{"id":"p8niHQvrVII","title":"Intelligent Machines in the Physical World","description":"How humans and AI systems collaborate in real workflows today, and where human judgment remains essential.\n\nNEAR is the unified commerce layer for assets and agents.\n\nWebsite: https://near.org/\nNEAR AI: https://near.ai/\nIntents: https://near.org/intents\nChain Abstraction: https://www.proximity.dev/\nSharding: https://nearone.org/\n\nFor the devs\nGitHub: https://github.com/near \nDev Docs: https://docs.near.org/\nDevhub: https://devhub.near.page/ @NEARDevHub\n\nConnect with socials\nTwitter:   / nearprotocol  \nBlog: https://near.org/blog/\nMedium:   / nearprotocol  \n#NEAR #nearprotocol #NEARAI","channel_name":"NEAR Protocol","published_at":"2026-03-02T00:00:00.000Z","thumbnail_url":"https://img.youtube.com/vi/p8niHQvrVII/hqdefault.jpg","transcript_text":"Hello and welcome everybody. We are going to try and be a bit uplifting, a bit positive to finish the day off. Um, feels like it's time for that. Um, but first I want to uh introduce I'm going to let my panelists introduce themselves. So we'll start from the far end. David, tell us a bit about yourself, what you're doing here. Hello, I'm David Scholhoff. Uh, thank you for having me. Welcome to be here and uh I'm with Tobore Life AI. I'm the founder. Tobor Life is a developer of robot applications and a master distributor for unitry robots. Excellent. Tinder. Uh my name is Tinder. I lead AI at Frontier Tower and I also consult in the space. I primarily focus on reducing anxiety for my clients uh with respect to tech. Sure. Increasing anxiety or reducing anxiety? Reducing. Reducing. Okay, that's probably better. I'm Mike Basel. I am among other things the exec director of the confidential computer consortium which is uh part of the Linux Foundation. Uh I'm a uh a techie with delusions of management and uh I'm just here to ask questions and hopefully get some conversation going. I've also been doing distributed systems and stuff since probably before most of you were born which is really embarrassing but that's where we are. So you may have noticed that we've got this this physical world uh is part of the the definition of what we're supposed to talk about and uh we were having a bit of discussion backstage about what we mean by the physical world. So um the vinda do you want to give your take on what you think we should be talking about and then I can disagree with you. Yeah sure. Uh I think the physical world is uh very real like there's physical AI so there's robotics there's that aspect of the physical world and really anything that impacts people in their lives and um is real for them is part of the physical world as far as I'm concerned okay so uh you have a problem with that David are you going to let us roll with that? Um, no, I think it's consistent. I mean, I think people are familiar with AI in a conceptual cloud-based format. And I think with robotics now, it brings it into the physical world to impact us even more. So, sure. Okay. So, I mean, part of the the thinking behind here is that we all know people who are scared by AI, right? And scared about where it sits in the world and how they work with it. My kids are both at university and they're both reading liberal arts degrees and they are concerned that if they touch AI, they're basically going to be kicked off the course, right? Because that's what happens if in in in the way they think of it. But I'm I'm really hoping that most of the people here, if not all of you, kind of take a somewhat different view. So I kind of want to start off with uh with asking the question, you know, are there any workflows you're seeing where AI is actually adding real value right now rather than just R&D stuff? Dinda, why don't you start off? Yeah, I'll start with your kids. Um working with schools to help them figure out uh sh reshaping their curriculum. kind of have a fast and slow mode where in the fast mode you're actually figuring out like the skills that you need this month because the new technology came out and it's very useful for work. Um, and in the slow mode is like can we get the whole degree done in 3 years. Um, community colleges are also taking a different approach to this where they're trying to figure out the most useful uh degrees but so are other universities. So that's one example. Uh certainly everyone here sees the impact of like all of these bots and agents kind of speeding them up and obviously the amount of work then that's coming your way is also increasing and so we have to find the new equilibrium. Okay, I like that. And and David, what about in your world? You say you're strongly in the robotics kind of world. Where do you what give me some of ideas of sectors or usage cases where AI is really impacting that? Um sure. And you said outside of research. Um, so we've seen deployments or working on deployments in security patrolling. So where the robots can go around and look for unusual things that they've been trained to spot also um in construction sites um very good at transporting heavy loads up rough terrain and around um and as well as like wellness checks on people or inspecting. Another application is carrying around um measuring devices. So where so maybe a human had to go around um carrying a um some type of measuring device for radioactivity or heat sensors, now a robot can go around with that same route and record the information. And and where does AI fit in there? Because we've had robots, you know, going into hazardous areas for a while now, right? Where does AI help? What's the what's the the step up that AI gives us? Um a few things. I mean, off the top of my head, one would be it can recognize objects or its environment. So, it it can navigate more easily um in unfamiliar terrain and it can identify objects that it sees and report on that. So, it it enables it to um do missions rather than just be, you know, a mechanical tool. Okay. So, it's be it can be more intentional and it can it can actually have more autonomy than it might otherwise do. Make decisions as to what it should exactly make decisions and report back to you based on those decisions. So it's agentic in some ways, one might say. Very much. Excellent. Sorry, Deinda, you were looking like you had something to say. No. Uh, uh, that's exactly right. I mean, the reason we're use um, we started using deep learning and and all of these different types of AI models is because you can now deal with variability. You obviously have to make some adjustments for the things that you previously knew were deterministic from your software, but now you can handle cases where instead of a 45 minute wait time on Bank of America's like call line, you can get done with something in 5 minutes, right? That's nice value. Um, but it also applies in different areas like you'd mentioned for robotics. Uh if you can figure out obstacle detection and you don't have to use classical computer vision models to deal with every aspect of like a factory automation um then you can at the very least get to your next cycle of automation much much faster um and in the best case scenario deal with a lot of these like corner cases uh without having to intervene. Okay. So there's some positive things there, right? We're saying there's things that robots can do that we can't. I'm also hearing we don't need humans anymore. Uh tell me please that we need humans still David. Um sure. Um very very much so. So we're at very very early stage of this technology and as you've seen other technology waves happen. Um you know there's a lot of rough edges at the very early stage. So while they're deployable and they can definitely um take over some jobs, there's a lot of supervision need, there's a lot of occurrences that happen that maybe um weren't planned where you have to um intercede and then train the robot for those. So this we're again very early stage. The robots have a lot to learn before they can replace those completely. I think that's quite a ways off. Nenda, yeah, when hasn't this question been asked, right? when the calculators first came by, are humans forever replace uh you know um this question keeps coming up time and time again and it's not like we say okay all of human productivity that we have today we're going to stop with that because we have a technology that allows us to do the things that we need to do like Amazon didn't stop at 5day shipping they kept going faster because they could figure out like you know that's what people want and I think the same is true for a lot of these technologies there's no stopping us when it comes to figuring out new ways to either you know create value or to have fun or to build around the world. Yeah. But I I kind of feel I mean we're on the singularity stage, right? Sure. And I kind of feel that there is a there's a bit of a difference here that if we if we as we near the singularity, whatever that looks like, are we not getting to a stage where there's a whole bunch of stuff that really machines aren't going to need us for? And you know, where does that stop? Will there be things still? And if so, what are they? I mean, I've got a few ideas, but what do you think? There's Do you want to go first? Go on, go ahead. Yeah. Yeah. So, there's this really cool niche piece of research that I didn't know about. Um, my buddy at the tower, uh, Clovis, uh, he actually used leads it alongside me. Um, he explained to me this idea of post labor economics, this whole field of research where people are trying to figure out when everything is figured out and a single person can build a billion dollar company, which is apparently happening. um like what what does that leave us? And we have healthy debates about this because I'm not convinced that we'll ever reach a point where we'll have figured everything out. Um if you are saying that Singularity as it stands exists in a mode where it has figured out every single thing out for us and we don't need to do anything else, we're kind of in Star Trek world, right? But I very carefully didn't say that. But that that but yeah. Okay, that's that's a fair point. Um so you don't believe in that or you think it's I think we're so far away from this lines are different right everyone defines AGI different ways uh otherwise we some by some definitions we're there we have been there since like the first agents came about and by other definitions we won't be there for another 20 30 years um so depending on how you define whether we're still useful I don't code nearly as much as I used to directly but I'm writing a lot more coding projects am I still useful I think they're they're of value to me I think the nature of how these things are shaping my life is changing. It's an evolving situation. I don't think we can just put a pin in it saying we're useless at some point. David, are you still useful? Sorry, that sounds really mean. Tell me. It depends who you ask, I think. But um you know, I I think you made some good points. I I people are used to concepts of of robots from movies and television and you know, we're maybe at step one compared to step 10,000 we need to get to. So um I think it's a long time to imagine you know a world without us but even then creativity invention um human ingenuity um you know how we dream things um there's elements that I don't think will ever be created that replace us. Yeah, I I've got ethics and guardrails and factchecking down there with creativity, but I I sometimes wonder whether there's this asymmetry, right? We we talked about how this technologies have always come along and you know, calculators provided a a new way of doing things much faster, much more easily, but just with the speed and the uh the the scale at which you can can do things. I do worry even about stuff like creativity. the uh I don't know if you've been following that people are doing maths proofs with AI but the the AI doesn't get them right it often is giving good ideas towards and I wonder whether we will get to a stage where the you know AI can actually solve some of these math problems but I'm not sure that I expect a world where it sees why it should make some sense David yeah well this is kind is something you're dancing around. But um we're focused on robotics AI and as you know robots are AI in the physical world just manifested and um we think that to get to AGI to get to where you know the robot can function like us and navigate the world that that it has to be you have to put AI in a physical body because without the physical body it's kind of in a dream state like when we're dreaming we don't have our bodies just a we're just a floating consciousness in our dreams. you don't have all your capabilities. There's many things you can't do um and things you don't understand. So AI really needs to understand the physical world to advance is is our belief. Um so getting it into robotics, training robots about the physical world um should help advance that and that's what Tober life is about is advancing it till so robots can take over the world a little faster. So there's I think the argument is that without real world actual interaction physical interaction AI can't actually be getting closer to us and as that happens there are some physical constraints that will slow things down. Am I kind of getting that right or not? Um yeah, I didn't follow the last part but but the physical constraints um will help build an understanding of the world for AI so they can be more useful for us and um and and moves it towards the AGI and yes there'll be phys there'll be limitations in there to protect us. The posetronic brain is not yet upon us but azimoth fans uh anything yeah there's also different measures of what intelligence even means. we haven't really figured this out for ourselves. Like there's consciousness questions, there's intelligence questions, there's all of these layered questions. So I really try to stay very close to like how is this useful in my real life? And right now as an assistive tool uh in every layer, it's extremely useful to utilize AI as an independent tool doing everything by itself. Maybe with agentic orchestration, you can say it's able to do some of those things, but there's still quite a few steps left. Um, and from my my point of view, first they completely forget the idea that the the way that neural networks work, even though we name them that way and they're modeled after the brain, it's not exactly the way our intelligence works, right? So, it's a different type of thing and the amount of data it has at at boot like the scale is enormous. There very different ways that these things work. Um, but even given that it's helping us in all these ways, I think there's we're a few few days, months, weeks, years away. I don't like the days but I you know I I think that's a a fair point and you know the fact that we don't know how our own consciousness works means it's rather difficult to say whether AI is or will be conscious even right it's a completely different framework altogether when you think of like classical deep learning like essentially you're treating teaching patterns and the ability to reprint on those patterns over the course like understand like fairly deep uh ideas is basically we'll give you enough layers we'll figure out there's a depth to this there's different types of patterns that can be learned And maybe that means it's learned. And because of LMS and the latest advances in machine learning, we're saying, \"Oh, we're getting pretty close to this pattern matching. That must be intelligence.\" It might not be, but it's very useful. And so we should use that. Yeah. Unless, of course, our consciousness is just pattern matching. Yeah, that's true, too, David. Sorry. So, um, just touching again on what you're saying. So, voice recognition has been around for 50, 60 years. I mean when I was a little boy there was Eliza and there was voice recognition as part of that but it was very elementary compared to what we can do today of course and I I I see where we are today maybe like voice recognition was 50 years ago and if you project that out like these AI factories that they're building for $50 billion with nuclear power plants next to them and the level of computing power that exists there I think eventually we'll be in a handheld device or in a robot and when that level you know, it might take decades, but when that does occur, I mean, I think essentially they'll reach us when you have that level of power locally um to exploit, but I think, you know, right now we're simulating something that eventually will be the reality of it, just like voice recognition from 50 years ago. So, you said the word exploit, and which kind of brings me to my next question is should should humans be scared and if so of what? I think of a Mark Twain quote um that comes to mind where he said um he said if you take a starving um sick dog and you feed it and you make it well it'll never bite you. And he said that is the difference between dogs and man. And uh I reflect on that that dogs are created by God and you know AI and robots are created by man. So I just think we have to be careful but um you know I I believe in a bright future with them. So defender AI is basically a dog. No, I don't you know you could you could guard rails. Um yeah, you could you can make the argument that we can certainly train certain behaviors into AI and and there's a lot of value in that. Sure. Uh I don't see it as a a dog per se. I think the despair thing is so weird to me because think about all the things you can do now. Like you can take your government website that has this really terrible civics like setup and like you can't find the right form or the right thing that you need to do and you can feed it to uh openclaw or some agent and some orchestration unit and you can get through a problem that used to take a lot of time and was completely unreachable for vast majority of people who are like who can't afford these services and now they can figure out how to like deal in their civics environment. you have transfer learning happening from all these people who are learning to curate all of this knowledge uh utilizing AI because they're forced to use these systems so much like we're kind of forced to use the internet today and the cloud um and that transfer learning is is essentially teaching you how to curate data that's coming from your social media and and so you better understand how to be a curator of knowledge and so all of these wins that we're getting have very clear value so the idea that should be we were of course every great thing should come with like a healthy amount of skepticism and this idea that you should be careful of all the terrible things that can happen, but there's so much opportunity to do good. So, I I I agree with that and I'm also aware that we're speaking to a captive audience of the right people, right? Of all the people who are like to be able to survive and have jobs of some type and living a lot of them are in this room, right? And I I think back to, you know, example of when when sewers started being invented, right? Yeah. uh they were obviously a a a common good for most people, right? But the people who used to take away the stuff from the streets, they lost their jobs and they couldn't necessarily retrain and that is going to happen. So I think at the individual level there will be and we're seeing you know entry-level jobs are disappearing in some sections and I think they're going to come back but Oh okay that's what say? Yes, you agree right? Exactly right. Things will impact us in a poor way, which is why I'm saying it's not all all roses, but we have the tools today to get back up from that. Who who is we? Who is we? All of humanity have tools uh to to get back up from this idea of like, hey, I might not have a job, but you can learn new skills faster than you could have ever learned it at a deeper level than you could have ever learned it. And you can start implementing it today. Like that has never happened either. So there's this balance that happening we as humanity and the civics and the governments and the organizations actually care about their people um which I hope most do um there's an opportunity to really just push forward through the despair part which inevitably happens with all of these changes and I think one of the other changes to come to a point I think is relevant to you David is that quite a lot of the previous changes have been technological orformational changes you know, think of the internet or printing. Um, but this one has relevance and uh instantiation in the physical world through robots, through uh other types of uh automation and AI in ways that maybe haven't before. Maybe that's a thing to be positive about. Yeah. And I I I just would build on what you said about um the opportunity. So, you know, if you don't have your job moving boxes or pushing a paper anymore, cleaning up something on the street, the power that a robot with AI gives you, um, I mean, all you have to do is imagine an application or something that could be valuable to somebody, and you have a machine there that can do things that have never been able to be done before. So, I think, um, um, it empowers people to to create great businesses on their own with these tools and make a lot more money than, but not everyone is a business owner. Not everyone wants to do that or can do that. And we there will be there will be I don't want to use the word suffering. There will be people who do not do well out of it. But I think it's going to float more boats than it that it sinks. And that's that's the hope. And we need I think we have a responsibility as the people in this room to think about the people who it's going to be most difficult for uh to do that. So I we haven't got a huge amount of time left. I had kind of one final thing and I I'm not sure it's even the right question. Okay. but is what does AI do badly now that it's going to do better soon that is relevant to this particular problem? Do you want to take that first? I got to think on that. Okay. Um I have a few things and it's not badly. It's say it again. I have a few things. It's not badly but it's things that we're constantly improving on. Obviously we still need a lot of data to be able to learn a new mo like a new mode of operation. That's a technology that's being constantly improved. Uh you see it at every level in the LLM's world in the U. And I would I would add energy usage to that as well. Energy usage is a thing that we're terrible at right now. And I think um not only can we utilize not only can like we work with machines to like make them better faster. We can utilize AI to make them better faster even faster. Right. There's a song about that. Yeah. Sure. Uh there's so many things that it does well too that we misunderstand and there's a lot of lack of knowledge. Sometimes people use these AI systems. A lot of clients I help, they're trying to use it like a deterministic system. Forget the fact that not everything is deterministic that we think it is, but just the idea that they're trying to use the tool in the wrong way. If we start teaching people how to use the tool the right way as we're improving the tool over time, uh there's a lot like there's a lot to be gained still. Okay, I like that, David. Well, yeah, just as you said, the hammer was probably used to build this great hall we're in, but in the wrong hands, you know, you have broken bones and broken glass if it's in used correctly. So, um I think the training of the tools is very important. Also, you know, areas that um in terms of robotics for AI, navigating the world in the physical world is still a big challenge for the robots. Okay. So um um once that is overcome and there's a lot of work done on it and and training data is key for the AI that people are familiar with there's been all the writings of humanity up till now for it to absorb. So there's a lot of training data but for the physical world there's very little training data and it's being created but um that's going to take a lot longer for the robots to learn how to navigate and deal with our world even just to come and pick up a bottle. So once that happens, we're Yeah. And I I I go further than that because a physical world isn't just these things, right? It's how we interact. It's how organizations work, how communities work, and navigating what that looks like in AI is is a next level problem and something hopefully we can continue to help with. Uh and luckily there's a lot of science fiction in this area. Not all of it positive, but you know, a fair amount of it. Um okay, so we have a very short time left. Um, Dinda, what would you send people out with some hope? Give them give the give our audience some hope around AI in the real world. I think the opportunity to improve your life and do less of the boring things is is immense. I think the technology is constantly improving, but if you were to stop today, you could still get a lot more value out of it. I think every single person here who wants to can build a business. I think every single person who wants to can learn a new skill faster than ever before possible. And for sure, you can also make an impact in new ways that you might not have had the tools or just the resources to be able to do. So, you're empowered in ways that you've never been empowered before. Cool. David, some hope, please. Um well, eventually um robots will learn how to build robots and and repair them and make them and then um they'll make, you know, thousands, tens of thousands, millions of robots. And the robots will eventually then maybe make a city of robots and build roads between their their cities. And just like we do, we don't really ask the animals for permission when we put a road in through a city. So, um you know, we we might look like that to them. But, um again, I think that's a long way off. And I really more echo the um the idea that, you know, we can have um zerocost goods like with robots making everything and doing everything. um we can build a utopia if it's engineered right, if the right thoughts are presented. So I think it's really up to us to formulate the vision for the future that we want and make it happen. Thank you. So my and my last takeaway is that I'm hopeful because you lot are here right you're you some of you stay well we've had more people arrive this people care about this and we can engage on this and I'm hoping that in a year or two we'll have AIs in the audience or on the panel that we can engage with as well that's going to be really interesting when that happens right I know people are beginning to do these talk shows but when we when we get to that stage and we can actually start having these conversations and and seeing how we can interact. That's when I think it gets really interesting. And please go out, be hopeful, do the right things, and thank you very much indeed.","transcript_synced_at":"2026-11-03T04:37:39.264Z","segments":[{"start_seconds":10,"text":"Hello and welcome everybody. We are going to try and be a bit uplifting, a bit positive to finish the day off. Um, feels like it's time for that. Um, but","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=10s"},{"start_seconds":19,"text":"first I want to uh introduce I'm going to let my panelists introduce themselves. So we'll start from the far end. David, tell us a bit about yourself, what you're doing here.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=19s"},{"start_seconds":26,"text":"Hello, I'm David Scholhoff. Uh, thank you for having me. Welcome to be here and uh I'm with Tobore Life AI. I'm the founder. Tobor Life is a developer of","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=26s"},{"start_seconds":35,"text":"robot applications and a master distributor for unitry robots. Excellent. Tinder.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=35s"},{"start_seconds":41,"text":"Uh my name is Tinder. I lead AI at Frontier Tower and I also consult in the space. I primarily focus on reducing anxiety for my clients uh with respect to tech.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=41s"},{"start_seconds":50,"text":"Sure. Increasing anxiety or reducing anxiety? Reducing. Reducing. Okay, that's probably better.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=50s"},{"start_seconds":55,"text":"I'm Mike Basel. I am among other things the exec director of the confidential computer consortium which is uh part of the Linux Foundation. Uh I'm a uh a","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=55s"},{"start_seconds":64,"text":"techie with delusions of management and uh I'm just here to ask questions and hopefully get some conversation going.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=64s"},{"start_seconds":70,"text":"I've also been doing distributed systems and stuff since probably before most of you were born which is really embarrassing but that's where we are. So","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=70s"},{"start_seconds":78,"text":"you may have noticed that we've got this this physical world uh is part of the the definition of what we're supposed to talk about and uh we were having a bit","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=78s"},{"start_seconds":86,"text":"of discussion backstage about what we mean by the physical world. So um the vinda do you want to give your take on what you think we should be talking about and then I can disagree with you.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=86s"},{"start_seconds":94,"text":"Yeah sure. Uh I think the physical world is uh very real like there's physical AI so there's robotics there's that aspect of the physical world and really","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=94s"},{"start_seconds":101,"text":"anything that impacts people in their lives and um is real for them is part of the physical world as far as I'm concerned okay so uh you have a problem with that","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=101s"},{"start_seconds":110,"text":"David are you going to let us roll with that? Um, no, I think it's consistent. I mean, I think people are familiar with AI in a conceptual cloud-based format.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=110s"},{"start_seconds":120,"text":"And I think with robotics now, it brings it into the physical world to impact us even more. So, sure. Okay. So, I mean,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=120s"},{"start_seconds":127,"text":"part of the the thinking behind here is that we all know people who are scared by AI, right? And scared about where it sits in the world and how they work with","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=127s"},{"start_seconds":135,"text":"it. My kids are both at university and they're both reading liberal arts degrees and they are concerned that if they touch AI, they're basically going to be kicked off the course, right?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=135s"},{"start_seconds":144,"text":"Because that's what happens if in in in the way they think of it. But I'm I'm really hoping that most of the people here, if not all of you, kind of take a","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=144s"},{"start_seconds":153,"text":"somewhat different view. So I kind of want to start off with uh with asking the question, you know, are there any","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=153s"},{"start_seconds":161,"text":"workflows you're seeing where AI is actually adding real value right now rather than just R&D stuff? Dinda, why don't you start off?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=161s"},{"start_seconds":170,"text":"Yeah, I'll start with your kids. Um working with schools to help them figure out uh sh reshaping their curriculum.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=170s"},{"start_seconds":176,"text":"kind of have a fast and slow mode where in the fast mode you're actually figuring out like the skills that you need this month because the new technology came out and it's very useful","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=176s"},{"start_seconds":184,"text":"for work. Um, and in the slow mode is like can we get the whole degree done in 3 years. Um, community colleges are also taking a different approach to this where they're trying to figure out the","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=184s"},{"start_seconds":192,"text":"most useful uh degrees but so are other universities. So that's one example. Uh certainly everyone here sees the impact","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=192s"},{"start_seconds":201,"text":"of like all of these bots and agents kind of speeding them up and obviously the amount of work then that's coming your way is also increasing and so we have to find the new equilibrium.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=201s"},{"start_seconds":210,"text":"Okay, I like that. And and David, what about in your world? You say you're strongly in the robotics kind of world.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=210s"},{"start_seconds":215,"text":"Where do you what give me some of ideas of sectors or usage cases where AI is really impacting that?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=215s"},{"start_seconds":222,"text":"Um sure. And you said outside of research. Um, so we've seen deployments or working on deployments in security patrolling. So where the robots can go","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=222s"},{"start_seconds":232,"text":"around and look for unusual things that they've been trained to spot also um in construction sites um very good at transporting heavy loads up rough","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=232s"},{"start_seconds":240,"text":"terrain and around um and as well as like wellness checks on people or inspecting. Another application is","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=240s"},{"start_seconds":248,"text":"carrying around um measuring devices. So where so maybe a human had to go around um carrying a um some type of measuring","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=248s"},{"start_seconds":255,"text":"device for radioactivity or heat sensors, now a robot can go around with that same route and record the information. And and where does AI fit in there?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=255s"},{"start_seconds":263,"text":"Because we've had robots, you know, going into hazardous areas for a while now, right? Where does AI help? What's the what's the the step up that AI gives us?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=263s"},{"start_seconds":273,"text":"Um a few things. I mean, off the top of my head, one would be it can recognize objects or its environment. So, it it can navigate more easily um in","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=273s"},{"start_seconds":281,"text":"unfamiliar terrain and it can identify objects that it sees and report on that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=281s"},{"start_seconds":286,"text":"So, it it enables it to um do missions rather than just be, you know, a mechanical tool.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=286s"},{"start_seconds":292,"text":"Okay. So, it's be it can be more intentional and it can it can actually have more autonomy than it might otherwise do. Make decisions as to what it should exactly make decisions and report back to you based on those decisions.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=292s"},{"start_seconds":303,"text":"So it's agentic in some ways, one might say. Very much.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=303s"},{"start_seconds":307,"text":"Excellent. Sorry, Deinda, you were looking like you had something to say.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=307s"},{"start_seconds":310,"text":"No. Uh, uh, that's exactly right. I mean, the reason we're use um, we started using deep learning and and all of these different types of AI models is","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=310s"},{"start_seconds":318,"text":"because you can now deal with variability. You obviously have to make some adjustments for the things that you previously knew were deterministic from","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=318s"},{"start_seconds":325,"text":"your software, but now you can handle cases where instead of a 45 minute wait time on Bank of America's like call line, you can get done with something in","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=325s"},{"start_seconds":333,"text":"5 minutes, right? That's nice value. Um, but it also applies in different areas like you'd mentioned for robotics. Uh if","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=333s"},{"start_seconds":340,"text":"you can figure out obstacle detection and you don't have to use classical computer vision models to deal with every aspect of like a factory automation um then you can at the very","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=340s"},{"start_seconds":349,"text":"least get to your next cycle of automation much much faster um and in the best case scenario deal with a lot","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=349s"},{"start_seconds":356,"text":"of these like corner cases uh without having to intervene.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=356s"},{"start_seconds":360,"text":"Okay. So there's some positive things there, right? We're saying there's things that robots can do that we can't.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=360s"},{"start_seconds":367,"text":"I'm also hearing we don't need humans anymore. Uh tell me please that we need humans still David.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=367s"},{"start_seconds":375,"text":"Um sure. Um very very much so. So we're at very very early stage of this technology and as you've seen other technology waves happen. Um you know","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=375s"},{"start_seconds":384,"text":"there's a lot of rough edges at the very early stage. So while they're deployable and they can definitely um take over some jobs, there's a lot of supervision","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=384s"},{"start_seconds":391,"text":"need, there's a lot of occurrences that happen that maybe um weren't planned where you have to um intercede and then train the robot for those. So this we're","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=391s"},{"start_seconds":400,"text":"again very early stage. The robots have a lot to learn before they can replace those completely. I think that's quite a ways off.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=400s"},{"start_seconds":406,"text":"Nenda, yeah, when hasn't this question been asked, right? when the calculators first came by, are humans forever replace uh","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=406s"},{"start_seconds":414,"text":"you know um this question keeps coming up time and time again and it's not like we say okay all of human productivity that we have today we're going to stop","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=414s"},{"start_seconds":423,"text":"with that because we have a technology that allows us to do the things that we need to do like Amazon didn't stop at 5day shipping they kept going faster because they could figure out like you","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=423s"},{"start_seconds":431,"text":"know that's what people want and I think the same is true for a lot of these technologies there's no stopping us when it comes to figuring out new ways to","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=431s"},{"start_seconds":439,"text":"either you know create value or to have fun or to build around the world.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=439s"},{"start_seconds":444,"text":"Yeah. But I I kind of feel I mean we're on the singularity stage, right? Sure.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=444s"},{"start_seconds":447,"text":"And I kind of feel that there is a there's a bit of a difference here that if we if we as we near the singularity, whatever that looks like, are we not","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=447s"},{"start_seconds":456,"text":"getting to a stage where there's a whole bunch of stuff that really machines aren't going to need us for? And you know, where does that stop? Will there","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=456s"},{"start_seconds":465,"text":"be things still? And if so, what are they? I mean, I've got a few ideas, but what do you think? There's Do you want to go first? Go on, go ahead.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=465s"},{"start_seconds":473,"text":"Yeah. Yeah. So, there's this really cool niche piece of research that I didn't know about. Um, my buddy at the tower, uh, Clovis, uh, he actually used leads","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=473s"},{"start_seconds":482,"text":"it alongside me. Um, he explained to me this idea of post labor economics, this whole field of research where people are trying to figure out when everything is","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=482s"},{"start_seconds":490,"text":"figured out and a single person can build a billion dollar company, which is apparently happening. um like what what does that leave us? And we have healthy","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=490s"},{"start_seconds":499,"text":"debates about this because I'm not convinced that we'll ever reach a point where we'll have figured everything out.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=499s"},{"start_seconds":505,"text":"Um if you are saying that Singularity as it stands exists in a mode where it has figured out every single thing out for us and we don't need to do anything","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=505s"},{"start_seconds":512,"text":"else, we're kind of in Star Trek world, right? But I very carefully didn't say that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=512s"},{"start_seconds":516,"text":"But that that but yeah. Okay, that's that's a fair point. Um so you don't believe in that or you think it's","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=516s"},{"start_seconds":522,"text":"I think we're so far away from this lines are different right everyone defines AGI different ways uh otherwise we some by some definitions we're there","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=522s"},{"start_seconds":531,"text":"we have been there since like the first agents came about and by other definitions we won't be there for another 20 30 years um so depending on how you define whether we're still","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=531s"},{"start_seconds":540,"text":"useful I don't code nearly as much as I used to directly but I'm writing a lot more coding projects am I still useful I think they're they're of value to me I","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=540s"},{"start_seconds":548,"text":"think the nature of how these things are shaping my life is changing. It's an evolving situation. I don't think we can just put a pin in it saying we're useless at some point.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=548s"},{"start_seconds":556,"text":"David, are you still useful? Sorry, that sounds really mean. Tell me.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=556s"},{"start_seconds":560,"text":"It depends who you ask, I think. But um you know, I I think you made some good points.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=560s"},{"start_seconds":566,"text":"I I people are used to concepts of of robots from movies and television and you know, we're maybe at step one compared to step 10,000 we need to get","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=566s"},{"start_seconds":575,"text":"to. So um I think it's a long time to imagine you know a world without us but","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=575s"},{"start_seconds":582,"text":"even then creativity invention um human ingenuity um you know how we dream things um there's elements that I don't","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=582s"},{"start_seconds":591,"text":"think will ever be created that replace us. Yeah, I I've got ethics and guardrails and factchecking down there with creativity, but I I sometimes","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=591s"},{"start_seconds":599,"text":"wonder whether there's this asymmetry, right? We we talked about how this technologies have always come along and you know, calculators","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=599s"},{"start_seconds":608,"text":"provided a a new way of doing things much faster, much more easily, but just with the speed and the uh the the scale","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=608s"},{"start_seconds":615,"text":"at which you can can do things. I do worry even about stuff like creativity.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=615s"},{"start_seconds":618,"text":"the uh I don't know if you've been following that people are doing maths proofs with AI but the the AI doesn't","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=618s"},{"start_seconds":627,"text":"get them right it often is giving good ideas towards and I wonder whether we will get to a stage where the you know","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=627s"},{"start_seconds":636,"text":"AI can actually solve some of these math problems but I'm not sure that I expect a world where it sees why it should make","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=636s"},{"start_seconds":645,"text":"some sense David yeah well this is kind is something you're dancing around. But um we're focused on robotics AI and as you know","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=645s"},{"start_seconds":654,"text":"robots are AI in the physical world just manifested and um we think that to get to AGI to get to where you know the","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=654s"},{"start_seconds":663,"text":"robot can function like us and navigate the world that that it has to be you have to put AI in a physical body","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=663s"},{"start_seconds":670,"text":"because without the physical body it's kind of in a dream state like when we're dreaming we don't have our bodies just a we're just a floating consciousness in","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=670s"},{"start_seconds":678,"text":"our dreams. you don't have all your capabilities. There's many things you can't do um and things you don't understand. So AI really needs to","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=678s"},{"start_seconds":685,"text":"understand the physical world to advance is is our belief. Um so getting it into robotics, training robots about the","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=685s"},{"start_seconds":692,"text":"physical world um should help advance that and that's what Tober life is about is advancing it till so robots can take over the world a little faster.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=692s"},{"start_seconds":701,"text":"So there's I think the argument is that without real world actual interaction physical interaction AI can't actually","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=701s"},{"start_seconds":711,"text":"be getting closer to us and as that happens there are some physical constraints that will slow things down. Am I kind of getting that right or not?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=711s"},{"start_seconds":719,"text":"Um yeah, I didn't follow the last part but but the physical constraints um will help build an understanding of the world","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=719s"},{"start_seconds":726,"text":"for AI so they can be more useful for us and um and and moves it towards the AGI and yes there'll be phys there'll be limitations in there to protect us.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=726s"},{"start_seconds":736,"text":"The posetronic brain is not yet upon us but azimoth fans uh anything yeah there's also different measures of","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=736s"},{"start_seconds":745,"text":"what intelligence even means. we haven't really figured this out for ourselves.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=745s"},{"start_seconds":749,"text":"Like there's consciousness questions, there's intelligence questions, there's all of these layered questions. So I really try to stay very close to like","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=749s"},{"start_seconds":756,"text":"how is this useful in my real life? And right now as an assistive tool uh in every layer, it's extremely useful to","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=756s"},{"start_seconds":763,"text":"utilize AI as an independent tool doing everything by itself. Maybe with agentic orchestration, you can say it's able to do some of those things, but there's","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=763s"},{"start_seconds":771,"text":"still quite a few steps left. Um, and from my my point of view, first they completely forget the idea that the the way that neural networks work, even though we name them that way and they're","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=771s"},{"start_seconds":780,"text":"modeled after the brain, it's not exactly the way our intelligence works, right? So, it's a different type of thing and the amount of data it has at at boot like the scale is enormous.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=780s"},{"start_seconds":787,"text":"There very different ways that these things work. Um, but even given that it's helping us in all these ways, I think there's we're a few few days,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=787s"},{"start_seconds":796,"text":"months, weeks, years away. I don't like the days but I you know I I think that's a a fair point and you know the fact that we don't know how our own","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=796s"},{"start_seconds":804,"text":"consciousness works means it's rather difficult to say whether AI is or will be conscious even right it's a completely different framework altogether when you think of like","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=804s"},{"start_seconds":813,"text":"classical deep learning like essentially you're treating teaching patterns and the ability to reprint on those patterns over the course like understand like","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=813s"},{"start_seconds":821,"text":"fairly deep uh ideas is basically we'll give you enough layers we'll figure out there's a depth to this there's different types of patterns that can be learned And maybe that means it's","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=821s"},{"start_seconds":829,"text":"learned. And because of LMS and the latest advances in machine learning, we're saying, \"Oh, we're getting pretty close to this pattern matching. That must be intelligence.\" It might not be,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=829s"},{"start_seconds":838,"text":"but it's very useful. And so we should use that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=838s"},{"start_seconds":841,"text":"Yeah. Unless, of course, our consciousness is just pattern matching.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=841s"},{"start_seconds":844,"text":"Yeah, that's true, too, David. Sorry.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=844s"},{"start_seconds":846,"text":"So, um, just touching again on what you're saying. So, voice recognition has been around for 50, 60 years. I mean when I was a little boy there was Eliza","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=846s"},{"start_seconds":854,"text":"and there was voice recognition as part of that but it was very elementary compared to what we can do today of course and I I I see where we are today","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=854s"},{"start_seconds":863,"text":"maybe like voice recognition was 50 years ago and if you project that out like these AI factories that they're building for $50 billion with nuclear","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=863s"},{"start_seconds":872,"text":"power plants next to them and the level of computing power that exists there I think eventually we'll be in a handheld","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=872s"},{"start_seconds":878,"text":"device or in a robot and when that level you know, it might take decades, but when that does occur, I mean, I think","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=878s"},{"start_seconds":886,"text":"essentially they'll reach us when you have that level of power locally um to exploit, but I think, you know, right","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=886s"},{"start_seconds":894,"text":"now we're simulating something that eventually will be the reality of it, just like voice recognition from 50 years ago.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=894s"},{"start_seconds":900,"text":"So, you said the word exploit, and which kind of brings me to my next question is should should humans be scared and if so of what?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=900s"},{"start_seconds":911,"text":"I think of a Mark Twain quote um that comes to mind where he said um he said if you take a starving um sick dog and","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=911s"},{"start_seconds":920,"text":"you feed it and you make it well it'll never bite you. And he said that is the difference between dogs and man. And uh","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=920s"},{"start_seconds":929,"text":"I reflect on that that dogs are created by God and you know AI and robots are created by man. So I just think we have to be careful but um you know I I","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=929s"},{"start_seconds":938,"text":"believe in a bright future with them. So defender AI is basically a dog.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=938s"},{"start_seconds":943,"text":"No, I don't you know you could you could guard rails. Um yeah, you could you can make the argument that we can certainly train certain behaviors into AI and and","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=943s"},{"start_seconds":952,"text":"there's a lot of value in that. Sure. Uh I don't see it as a a dog per se. I think the despair thing is so weird to","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=952s"},{"start_seconds":959,"text":"me because think about all the things you can do now. Like you can take your government website that has this really terrible civics like setup and like you","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=959s"},{"start_seconds":967,"text":"can't find the right form or the right thing that you need to do and you can feed it to uh openclaw or some agent and some orchestration unit and you can get through a problem that used to take a","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=967s"},{"start_seconds":976,"text":"lot of time and was completely unreachable for vast majority of people who are like who can't afford these services and now they can figure out how to like deal in their civics","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=976s"},{"start_seconds":984,"text":"environment. you have transfer learning happening from all these people who are learning to curate all of this knowledge uh utilizing AI because they're forced to use these systems so much like we're","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=984s"},{"start_seconds":992,"text":"kind of forced to use the internet today and the cloud um and that transfer learning is is essentially teaching you how to curate data that's coming from your social media and and so you better","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=992s"},{"start_seconds":1001,"text":"understand how to be a curator of knowledge and so all of these wins that we're getting have very clear value so the idea that should be we were of","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1001s"},{"start_seconds":1009,"text":"course every great thing should come with like a healthy amount of skepticism and this idea that you should be careful of all the terrible things that can happen, but there's so much opportunity to do good.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1009s"},{"start_seconds":1019,"text":"So, I I I agree with that and I'm also aware that we're speaking to a captive audience of the right people, right? Of all the people who are like to be able","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1019s"},{"start_seconds":1026,"text":"to survive and have jobs of some type and living a lot of them are in this room, right? And I I think back to, you","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1026s"},{"start_seconds":1035,"text":"know, example of when when sewers started being invented, right? Yeah. uh they were obviously a a a common good","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1035s"},{"start_seconds":1043,"text":"for most people, right? But the people who used to take away the stuff from the streets, they lost their jobs and they","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1043s"},{"start_seconds":1051,"text":"couldn't necessarily retrain and that is going to happen. So I think at the individual level there will be and we're seeing you know entry-level jobs are","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1051s"},{"start_seconds":1061,"text":"disappearing in some sections and I think they're going to come back but Oh okay that's what say? Yes, you agree right? Exactly right.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1061s"},{"start_seconds":1070,"text":"Things will impact us in a poor way, which is why I'm saying it's not all all roses, but we have the tools today to get back up from that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1070s"},{"start_seconds":1079,"text":"Who who is we? Who is we?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1079s"},{"start_seconds":1080,"text":"All of humanity have tools uh to to get back up from this idea of like, hey, I might not have a job, but you can learn new skills faster than you could have ever learned it at a deeper","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1080s"},{"start_seconds":1089,"text":"level than you could have ever learned it. And you can start implementing it today. Like that has never happened either. So there's this balance that","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1089s"},{"start_seconds":1097,"text":"happening we as humanity and the civics and the governments and the organizations actually care about their people um","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1097s"},{"start_seconds":1104,"text":"which I hope most do um there's an opportunity to really just push forward through the despair part which inevitably happens with all of these","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1104s"},{"start_seconds":1113,"text":"changes and I think one of the other changes to come to a point I think is relevant to you David is that quite a lot of the previous changes have been technological","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1113s"},{"start_seconds":1121,"text":"orformational changes you know, think of the internet or printing. Um, but this one has relevance","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1121s"},{"start_seconds":1130,"text":"and uh instantiation in the physical world through robots, through uh other","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1130s"},{"start_seconds":1138,"text":"types of uh automation and AI in ways that maybe haven't before. Maybe that's a thing to be positive about.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1138s"},{"start_seconds":1145,"text":"Yeah. And I I I just would build on what you said about um the opportunity. So, you know, if you don't have your job","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1145s"},{"start_seconds":1152,"text":"moving boxes or pushing a paper anymore, cleaning up something on the street, the power that a robot with AI gives you,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1152s"},{"start_seconds":1160,"text":"um, I mean, all you have to do is imagine an application or something that could be valuable to somebody, and you have a machine there that can do things that have never been able to be done","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1160s"},{"start_seconds":1168,"text":"before. So, I think, um, um, it empowers people to to create great businesses on their own with these tools and make a lot more money than, but not everyone is a business owner.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1168s"},{"start_seconds":1178,"text":"Not everyone wants to do that or can do that. And we there will be there will be I don't want to use the word suffering.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1178s"},{"start_seconds":1184,"text":"There will be people who do not do well out of it. But I think it's going to float more boats than it that it sinks.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1184s"},{"start_seconds":1189,"text":"And that's that's the hope. And we need I think we have a responsibility as the people in this room to think about the people who it's going to be most","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1189s"},{"start_seconds":1196,"text":"difficult for uh to do that. So I we haven't got a huge amount of time left.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1196s"},{"start_seconds":1201,"text":"I had kind of one final thing and I I'm not sure it's even the right question.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1201s"},{"start_seconds":1206,"text":"Okay. but is what does AI do badly now that it's going to do better soon that is relevant to this particular problem?","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1206s"},{"start_seconds":1217,"text":"Do you want to take that first? I got to think on that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1217s"},{"start_seconds":1220,"text":"Okay. Um I have a few things and it's not badly. It's say it again. I have a few things.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1220s"},{"start_seconds":1225,"text":"It's not badly but it's things that we're constantly improving on. Obviously we still need a lot of data to be able to learn a new mo like a new mode of","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1225s"},{"start_seconds":1233,"text":"operation. That's a technology that's being constantly improved. Uh you see it at every level in the LLM's world in the U.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1233s"},{"start_seconds":1239,"text":"And I would I would add energy usage to that as well.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1239s"},{"start_seconds":1242,"text":"Energy usage is a thing that we're terrible at right now. And I think um not only can we utilize not only can like we work with machines to like make","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1242s"},{"start_seconds":1250,"text":"them better faster. We can utilize AI to make them better faster even faster. Right. There's a song about that.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1250s"},{"start_seconds":1256,"text":"Yeah. Sure. Uh there's so many things that it does well too that we misunderstand and there's a lot of lack of knowledge. Sometimes people use these","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1256s"},{"start_seconds":1263,"text":"AI systems. A lot of clients I help, they're trying to use it like a deterministic system. Forget the fact that not everything is deterministic that we think it is, but just the idea that they're trying to use the tool in","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1263s"},{"start_seconds":1272,"text":"the wrong way. If we start teaching people how to use the tool the right way as we're improving the tool over time, uh there's a lot like there's a lot to be gained still.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1272s"},{"start_seconds":1281,"text":"Okay, I like that, David. Well, yeah, just as you said, the hammer was probably used to build this great hall we're in, but in the wrong hands, you","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1281s"},{"start_seconds":1288,"text":"know, you have broken bones and broken glass if it's in used correctly. So, um I think the training of the tools is very important. Also, you know, areas","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1288s"},{"start_seconds":1297,"text":"that um in terms of robotics for AI, navigating the world in the physical world is still a big challenge for the","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1297s"},{"start_seconds":1304,"text":"robots. Okay. So um um once that is overcome and there's a lot of work done on it and and training data is key for","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1304s"},{"start_seconds":1312,"text":"the AI that people are familiar with there's been all the writings of humanity up till now for it to absorb.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1312s"},{"start_seconds":1318,"text":"So there's a lot of training data but for the physical world there's very little training data and it's being created but um that's going to take a","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1318s"},{"start_seconds":1325,"text":"lot longer for the robots to learn how to navigate and deal with our world even just to come and pick up a bottle. So once that happens, we're","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1325s"},{"start_seconds":1333,"text":"Yeah. And I I I go further than that because a physical world isn't just these things, right? It's how we interact. It's how organizations work,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1333s"},{"start_seconds":1340,"text":"how communities work, and navigating what that looks like in AI is is a next level problem and something hopefully we","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1340s"},{"start_seconds":1347,"text":"can continue to help with. Uh and luckily there's a lot of science fiction in this area. Not all of it positive, but you know, a fair amount of it. Um okay, so we have a very short time left.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1347s"},{"start_seconds":1358,"text":"Um, Dinda, what would you send people out with some hope? Give them give the give our audience some hope around AI in the real world.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1358s"},{"start_seconds":1367,"text":"I think the opportunity to improve your life and do less of the boring things is is immense. I think the technology is","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1367s"},{"start_seconds":1375,"text":"constantly improving, but if you were to stop today, you could still get a lot more value out of it. I think every single person here who wants to can build a business. I think every single","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1375s"},{"start_seconds":1383,"text":"person who wants to can learn a new skill faster than ever before possible.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1383s"},{"start_seconds":1387,"text":"And for sure, you can also make an impact in new ways that you might not have had the tools or just the resources to be able to do. So, you're empowered","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1387s"},{"start_seconds":1395,"text":"in ways that you've never been empowered before. Cool. David, some hope, please.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1395s"},{"start_seconds":1400,"text":"Um well, eventually um robots will learn how to build robots and and repair them and make them and then um they'll make,","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1400s"},{"start_seconds":1408,"text":"you know, thousands, tens of thousands, millions of robots. And the robots will eventually then maybe make a city of robots and build roads between their","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1408s"},{"start_seconds":1416,"text":"their cities. And just like we do, we don't really ask the animals for permission when we put a road in through a city. So, um you know, we we might","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1416s"},{"start_seconds":1425,"text":"look like that to them. But, um again, I think that's a long way off. And I really more echo the um the idea that, you know, we can have um zerocost goods","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1425s"},{"start_seconds":1434,"text":"like with robots making everything and doing everything. um we can build a utopia if it's engineered right, if the right thoughts are presented. So I think","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1434s"},{"start_seconds":1443,"text":"it's really up to us to formulate the vision for the future that we want and make it happen.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1443s"},{"start_seconds":1448,"text":"Thank you. So my and my last takeaway is that I'm hopeful because you lot are here right you're you some of you stay","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1448s"},{"start_seconds":1455,"text":"well we've had more people arrive this people care about this and we can engage on this and I'm hoping that in a year or","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1455s"},{"start_seconds":1464,"text":"two we'll have AIs in the audience or on the panel that we can engage with as well that's going to be really interesting when that happens right I","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1464s"},{"start_seconds":1472,"text":"know people are beginning to do these talk shows but when we when we get to that stage and we can actually start having these conversations and and","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1472s"},{"start_seconds":1481,"text":"seeing how we can interact. That's when I think it gets really interesting. And please go out, be hopeful, do the right things, and thank you very much indeed.","url":"https://www.youtube.com/watch?v=p8niHQvrVII&t=1481s"}],"youtube_url":"https://youtube.com/watch?v=p8niHQvrVII","url":"https://near-pedia.xyz/videos/p8niHQvrVII"}