to say your request, then wait for it to finish parroting back what you just said to see if you have to right it. The pin has inherent ambiguity and does not know what you are explicitly referring to. I'd go further to say that inputs are more like suggestions. It's interpreting your words through an LLM that cannot intuit the world. In certain contexts it's fine, like with voice assistants, but it isn't optimal when there is too much abstraction. These models do pretty well when they have their bearings, but their internal world models fall short in novel situations or are given ambiguous context and they hallucinate. Those hallucinations make the AI not just useless, but actively unhelpful; it's extremely frustrating to be on the other end of. They kind of get it, but they also don't.
It leads into the second point, the lack of "agentic" AI. On phones, we rely on apps being intuitive, but mostly ourselves getting things done. Agents are becoming more sought after and worked on in the LLM space, but they aren't nearly where they need to be for a device like this. An agent would need to be able to complete multiple chained tasks accurately and completely in a way that we trust it without supervision. After all, the pin fundamentally requires you to relinquish all control to the LLM since there is no screen. Unfortunately, even the smartest models are not ready to become full agents as they are still to sensitive to the context and prompts.
Another thing I want to touch on is privacy. Our phones are the most private device we own. It's with us in many intimate moments, it contains our daily correspondences, images/video, our habits, our searches, etc.. you get the point. You cannot have anything private with the pin since it has to read it out loud. Yes, you can wear headphones, but that means your input will still have to be spoken without a screen. Going back to that Uber example from earlier, IT READS YOUR ADDRESS BACK TO YOU ALOUD TO CONFIRM AND YOU HAVE TO SPEAK THE ADDRESS YOU'RE GOING TO IN THE FIRST PLACE. It is very frustrating seeing the cope and oversight that people are having. You don't want it accidentally sending a wrong text, or equally a text to the wrong person. You have no way to prevent that directly. There's always the moment where you have to sit there and wait for it to talk back and there's panic that it will mishear you. Have you ever asked Siri to send a text and it says "Sending (insert message) to (insert name of wrong person)"? It's so frustrating and nerve racking - and you can verify on screen before it sends.
For those reasons, we need to keep screens. It's not just more convenient, it's necessary until we can trust these systems to be accurate, accomplish tasks efficiently, and do nothing to harm the user.
Thanks for coming to my Ted Talk
This video is excellent.
thank you!! 🙏
"The 5 Point is the first Seattle business to ban in advance Google Glasses. And ass kickings will be encouraged for violators." -- 2013
Sensing versus Surveillance... another market miscalculation? The example of automatically recording a concert performance is problematic because many artist ask fans not to pull out their phones to record.
love love love the concept of sensing vs. surveillance - i tend to think that our current behavior is hugely shaped by the mere presence of smartphone cameras, and that only gets more intense when more people have cameras available even quicker than they currently do, and even *more* intense when people are constantly recording
I really feel they dropped the ball on selling a viable use-case for the pin. It doesn't seem to replace anything essential yet. Smartphones weren't a thing the average phone buyer thought they needed back in the day. The name itself was a joke because why would a phone need to be smart? Phones were around forever and couldn't be easier: punch a number and start talking. Most (non-tech) people just considered them a new type of 'feature phone'. We'd had phones with mp3 players, cameras, radios etc, and most presumed smartphones were something similar.
We bought them because they were phones. They worked as phones. I still know people who use a smartphone like a dumbphone. Calls and texts. The rest of us bought phones and were pleasantly surprised to find computers in our pockets. Humane AI wants to leapfrog gradual adoption and go straight to cultural icon status without earning the transition. It's a pity because that's a neat little device.
yessss, you nail it here. i talk constantly here about the smartphone is fundamentally different than other computing devices from the past - and since!! even iPad struggles with the Human Ai Pin-esque “what do i use this for” issue
29:35 I think you were struggling to remember the word signifiers possibly? There’s a line about how the device doesn’t have any affordances and that screens allow you to have them because of the UI. Technically and referenced as often confused in Don Norman’s human center design. You’re referencing the issue with the signifiers to the affordances of the pin. The things that can do is the affordance. The signifiers are the things that lead a user to discover the affordance. But your point still stands. I just wanted to make a note for anyone interested in this to checkout Human Centered Design where he defined these terms and Don’s legendary book and word. (First human to have the role with the title UX and it was for Apple!) also I enjoyed the door handle analogy but the push and pull analogy is my favorite in his infamous door analogy in his book. I got my office to change their doors for our bathrooms because of it 🤣
whoa, you've now sent me down a rabbit hole trying to distinguish between signifier and affordances - it looks like Norman added "signifiers" to his revised edition in 2013? I used the word "affordance" because that is what I was taught in college - but I can see how a distinction between the two could have implications for design.
also, I didn't know about the UX/Apple thing! do you have a link for me where I can find out more??
thanks for the thoughtful comment :) hope to see you the comment section more!
@@uncoverage yah I believe in the revised version he specifically mentioned how we had to add it because it was being confused with his interpretation of affordances.
There’s a few podcasts where the Apple/UX things come up. I gotta find links but one was a recent once he did where he’s now speaking for economic human design (thinking of implications on environment, broader issues with mass adoption, etc.)
And my pleasure! Really enjoy ur discussions on these topics! I think the nuance is missing in most tech coverage/opinion pieces on TH-cam lately and I’m glad you are filling that gap. I feel like I wish my IRL friends cared hearing me go on like you do sometimes cause I could go off! 🤣
Oh and I’ve watched a handful of your Vision Pro videos. Since you add clips for context, there’s a really good clip of Steve Jobs that I think could be an interesting topic to explore. He talks about VR/AR/Glasses. About how when it’s like the “headphones for video” in the sense that video is better than a theater like some headphones can be better than speakers. With Vision Pro I think the best (maybe only for most) use case right now being consuming video content on huge screens better than a theater, I think it’s super interesting and relevant and speaks to how Apple would have/maybe should have marketed a bit differently. You can find it on TH-cam shorts let me get a link too.
I’ll see you in the comments more!
@@uncoverage Steve clip about vr/ar/glasses th-cam.com/video/bQECSInWVPY/w-d-xo.htmlsi=h0vgjDUpAxl9dOAS
I wouldn't get the Humane thing just based on the TED presentation alone. The melodrama of his voice with the background music was like a bad skit on SNL. Why would anyone want to be limited to a device that you have to talk to? There are so many situations you don't want to be talking to your computer.
i’m glad the absurdity landed with you here!
They say it's a standalone device, but it feels like not one of Humane employees actually used it as one, because basic functionality is lacking. It should've been the "Humane Pin" not the buzzword "AI Pin," as in: it should have focused on being a wearable that doesn't get in the way, processes things locally and sometimes helps you with "intelligence" queries. Not even the device itself knows what it should be doing. It feels like a Light Phone done wrong.
I do think that Apple is coming with AI stuff and I hope it's mostly in the device, not in the cloud; Humane says it's "safe" but would you give all your personal data and current interactions to the internet?
i’ve been very curious about your point as it relates to the Rabbit R1, which arguably is even less private and secure than the Ai Pin
Your dog is beautiful
she is!! thank you :) her name is Miso, by the way 🙏
18:13 graphical interface vs ai voice
1:24 💗
she is so cute, right‽ 🐕
Love your dog. What’s his name?
Miso!
Thoughtful as usual. But I think you’re confusing Humane’s aim to reduce our dependency on screens with an aim to replace screens entirely.
interesting - I'd love to hear more about this. to be clear, I don't think they're aiming to get rid of screens altogether, but an utter minimization of screen time seems to be their primary goal, to the point that they compromised the product to achieve this idealized vision
The third thing that I use most with voice devices might be asking for the weather….
yes!! can't believe I missed this use case
Really like what you had to say but found myself very underwhelmed by how far away you seemed. Ended up zooming in 7x. Please consider a closer angle next time.
oh no; do NOT watch today's video th-cam.com/video/9QAY3_y5iT4/w-d-xo.html
AI won't really replace the employees either
I do think it's easy to go far in one direction or the other when you make claims about job replacement, but the argument that has resonated most with me about AI is that it is a very fast, very rote intern.
Humane founders watched too much Star Trek. They basically created a shody version of the Star Trek communicator with visual tech included, without considering the fashion aspects of the device. The Meta Raybans or the Apple watch seem like a better UI, at least in terms of fashion and probably in performance. If those can't do what Humane can do now, they soon will.
it definitely feels Star Trek-y! i agree that glasses or phones will ultimately be a big platform *for* AI in the short run here.
Let's also forget the hypocracy of going on a Ted Talk to sell this thing and not just doing an audio only podcast. It invalidates the premise that we should have a screenless experience.
ha!
He should have called it the Boomer Pin. 😂
Also no device with 2-4 battery life is invisible.
wait wym??
@@uncoverage it just feels like an anti-technology device. My second statement is about if this is your main device and the battery is so poor it’s gonna be noticeable how inconvenient it is.
Ageism eh? Would love to see you make such a comment to Geoffrey Hinton.
@@Mf_Cooldawg Didn't "Boomers" create most of the technology that has led to AI, and aren't there a lot of "Boomers" developing such tech?