AI Patents Itself: The Invention Machine Is Here

What happens when you feed an AI its own code and ask it to improve itself? It invents something better, writes a patent for it, and files it. In 8 minutes. Dr. Marcus Weller did exactly that with Deep Invent, and the result made him realize his life had just changed.

We're talking about an AI platform that analyzes patterns across all human innovation to create new inventions. Real ones. With actual patents. A 7-year-old used it to co-invent a system for removing microplastics from water. An Amazon robotics director quit his job after the system showed him an AR speed-reading invention he couldn't stop thinking about.

Marcus breaks down how Deep Invent works - scraping global scientific literature, patents, and market data in real-time to find the "white space" where innovation hasn't happened yet. Then it generates clusters of patentable inventions in those gaps.

We dig into the difference between AI hallucination and imagination, why cross-disciplinary insights matter, what "recursive evolutionary inference" means, and whether we're looking at steps toward superintelligence. Plus Marcus shares his grandmother's story about the resistance to electricity in homes - a reminder that every transformative technology faces the "this is too weird" phase.

Deep Invent: https://deepinvent.ai

Click Here to Subscribe:

FUTR.tv focuses on startups, innovation, culture and the business of emerging tech with weekly podcasts talking with Industry leaders and deep thinkers.

Occasionally we share links to products we use. As an Amazon Associate we earn from qualifying purchases on Amazon.

Transcript

0:00

I've got an absolutely amazing show that's gonna blow your mind today. What if AI could discover breakthrough inventions by analyzing patterns of innovation? Stay tuned. Today we're joined by Dr. Marcus Weller. Marcus has spent more than a decade advancing US interests as a science diplomat, author and speaker on AI and foreign policy. He's been an advisor to at ARC and DARPA and has a PhD in industrial psychology. This year, Marcus launched Deep Invent an AI platform that's designed to identify innovation patterns across disparate industries to identify market opportunities, produce actionable strategies. Even drafting patent applications. Deep Invent uses what Weller calls engineered serendipity to reveal hidden patterns of innovation and connections in scientific data to generate new IP concepts instantly, potentially reshaping how we discover tomorrow's technologies. So let's talk with Marcus about how it all works. Welcome, Marcus. Thank you for having me, Chris. I'm excited to have you. I'm, I'm super excited for this conversation.'cause we talked earlier and our call went very long 'cause we were just coming up with all sorts of new interesting areas where all this kind of technology could go. Um, so I'm sure we're gonna have an incredibly fascinating conversation, but I wanna start off with, you know, like how did your journey bring you to Deep Invent? Yeah, so it was my, uh, sophomore year of college and, um, I actually, it was, it was, this was 2004, 2003. And, uh, I learned, I was sitting in a cognitive psychology, uh, course and I learned about neural networks. They had this mathematical formula for basically how cognition happens in the human brain in, in our wet net. Yeah. Behind, uh, between our ears and behind our eyes. Right. And, uh, I thought. Wow, this is the most important thing I've ever learned about. Because nothing in the world exists unless it's happening through a mind. Unless it's perceived by a mind. Right? Yeah. And so, um, it just seems so fundamental and important. And, you know, at the time I was kind of considering maybe I wanna study cosmology, you know, um, and, and learn about the cosmos. But, you know, in that moment it really crystallized for me that. The brain, the human mind is the most complex structure that we know about anywhere in the universe. It's sort of the distillation of, uh, complexity that all the universe has, has sort of produced. Yeah. And, um, so I thought, you know, I think this is what I want to do. And I decided in that moment that I was gonna do a, a PhD in, in this related field and study the essence of cognitive ability and also what the result of, of cognitive ability was. And that ultimately led me to studying innovation. Right? Yeah. Like the, the, the most important out and, and sort of sophisticated and complex thinking that human cognition can produce is, is this, this ability to, to come up with novel concepts and, and innovate things that should exist that, that don't yet exist. And, um, so that's sort of captured and captivated my, my attention. And, um, I ended up sort of dedicating my, my life to that. And eventually that led to, um, you know, uh, building a system that could. Emulate and augment the way that we as humans are able to produce this type of thinking. I gotta imagine a lot of that is kind of counterintuitive too. Um, because, you know, when we look back inward at how we think about things, um, it's not always clear to us, like how we get to from point A to point Z. Right. Innovative thinking is very hard to trace back sometimes. Yeah. You know, it seems sometimes like a lightning bolt just hits you out of the blue. You know, you, you get the idea in the shower or, um, you know, you're, you're, you're sleeping and you have the idea or you're waking up, or you know, meditating and it just seems like it just strikes you and it's this emergent, you know, concept, right? Yeah. But what's actually happening under the surface. Is there are, you know, there are neurons that they're interacting and they're spreading their activation and they're sharing information below your perceptual threshold. And then eventually, you know, an idea will crystallize and then it will surface itself up to your frontal lobe and your, your direct, you know, conscious awareness and you become aware of that idea, but that your subconscious is sort of always processing these sort of, you know, um, different probabilistic states and the information's flowing and it, it's always happening. And, uh. That's what's so fascinating about neural network, artificial neural networks and these new LLMs is you can actually use some of the human, the best of the human cognitive heuristics, embed those into systems of interacting AI models, and then you start to get this sort of emergent behavior where you can actually get innovative concepts that that sort of are derived or synthesized from undiscovered. Public knowledge. So there's so much information and data out there everywhere. Yeah. And, um, it's, it's, there's only a fraction of the human knowledge that's going to be generated, that's been generated yet, you know, it's some tiny, tiny fraction. Right? And so there's so much, you know, undiscovered, untapped knowledge out there that comes from synthesizing information from disparate, you know, regions in a knowledge graph or disparate fields of study that maybe they use different language, but they're referring to the same sort of phenomena, you know, under the surface latent in that data. And, uh, that's very hard to do for a human brain or a given human to do. It's sort of what we say, b beyond the cognitive horizon of, of humans. But with AI models you can do really, really interesting things with data science to recognize. Latent patterns or structures in the data, and then look for similar structures from different fields that maybe they use different language, but because they have common structures that you can recognize deeper in that latent data structure, you can actually map those onto it and give these models the way they interact with each other, the ability to do. Reason by analogy and, and find analogies in other concepts that are pre linguistic that don't even use the same language to refer to the concept. And when you do that, that's how you get these really transformative, um, interdisciplinary innovations that link one thing to another. And that's how you sort of mine, uh, knowledge from undiscovered sort of public, public information. The last three or so years have been really, really interesting because I have never seen technology change as fast as it has recently. And it's all due to this sort of AI revolution that we've had largely around the LLMs. And I think that, you know, the level of innovation I've seen, you know, for, there's a lot of things you can say about LLMs. And they, they have their challenges here and there, but the reality is like there's a lot of innovation happening, and it's not just like, it, it's the pace of innovation that, that has increased so drastically. I mean, what, what do you, you know, like we have sort of like this, this human concept of intelligence and innovation and then, you know, like, and, and we're kind of mimicking some of those functions. In ai, but, but it is kind of fundamentally different in a lot of ways. I mean, like, could you talk to how, like the difference between sort of human intelligence and like where machine intelligence is going there? There's the direction of machine intelligence that if sort of left unchecked it becomes very exotic, hard to explain, not necessarily transparent, um, how those conclusions are arrived at and um. I think that there's a, there's a friction there because it causes humans who need to make use of these innovations question in, in some cases their validity when they're not as explainable, right? Um, and so there's another way to do it, which is to try to use some of these artificial systems to augment the human's ability to innovate. And so sort of keeping that human in the loop, and not replacing them, but sort of capitalizing on the two types of neural networks, the wet nets of the humans and the, and the dry nets, you know, the, the, the artificial neural networks, um, of these, of these LLMs, um, and even other, um, of artificial intelligence paradigms. And, um, I think that. You know, the way we look at it at Deep Invent is that we're trying to, um, bring the best of both worlds together and then allow the human to sort of be the, the progenitor of the initial concept. And then to allow these systems to do things that are beyond the human cognitive horizon, you know, that are beyond those limits that we have as a human. So even as a PhD and sometimes especially as a PhD, you only know what you know, right? You get really, really narrow on a specific field of study. And in order to have that expertise, but to stay relevant in that field, you, you, you need to keep up with the volume of literature that's exponentially increasing in, in, in speed and volume. And so you have to get progressively more and more narrow as your career progresses. And what that means is that you actually. It may become less and less likely that you have these transformative, you know, leaps of innovation be, and, and you tend to, you know, innovate in a more incremental way because you're so narrow. And the, these leaps tend to come from cross-disciplinary insights, right? And so if you're not able to be familiar with, at the level of depth that it would be, be required to sort of operate at the frontier in multiple fields at once, that's a perfect area where artificial systems can come into play and sort of be your innovation copilot, because they're inherently polymathic. They are deep in a broad range of fields all the time. And always, and the way that we sort of run them is we augment that even further by scraping all of the world's scientific literature about a given human's idea all the way up to the minute. Into a knowledge graph, right? Yeah. That, that scientific literature is, it's, it's this ephemeral knowledge graph that happens right in this moment. It's the first time in history that that knowledge graph has been generated for that concept pulls in that scientific literature. And then not just the, the scientific data, it's gonna pull in the technical data. So the prior art, the patents, the existing patents up to the minute, and then the commercial data. So you've got the scientific, technical, and commercial data all into one knowledge graph. And imagine now your idea is contextualized so much more deeply and profoundly than it was before you really understand the landscape and the context in which that idea lives. And from that starting point, you can then generate. Innovations in the under innovated regions of that knowledge graph. Well, that's what we call the white space. Yeah. Where you're able to actually synthesize new inventions from, from, um, cross-disciplinary data. So you can imagine, you know, a human PhD, if they're in there using this, uh, they're able to then have a team of PhDs that are multidisciplinary where the system has an intuition about what other fields of studies would be relevant for this to really, you know, take the idea the next level. And then not only that, but you get through that process. And if that research wasn't valuable enough, you can actually crystallize this into patents, drafts, you can actually generate. Patents that you can then protect so that every time that you run a session in, in the system, in Deep Invent, you're able to generate an economically valuable output that you can commercialize. You know, when you first go to Deep Invent, you're, you're kind of presented with like that standard chat window. Interface is sort of your, your starting point. The flow is really, really simple. We call it cursor for deep tech. So it's kind of like your vibe inventing, you know, you're gonna put in a couple of sentences about whatever the concept is that you wanna investigate. It can be a problem that you're trying to solve. It can be a hypothesis that you have. It can be an industry that you're curious about, or it can be a specific product that you have in mind or, or a service or a framework. Um, you can put that in, in a sentence or two. That's all it takes. If you have more, you can add more. But it really, all it takes is a couple of sentences. You put those into the system, you just, you just type 'em in. And then what it'll do is immediately it starts this agentic flow and carries you through every step of this reasoning process that is, um, specifically an interaction. Of debating AI models, they, they, it's like a team or a fleet of models that are sort of conspiring in your favor, right? To figure out how to innovate around this concept and take your idea to the next level. So what it'll do is it'll first characterize it and it'll tell you, okay, this is the idea that you have and these are all the things that you know are related to this idea and that this is sort of the broader context, and this is the market around this. These are the people that are doing things like this. So it really helps you understand at a deeper level, immediately the quality of your idea. So you sort of get to in initially just test the quality idea right out of the gate. Then what it does is it starts building on the idea immediately. So it. Starts to build your own custom knowledge graph for the, for that idea immediately. And what that'll do is it pulls in, first it's, it's like a team of PhD scientists, and it'll look at all of the research that's happening all the way up to the minute to right now on that specific subject in any language all over the world. Um, and so then it'll pull that data into this knowledge graph. And then what it'll do is it'll, it'll look for all of the patents anywhere in the world that have been filed that are related to this, and that's called prior art. That, that, that technical data is sort of, it helps you understand. You know how these, um, this idea and this science has been sort of crystallized into something productizable. It's been, you know, converted into something that, that you can actually make, because an embodiment for a patent is, you're supposed to describe it and disclose it in a way that's sufficient for someone to actually, you know, build it, um, or create it, right? And then what it'll do is with that, now you've got sort of, you understand the, the sort of products or services, um, or software or whatever that are that, that are, that exist in the world is it will then under help you understand what is the commercial landscape in which those products live. Where is that, what industry do they live in, and where is that industry going, right? So it's not just telling you about what the commercial landscape is today, but it's understanding the trend line of where that industry is heading. And then with that contextualization, it will help you then generate. New provisional patents, specifically addressing those gaps in that future competitive landscape. And it will generate not just one patent, but a cluster of patents to address those gaps. And what that helps you do is really understand, not just, you know, what you can file and what you can protect, but these patents serve as sort of a disclosure or a product description, or a product definition we say in, in the technical landscape that tells you what to do your r and d on. So it actually helps you understand, where am I gonna invest my time or capital or energy or research to actually start to create products that the world in fact needs. When I tried it out, um, I was, I was, uh, impressed by the level of a market analysis that it did as well. I mean, 'cause like I, I, I chose a very niche kind of subject and it, um. It, it really like nailed it. I was, I was kind of surprised. I was expecting it to not, not like, hit it as, as well as it did and, and I know some of the players in that space and I kind of ran these ideas, you know, by them and they're like, wow. Like that really, you know, sums it up. And, and it even cited, you know, this company is one of the, the innovators in this space and he is like, one, it's great that it's validating the direction we're going. And, and two, it's, it's like there's some, there's some interesting ideas in there actually, you know, so, so I thought that was really, uh, the depth of it was surprising at the stage of development you guys are at right now. I noticed you're not telling us specifically what the product is, so you must really wanna protect it. No, I had some ideas. Yeah. No, I got some ideas. I got some ideas. No, no, but I mean, it, it really, uh, I thought it was, it was fairly impressive. Um, you know, where, I mean, I just feel like this is just making you a big brain, you know? Um, but, but the thing that's interesting about it is it's, um. The way it's iterating through ideas and iterating through the landscape is, is like you say, you're, you're bringing in all these PhDs to kind of instantiate the idea. And then you've got like these marketing PhDs almost that are going out and doing the market analysis. Um, I mean, it, it, it, and, and it's, and, and it's, it's interestingly providing really kind of non-obvious, some obvious things that you would anticipate, you know, like you can, you can kind of see where the market's going, but there was, there's definitely some non-obvious things emerging from there. And I, I know when we talked before, you know, and you, you mentioned earlier about the idea of, of taking, of, of analyzing design patterns, you know, like, or innovation patterns, and taking innovation patterns from like one industry and bringing them to another industry. Can you talk a little bit about how that all plays out in there? Yeah, this is a really important piece of it, is that we wanted to make an architecture that was going to, um, significantly help the user see far beyond where they would've been able to see as a human innovator. Right. Um, and the way to do that, uh, is to, what we looked at was, you know. Cognitive heuristics of great human innovators and the approaches that they would take, uh, to generate new, meaningful or profound innovations for humanity. And the way that that's often done is through analog reasoning or finding analogies from one field to another field. And that for a human often feels like this really intuitive, hard to trace, uh, thought pattern or concept. They just suddenly recognize something is similar from another field. And there was, there's a sort of a gestalt hypothesis, hypothesis around this, which is like recognizing the structure, um, or arrangement. Of, um, data, uh, so like structural mapping of, uh, a structure in the data to another data structure. And that's like what kind of is happening like deep in the, in the humans neural network. And this is a, this is a pattern that we wanted to try to follow and implement in deep Invent. And so this is what we're doing, um, to, to help the system augment the human's ability to recognize analogous latent patterns in data from one field to another field. So what it will do is it will intuitively sort of hook and find those and pull them in because it's inherently polymathic and it is aware of all of those data structures all at once. And that's that component that's specifically beyond the human cognitive horizon. Even if you had a thousand humans. That had diverse, you know, fields of expertise, you'd still have to coordinate the knowledge transfer between those human nodes. So this is just a seriously difficult and intractable problem, and that's why it's so rare for humans to have these transformative innovations where there's a sudden insight from a completely different field of study and a link between the two. But we can drive that now with these new modern tools and these artificial systems, we can start with the human's idea and intuition and then put that on steroids and really drive that linking and force the, an analog reasoning across fields of study every single run in every single session. And we can do that in big ways, and we can do that in small ways and we can sort of turn the dial on that, um, to optimize it for, you know, something that's productizable and buildable right now that the world needs immediately. And that is, is is technically achievable, you know, with, with, with modern technology. And so that's kind of the, the stage we're at now. And, and what we're seeing is, you know, users, uh, you know, putting their ideas in the system. And coming up with things that, uh, are definitely outside of, of what they thought was possible with that concept, but still achievable. And they get, they get super excited. And we've had a couple of instances now where people have like been using this on the side and they're working at, you know, Amazon for example, is one of our users. And, um, he was a research, um, uh, an r and d director, um, working in robotics at Amazon. Yeah. And then he came up with something in our system, this augmented reality, uh, augmented reality speed reading app. And he got so excited about, its the way that it embodied or, or, um, structured and, and framed the invention. That he quit his job at Amazon to go and do it full time. Um, and so, you know, and, and that's not actually the first time that that's happened, even when we're only in beta right now. Um, yeah. And so I do think that there's a lot of downstream, you know, impact that we're yet to see about how people can unlock their ability to innovate and really like, get to see an idea sort of manifest itself into the world in a way that is so much more, um, contextualized and well, well conceived and, um, is, is ultimately an idea that they can realize that they can, they can put into the world and, and have a, have a positive impact. And I think that's just what's so inspirational for me to watch, you know, is we created a platform for innovation that that can accelerate innovation. And there's just. Innovation is unbounded. There's, there's no such thing as too much human innovation. Right, right, right. And now we're seeing these just our users, you know, imagining and, and, and, and generating inventions that we could never have possibly imagined. And that, that's what I'm so excited about as we bring this to market. We talk about taking a design pattern from one, one market and bringing it to a different market. Um, but, but there's also that, that, that sense of intersectionality too, you know, like, 'cause the thing that's really. Uh, interesting about this is like, it's deep, but it's also wide, you know, so like there's that, that, like, I have this idea for this market, but it's like, well that idea applies across all these markets in this sort of way. And it's not just a design pattern, but it's sort of like it's a usage that, you know, like you wouldn't necessarily, that's a, it's sort of a non-obvious use use case that, that it, it, it designs for. And I think that's, that's really, um. Really transformative. Yeah. I mean, there's, there's so many ways to innovate, right? Like there's the technical innovation itself, you know, the engineering of it or how the, uh, invention is built. There's also, you can innovate how the, uh, how the, how the invention is manufactured, right? You could like the manufacturing process for a semiconductor, right? There's a, there's the function of the chip. You can innovate. Um, there's the, the, there's the structure. Of the chip that you can innovate. There's the manufacturing process of the chip that you can innovate and protect and patents. Right, right. That's very common in semiconductors. A lot of people don't realize a lot of the IP and semiconductors is around their manufacturer. Not necessarily, not just the, the function of the chip. For sure. And um, and then there's of course business model innovations. There's how the pro, how, how the product is monetized or how the concept is monetized. Those are also super exciting and I think there's a huge a, a amount of untapped potential in business model innovation.'cause if you think about it, there are only a handful of ex extent business models that, that we use. And business model innovation is quite rare. Um, and it doesn't have to be, I think that there's, there's a lot of potential there too. And that's why we wanted to make sure that the system would also have under its purview, innovating on the commercial data. That's why we have all three components or those dimensions of data in the knowledge graph, the, the scientific, the technical, and the commercial you need to in, in order for it to be innovation and from our perspective is you really need to have all three. Yeah.'cause that allows you to understand what's, what's new or novel. And it also helps you understand what's non-obvious and what's useful, you know? Yeah. And when you, when you've got those three dimensions and you can go from an ideation then to something that you can implement or implementation, and that's, that's, you know, what the research shows in innovation is necessary for that full spectrum of impact is you need to go from, be able to go from ideation all the way through implementation. Yeah. I mean, it's like you get a year's worth of, of thinking in like 20 seconds, which is wild. Um, you know, uh, one of the thing, there's, you have that concept of recursive, evolutionary inferencing. Could you, could you, uh, like explain how, how, like, how this. You know, you, you're, you're putting data into this thing and it's kind of churning against itself and, and, and creating. Can you explain how that all all works? There's a funny story about that actually. Uh, so we had the system, uh, last year, late last year working, um, sort of end to end. And it was generating new, uh, intellectual property, new inventions, and it was starting to saturate our eval benchmarks, our evaluation, uh, benchmarks, meaning that. Um, AI generated inventions were starting to be rated even higher by subject matter experts than the human subject matter experts inventions. Wow. Yeah. And, um, we were having them dual evaluated, meaning we were having a subject matter expert in the field and a patent attorney that is used to patenting in that field, uh, evaluate the inventions. Right. And so when that started happening, we started getting, you know, understandably pretty excited that hey, this, this might be the first system that it, that AI system that's truly innovative and is creating, is extrapolating, is generalizing, and not just interpolating, you know, in its little loop. And, um, and so what we decided to do was, before we were gonna go out and raise capital or, you know, commercialize this or put it in, in, you know, uh, people's hands. We wanted to see if it could innovate on something that we were the subject matter experts in and that it, we wanted to see if it could do it better than we could. Right. Um, 'cause it's one thing if it, like I, you know, exceeds a different kinda the eat your own dog food model. Mm-hmm. So we, so we decided to feed it its own product definition, its own architectural description of how it worked. We said, we fed it its own system description and we said, how would you innovate on this, on your, this is you, how would you innovate on yourself? That is real bootstrapping right there. It sure is. Um, it's a, or a snake eating its own tail. So, totally. So we, so basically what happened is it came up with something called recursive evolutionary inference. Now the interesting thing about that innovation is that in my handwritten engineering notebook at the time, and this was in December of last year. Uh, in my handwritten engineering notebook, and nowhere else was this invention I'd been working on called recursive evolutionary auto regression. And what that was, was a mechanism to use an evolutionary algorithm or a genetic algorithm to breed the top ideas that it would generate. And then with each recursive breeding or re with each recursion, it would inject genetic diversity. And that genetic diversity would be interdisciplinary data that it was able to recognize from one field to another. And that each spin, it would get more innovative. Right. Yeah. And I'm using these knowledge graphs and, and everything. Right. And so, um, I was pretty excited about it. I was pretty pleased, you know, uh, and then, you know, and that was, it took me about eight months of, you know, falling asleep, waking up, thinking about it, thinking about it in the shower, thinking about it while I'm eating, you know? Yeah. Just constantly day and night. And in about eight minutes. Uh, deep invent generated, recursive evolutionary inference, which was about 85% the same as what was in my notebook, except the other 15% was better. And I thought, whoa, man, this is, this is amazing. That's like a month per minute, right? Yeah, yeah. It's, it kind of feels like a time compression factor of a thousand. Um, and so, yeah, you know, basically then we just let it go through the rest of its process where it generated its own patent. Uh, on that innovation, which then I filed on its behalf as a good steward of, you know, the innovation. And to be fair, you know, I, you know, the, the, I did, you know, invent the concept. So we still have inventor of it, but the fact that it could catch, well, you, you invented the thing that invented it. So like, at the end of the day, even though you might not hold the patent directly, you still invented the thing that Yeah, you own the thing that owns the patent. Yeah. So it was, it, it was a fascinating thing. I mean, it was, this happened in a few minutes and it felt like we just got launched into a different. Dimension. Uh, we knew that our lives had changed in that moment. Um, I was sitting there, um, with my CTO and we took a screenshot of the screen of what was happening.'cause we could just, like, it just felt like one of those moments where we'll never forget it. Where, and, and, and as far as I know as, and, and you know, and I've looked around and talked to a lot of people, including the, the head of Microsoft Copilot and other people. This is the first AI that innovated on its own architecture and then in that flow, you know, patented itself or drafted its own patent on the resulting innovation of its own architecture. So, um, and we didn't, it's not necessarily that we set out to for that milestone, but it just. You know, it, it, it came to pass. And so that's when we knew, okay, we do have something here. It can deterministically invent the thing that should exist that doesn't yet. And this was just, you know, it was just undeniable proof directly in front of us. Empirically we could observe that that's what had happened. And that's when we decided to pursue it in earnest and build it out in a way, build guardrails in place, build infrastructure around it, get compute, get it ready for, you know, the servers ready to do the compute and set it up so that we could onboard beta customers and they could start to experience for the, for the, for themselves firsthand. Well, and and on that point, I think, uh, you know, one, one interesting thing here, and I'm just gonna pull this up real quick. Um, you just recently had the deep Invent for good innovation in, in Inventa on, I'm sorry, deep, deep Invent for good Inventa on, uh, and uh, there was a $10,000 grand prize, uh, and it was the. Integrated roadway, animal presence and deterrence system that actually won the $10,000 prize. And, uh, I'm just reading from your, your, uh, press release here. Uh, they got, they got the$10,000 prize and the a lifetime subscription deep event. Um, and that went to Laura Brown in the United States for her integrated roadway, animal presence and deterrent system with V two X communication for autonomous vehicles and smart traffic management, brown system integrates vehicle to everything V two X communication to detect and deter animals on roadways, reducing collisions while enhancing the safety of both drivers and wildlife. And then second place was, um. Alessandro DeAngelo of Germany who had realtime quantum inspired emotional state encoding and adaptive soundy synthesis system. That's a mouthful. Uh, or Kuma. And third place went to Brian, uh, Liat of the United States for his hybrid dual functional magnetic electro filter cartridge with smart se sensing, adaptive regeneration and integrated high efficiency. PFAS, microplastic waste Management. Um, I mean, those, those are, uh, that is not. Any small thinking going on there. These are, you know, as you might imagine, difficult for even me to, uh, you know, understand until, you know, you get in there and you're able to read it and you're able to see the sort of innovation flow. Like anybody that, that, that, um, goes to our, um, goes to the deep Invent for Good, uh, website, they can actually see these and go into our system and see their whole flow of how they invented it, how this system collaboratively invented these things with them. You get to see every scientific research, journal article, every, uh, prior art and patent that's cited. You can click into them. You can see how it was all comprised. You know, and that's what I find so fascinating is it's inventing transparently. It's like you can climb inside the mind that created it and see every chain. Of thought that it went through to arrive at that conclusion so that you can see it's an undeniable, valid conclusion. And I think that's what's so fascinating about this is that it completely democratizes this capability to invent profoundly to move us forward as a humanity. Yeah. The reason we did this whole campaign in the first place was that our, our sort of mission right, was when we release this to the world, we want to accelerate innovation for the good of humanity. You know, and we want to have massive economic impact as a result of doing that, you know, and across the globe. And we want to democratize this capability. Yeah.'cause that's what we need. That's the fundamental atomic matter of global GDP is innovation. And innovation is related to, yeah. All of the major most important quality of life outcomes for humans. Right. So, you know, that was sort of like our 10 year plan. Well. In the age of ai, that's not fast enough. And so we decided, we were, we decided with about a week's notice, my team was, um, let's just say a little frustrated, um, that we decided to accelerate this dramatically. But I said, let's do it in a week. Let's launch the ability to use this in a week. Well, you know, 10 years down to a week that, that, that tracks with your, you know, uh, month per minute, you know, kind of trajectory. Yeah. And so, you know, even our own team thought, you know, that Marcus, hey, appreciate the enthusiasm and the vision and. It's a crazy talk, you know? Um, but you know, we, we, to their credit, you know, they, they believed eventually that it was, that, that maybe we could give this a shot and, and, and, and at least try. It doesn't have to be perfect, but let's open it up. Let's make it totally free. Let's burden all the compute, we'll handle and pay for all the compute. We open it up to everyone in the world and we, what, what would happen if we just let everyone innovate all at once? And we said, and we provided an incentive to solve the most important problems that we have for humanity. And we give them incentive. If you do this and you generate it on the platform and you're willing to. You know, to, to contribute it to the public domain. So you open source, the invention that you generate, that the best idea to benefit humanity will win $10,000. And I was hoping, you know, that we would, we would open it up and maybe we get, you know,'cause we, we, we, we didn't plan or market it whatsoever. We just, right, we just opened it, you know, just to see if anyone would do it. And, um, we thought we, we'd maybe get like, you know, 10, 20, maybe 30 inventions and that would be awesome. And we got 10 x that. Um, and so many people flooding in, in the system and getting to experience it the first time. But, but 300 actual formal inventions that were, that were, uh, generated as part of the competition. And each one of those, it generates a cluster of inventions. So even within those, there's 10 within each account that was generated for each run. So the number of actual inventions for, you know, that, that, that were generated in the system is more like 3000. Um, and so that was just astonishing to see. And, um, then we had, uh, human expert judges like, uh, the head of Microsoft Copilot, um, the author of Blitzscaling, Chris Ye, um, people from, you know, uh, open AI and a bunch of amazing judges, right, that were, that are changing the world. They're people that are, you know, inventing at an absolutely prolific level, you know, um, or, or, or, uh, scaling inventions and innovation, uh, to humanity in a, in a, you know, in a prodigious way. Those were the people that were able to judge these. And the, the inventions that you mentioned are the ones that surfaced. Uh, and there's just, it was so hard to choose because they were all so, I'm sure you know, so brilliant. I mean, you know, and by the way, I'll point out one that I thought was just astonishing. So the one you mentioned, Brian Ette. Yeah. Which is like removing the forever chemicals and microplastics from the water supply. Right. He actually invented that his co-inventor was his 7-year-old child. So if you're wondering if it's, if you can use it or if anyone can use it, I think that the answer that says it all, the thing I wanna highlight here is how wildly different all of these ideas are. They're completely different markets, completely different industries, completely different technologies, but yet it was able to generate all these ideas across all that, um, that, that different landscape. You know, that's a great point that, um, that you know, I should have mentioned, which is that a lot of the best models right now. Their domain specific Superint intelligence for sure. You know, they're super intelligent Yeah. In a specific area or domain. And there's something qualitatively different about Deep Invent in that what we're trying to do is be super intelligent, meaning it's beyond the, the human capability, even with an infinite number of hours, um, or or a thousand person team, it's still beyond what that, that amalgamation of human thinking can do. And to do that across all fields, you know, to in any field. Yeah. Right. And, and that's where I think that, you know, I'm not saying that this is, you know, super intelligence, but what I will say is. If I were a consumer of super intelligence and I was, or I was waiting on super intelligence to arrive, I would certainly expect this as one of those steps on the road to be achieved to get there. Right? This is exactly what you would expect is something that can invent, actually come up with new things that don't exist, that should in all fields, and not have to rely on using a one model over here for, for, for pharmaceutical and this model over here for biotech and this model over here for industrial engineering and this, but to have one place right. To do everything that that's, yeah, that's what I find. It's inherently polymathic and it's, it's not just, you know, like a digital dilettante where it's at the superficial level. It's broad and deep everywhere. You know, we're, we're getting, um, the intuition of these machines that is kind of beyond, uh, human intelligence. And, and when we talked before, we, we kind of went down this rabbit hole about, um. The idea of like how we're seeing artificial intelligence have hallucinations, but what if hallucinations aren't necessarily wrong, but we just lack the foresight in our little human monkey brains to see like the potential of that. And, and, and like the way in, you know, like we try to suppress hallucinations in ai. But maybe, and, and, and part of what you do, I think too is, is not, is kind of encourage a little bit of that to, to get to that essence of innovation, right? Like if, if you think about when Einstein was doing the, the visual thought experiments, uh, about the train and understanding relativity and the clock moving. Yeah. Was that hallucination or was that imagination? It certainly wasn't something that existed in reality. Right, right. So I think what we actually need is a more granular way in the age of AI to look at hallucination, right? So if you are talking about data that has occurred, things that have, you know, deterministically occurred in the past, and a model tells you something that about the past that did not occur, that is a type of hallucination, right? Right. But if, let's say you could bound the models or the AI system in first principles in science, and you say, this is your fundamental basis. Now I want you to hallucinate into the future. Meaning that has, that does not exist. It has not existed yet. So you have to hallucinate something, right? But it's based on science that's more like imagination. That's a lot more like how humans innovate, isn't it? And so if you can, if you can, you know, base it on first principle science and then a capitalize on the stochasticity, capitalize on that sort of ability to hallucinate into the future, you can, you can, you can actually generate new things. And I think that was the fundamental insight for us is that, um, we just didn't look at hallucination with a fine enough lens, with enough resolution to be able to parse those and understand how to capitalize and where to constrain or tamp down innovation when looking at past data and when, when calling and scraping and digesting and ingesting past data. And then when to augment the hallucination and give it license to, uh, think analogically across fields and to think about things that don't yet exist and give it the permission to do that. We get into the idea of non-human, sort of emergent properties, right? I mean, they're, they're innovating in ways that we don't even recognize right now, which is really, I mean, it's both, it's both a little scary and wildly fascinating because, you know, that's how you, that's how you have giant leaps in innovation, I feel, where, where there's like, you know, it, it sometimes, and sometimes innovations like that, that it seems non-obvious until it's obvious. And then you then, once it's happened, you go, oh, how do we miss that? You know? And I, and I see this as kind of. Cracking that piece of it open, right? Yes. I think research transparency and sort of the, the, having an a transparent understanding of its thinking process is so critically important, and that's why we've set the architecture up the way that we have so that it's not hiding it's chain of, of thinking and, and reasoning. Um, it's, and, and it's, it's sort of chain of debate and innovation. All of that is wide open and, and, and visible. And you can zoom in on every thought it had and every piece of data that it used to trace back how that invention was generated. But there's something really fascinating about that when you're following through it. When it's using, we, instead of having a cold start, we gave it human cognitive heuristics of great human innovators. There are common right sort of thinking strategies that are layered that humans use to invent new concepts. And that's why some inventors are prolific inventors and they invent many things in many different fields because they have well trodden sort of, you know, neural pathways of these cognitive heuristic that they can use over and over again. Yeah. So we used those as our starting point and, and gave, gave those as the logic of how these different models in our system interact with one another to generate innovation. That's the first step. Yeah. So you could see the sort of the human fingerprint in those when you would be going through the innovation process. When you put your idea in and it's going through, it's, it's, it's sort of innovation process. But what we started noticing was sometimes it would make absolutely no sense from one step to another. And we're like, oh, oh, the system's broken. You know, when we were in early beta, we were like, oh, the system's misfiring. Let's just, you know, let's just go through the rest of it so we can do a full debug, you know, and see where it ends up. But sometimes it would, it would pull it all together in a later step. And we'd be like, oh my God. That's why it had that weird thing in there, you know? Yeah. Earlier in the process, like one of them was, um, one of the users, the, the user at Amazon that ended up quitting his job to pursue the invention. It generated, it was coming, it, it, it was this augmented reality glasses speed reading app that could allow you to read 400% faster by showing one word at a time, and then automatically adjusting the speed of the text depending on the complexity that the AI detected. So if it was like a neuroscience textbook, the, the, the words would flash slower, and then if it was like a news article that was written at a fifth grade reading level, of course it would go a lot faster. And that would sort of maximize both reading speed and comprehension. So it was this like, great in invention, but in, in, in a sort of one of the middle steps, it started pulling in scientific research and patents about QR codes. And we were like, oh, this thing's misfiring, dude. It's not, it's not working properly. And then in the la the the end step, it all comes together and it's like, oh, that's how you're getting the articles and, and, and reading information into the glasses is from either the glasses, if they have a camera or from your phone that links to your glasses. You scan the QR code and instantly the information is then shown for you to read in the glasses. So, you know, there were some emergent exotic heuristics that were happening that we did not program in. And um, yeah, so those were the ones that are harder to trace. So to your point, there are, now what I think we're gonna start seeing is these human cognitive heuristic were our starting point. But as we let these systems sort of improve upon themselves, they're gonna start figuring out these like sort of alien, exotic. Heuristics or thinking strategies. Yeah. That are, uh, some of them may be thinking strategies that we can then adopt as humans, that that's po. Right. That's certainly possible, and I do think they will, you know, interacting with these systems. The exciting part about it is they can actually make us smarter and think better, you know? We haven't, we don't have all of the end state cognitive heuristics of the people a thousand years from now that are living just like the scientific method didn't exist a thousand years ago. We have that now and that's a, that's a stack or of layered heuristics that allow us to generate new knowledge. And similar to that, this is deep invent. You can kind of think about it. It's a, it's a, it's a stack or a few layers of heuristics that allow us to generate new inventions from new generated knowledge, right? And so I think that eventually these systems will invent also new ways of thinking that are totally alien to us, and we'll, we'll never be able to adopt them with the, the, the, the nuanced differences between how our brains work and how these artificial systems will ultimately end up working. You made that comment earlier about super intelligence and like this is, you know, not super intelligence, but this is, this is one of the steps you would expect to see on the road to super intelligence. And, and I think one of the, one of the things about a super intelligence is it's going to have thinking patterns that are probably beyond our ability to understand even, and. I mean, that right there sounds kind of like what, what super intelligence would look like to me. You know, I, the only way for us to, um, you know, understand what super intelligence looks like is to start incrementally building it and interacting with it. Yeah. And there is never going to be a day where suddenly a new species of super intelligence is just immediately, you know, born. It's a gradient. Yeah. It's a process of emergence. Just like there's no one day when homo sapiens emerged. Right, right. You, you can't, you can't pin down what day that was. When you look back, only now you can see that the paths diverged between species. But you know, there wasn't a specific day when Hu when Homo sapiens started. And I think the very same thing is true. It may be a fast takeoff, but it's not, there isn't going to be like a specific moment. What I think there will be are a series of moments like this one where you're starting to see that it's doing something and you're forced to admit that wasn't possible before. I never experienced that. This is the first time I am personally experiencing that. I'm not saying it's the first time it's ever been experienced, but certainly I can deterministically say I've never experienced or heard of this until today. And that's what we felt like, you know, when it had innovated on its own architecture. And I, I think there's just moments like that where we have to sort of step back and say, look, okay, this is where we are, you know? And that. The, you know, we should have expectations of these systems that do certain things, and when they hit that milestone, we acknowledge that and we say, okay, we are on that path. There are a lot of companies right now that are saying they're working on super intelligence. They're doing it behind closed doors. They're aggregating, you know, they're doing a hundred million dollars offers to talent and they're, they're making a bunch of press and media stories. Where's the product? Which show me the, show me the thing it's doing that looks and feels and smells like super intelligence. That's the world that I wanna live in. That's why we're releasing this and building it out in public, out in the open, doing campaigns like Deep Invent for Good, letting people actually use it. You know, anyone for free to try it and experience it for themself. That's what I think we need to sort of do more of, is to innovate transparently out in the open. I think that's how we accelerate humanity. I totally agree, and I love the fact that like, all these became open source ideas too, which is, which is fascinating to me. I think, you know, I think, I think the biggest limitation on this, I mean, besides the potential for, you know, like we're gonna have to like, put in there how to, how to supply compute and, and cooling and all that for, for all this, you know, emerging technology. But I think the biggest limitation really on this is humanity's ability to adapt, to change. And I think I know what my next invention's gonna suggestion's gonna be in deep invent. I'm, I'm like, how do we get people, you know, ready to adapt to change at a, a huge scale, right? Um, because I think that's, that's where we, we constantly. Hit Roblox. And I think, I feel like we're in that period right now where we've got all this magic happening around us. You know, I mean, you know, a sufficiently advanced technology is indistinguishable from magic, right? And, and so I'm gonna start saying there's this magic happening around us. And, and, and yet people are fighting against it, you know? And, and, uh, and, and I, I don't know how that's all gonna play out, whether like some people go along with like, this evolution of our, our societies and, and, and that, and, and, you know, some stay behind, or we're gonna find some way to scoop everybody up and bring 'em along. But, you know, that, that, I always feel like, you know, it's like the Luddites back in the day, you know, fighting against the looms and things like that. You know, how, how do, how do, how do we, how do we get people on board with all this? I mean, I'm excited about it. You're excited about it. Some people are, you know. Like, what the hell is going on? I think this diffusion of innovation, you know, issue or concept, you know, it's permeated society in every major wave. Yeah. Um, I'll tell you a little story. So when I was a child, um, when I was, you know, nine, nine or so, eight or nine years old, um, be at my, my grandmother's house, you know, and, um, she lived in Wisconsin and we, and I would be there and I would be so fascinated to ask her about her life when she was my age, you know, and, and I wanted to just know what the world looked like and what, you know, you know, what the lack of technology, you know, like, and, and how, how could you live, you know, what was, what was it like? She told me something so fascinating, which was that, um, she told me the story of how electricity was being put in houses for the first time and how there was this backlash. Of electricity being installed in houses and there was this big resistance and people claiming it was gonna burn your house down and you couldn't do it. And in fact, in some cases that was true. They didn't have everything, like the installers didn't have it all figured out the installation well. They were running electrical through gas pipes. Which was slightly problematic. Right, right. And so, you know, um, but you know, in general it was like, it was an inevitable, we, we look back on it, you know, it was like an inevitable conclusion that we were gonna electrify America. Right? Yeah. But, uh, at the time it probably seemed like, yeah, this might, this might, you know, this, this new electricity fad, it's burning people's houses down. I don't know, it's probably a 50 50, you know, to, to some people. Yeah. And there were certainly Luddites that were like, I'm never gonna do that, you know? Yeah. Um, and I remember that my, my grandmother said that, um, her parents were, you know, not early adopters of it and, um, for that reason, for the fear of it. And, uh, you know, she, later in her life we was, um, I think she was a typist at IBM, um, you know, and so it just shows you like that we went from like, no, you know, reticent to install electricity in and then in her lifetime to thinking machines. Yeah. You know, and, and now like, so I think look in ev every major wave and shift of technology, like there's, there's resistance, there's diffusion of innovation. You have the early adopters and you have the laggards, and you have like sort of that, that main, you know, um, middle ground where, um, those, those people are, uh, where, where, where technology is like sufficiently diffused that it's like over 65% of the population. Right. And I think that the Rogers innovation curve, right, I think, which is what everybody talks about when they, they talk about crossing the chasm is sort of that, that shift from the early adopters to the, to, to, to the rest of the, the cycle of diffusion, right? And then also this, the sigmoid curves of um, how, how quickly it goes to mass adoption. Um, and I think that, that, that pace is getting faster and faster. And so it's starting to feel more and more magical. I mean, I even feel like that. I, I, I, I have like magical experiences with Deep Invent all the time where I, I just can't believe that it's doing what it's doing, you know? And, and so that just tells you how fast things are moving, that it's, it's hard to co uh, cognitively accommodate to them, right? To where you become adjusted and just used to it. Uh. We have a, a ma, an enormous capacity to like, just take technology for granted very quickly. Yeah. Um, Sam Altman talks about that a lot, and um, I tend to really agree with that. But there are ones where they just feel like they're a little bit further out there to where you almost can't believe that they exist right now. And, um, and, and, and it's gonna take a while to like kind of catch up to speed with that. So we have, when we talk to investors, you know, we, they'll, they'll, they'll, they'll start out a lot of times with like, how do we know this is like, you know, this even works. You know? And, and then they'll, they'll, they'll go back and they'll, they'll try it and they'll come back to the next meeting and they'll go, what are we gonna do if everyone's using this too much? You know, what, what does, you know? And it's like, it's, it's, it's pretty, yeah, it's a pretty wild time. I mean, I mean, you know, like, to, to get into it like yet another like famous chart. I mean, there's sort of that malthusian, uh, you know, chart where you have one part growing exponentially, one sort of growing linearly, I mean. You know, like certainly AI is that exponential growth. Avenue, but sort of the ability to ingest and adopt to change is kind of more linear in, its, in its progression. Right. And I, I do feel like we've got a little bit of that conflict going on, but you know, maybe that exponential curve will solve for that more linear growth pattern. Right. I think that technology adoption is, is still a major question. Um, yeah. And I think that that's important both for how people integrate it into their lives to capture value from it rather than just be subject to its changes on society. Yeah. Um, so I think that's a hugely important issue. And I think also from a macroeconomic perspective, the rate at which we're able to integrate it into our societal fabric of. Value generation and value distribution, that's going to be a critically important issue for us to solve as, as we move forward here, we need to make sure that the incentives are aligned so that the, the humans it's meant to serve are beneficiaries of the value, uh, that it, that it generates. Right? Yeah. And I, so I think there's some new thinking by David Shapiro that I really like on post labor economics and trying to understand, you know, that's huge first conceptualizing that Yeah. That, you know, labor might not be the, the end that the, the, the, the highest order pursuit. We've, we have oftentimes confused laboring or having a job with purpose itself, right. Um, or, or with the fulfillment that it provides, but the fulfillment. And the spending of our life hours in a way that's, that's, that's most, um, fulfilling, you know, for, for ourselves or for our family, our loved ones, people that we care about for society. If you're a person that likes to try to advance society, um, and, and. I think that those are like the aspirations that we should probably try to keep in mind and not clinging to necessarily that labor's the only way to get there, especially when we're at the precipice right now. We cannot deny this, that these systems will become capable of an enormous amount of labor embodied artificial intelligence through robotics is here. It's not coming. It's here now. Right. You know the, there, there are robots right now that can go into your home and do your laundry. Yeah. You know, you just haven't seen'em on your, on your corner yet. But they're coming. They're everywhere. And, um, and so I mean that's, you know, that means your plumber is gonna be a robot. That means that, you know, and, and then you've got also, of course the, the white collar work and the knowledge work that's, that's here now as well too. And that is diffusing through. And so we need to. Be aware that there are paths for value generation and sharing and value creation at where humans can be the beneficiaries of these systems. But we need our leadership and at the policy level, we need to work toward informing policy makers what's actually happening in these frontier labs. Right? And we need to be more communicative and more, you know, we need to interact with these policy makers in a way that's not just trying to funnel more, you know, government money to our, our private company to, to enrich ourselves, but to, to try to put policies in place that allow people to share in the benefits of these systems generating new. Value, you know, and so I think like letting people share in, in the value that the compute generates on a compute basis or setting up funds for every person that's that's born, you know, where they get $5,000 that, that goes into an account and then it, it earns a return that's, uh, you know, on the economy upon which all of this artificial compute is generating value. Because if you imagine, I mean, AI is going to com is eventually gonna generate more than 50% of the US GDP in value. Yeah. You know, it's not gonna be the humans. Right. And so we need to make sure that there are mechanisms to distribute that value generation from these artificial systems to the, to the people, you know, owning the wet nets. Yeah. Well, I mean like, like the post scarcity economy, right? I mean, and, and you, you look at, um, I mean, not to get super nerdy here, but you know, how, how can I not, uh, you know, like look at Star Trek. I mean that is sort of that utopian society that they created where it's like there was no. You know, there was no need for power there. You know, hunger was solved, all the like, kind of major problems with humanity were solved and everybody was sort of left to pursue their own interests and, and that brought about, uh, you know, uh, its own kind of revolution. Um, and, and I think that, you know, maybe in the face of all of this innovation and, and, and an emerging super intelligence, the role of humans in the universe needs to be sort of rethought. And, you know, maybe we don't ha I mean, you know, like e economies are, are entirely a fiction. And it's, it's, it's served us as a way to sort of distribute resources and, and, and those sorts of things and, and utilize resources in an efficient way, but. You know, what happens when you step beyond that and, and like, what, what, what is the role of humans in, in, in the world? You know, when that happens and we have maybe, you know, these SY machines and systems taking care of us and, you know, making wonderful lives for all of us. Um, and maybe that's where we belong. I mean, maybe that's how, how it all just ends up and we just kind of enjoy our. You know, our lives instead of just struggling and working all the time. Right. Yeah. And I think a lot of times, you know, there's, there, you know, humans are creatures of ideas. You know, that's, that's how we've progressed. That's what's uniquely different about our species. And so I think that allowing people to sort of have more opportunities to, to, to explore ideas and to ideate and then to hand them off to systems that can realize those, that's my version, my vision of a utopia. Right, right. Is like freeing up people to have the cognitive bandwidth to be able to ideate and then to be able to have a system that can work with that and interact with naturally that, that, that can, that can realize that idea and, and bring it into the world. And so I, I think that. That's where what I, I would say, you know, and I have, I, I have a bias, you know, here, uh, that I acknowledge, but I think that, you know, that, that that process of innovation and enabling people to do that is just that, that's what I see as a, as an aspiration above and beyond, um, just labor itself, you know, is, is freeing up that bandwidth. And then there's freeing up the bandwidth to love and to to love our family and to, to, to, you know, to love our life and our, our, uh, and create art present moment. Create art, you know, be, you know, be generative and creative. Uh, that, I mean, you know, I think that like, uh, a job that you're privileged enough to have that you can sit at a computer and you can crunch spreadsheets for 40 hours a week, uh, so that you can like, you know, incrementally move that line for that publicly traded stock of that company that you work for. That's great. You, and, and you can make a living, that's fine, but. We all know those people that experience that as a quiet hell and yeah. Um, you know, and they're, and they're spending time away from their family or they're, they're making trade offs of having children because they need to, you know, work this job. And we do need to acknowledge that there are trade offs that we make and that that might not be the end state lifestyle. That's just the, the period that we're in Now, remember that there was a time before we were an agrarian species before we did farming that, and that was a completely different way to live. And we had the same brains then. And we had the same brains. Right. And then there was a time when we were agrarian and we were forming little micro societies and we had a completely different way to live. Now we're sort of post agrarian, and that's been that, that that, you know, agricultural activ has been centralized and, you know, uh, you know, centralized to China in some cases. Yeah. And, and so now, you know, we have a completely other way to live, which is, you know, a lot of this knowledge work. Who would've thought in the 18 hundreds there would be something called knowledge work that you would call something that you do sitting in a chair work, you know? Yeah. Um, and so I do think that we have to understand that we are in, we are. Just a, a link in a long chain of human history of transformation and that, you know, there may be an enormous amount of prosperity ahead and there's no reason to just merely fear it, but to participate in the construction of a future that's fit for an optimist. You know, one of the things that distinguishes us from the machines, for better or worse, oftentimes for worse, but is, is we apply an emotional context to everything that we do. And, you know, maybe this is what frees us to explore that emotional context and explore the, that side of of humanity. And that brings about something different for us. Um, because, you know, all that other stuff gets out of our way that that's survival. Peace goes. We don't have to worry about that anymore. I'm, I'm just excited about, I'm excited talking to you. Makes me excited about the future. So, uh, you know, thank you for that. Well, th likewise, thank you. And for, for creating this platform for, for people to, you know, express their, their visions for the future. You know, it would be impossible for people like me to do this without getting people to know about it. I often say to founders that I advise that, you know, you've gotta do two things really as a founder, and that's it. You've gotta make something that's good. And I mean, the product has gotta be good, as in it works well and it should do good. It should be good, you know, net on net for society. So you gotta make something that's good and you've gotta get a million people. To know about it. Yeah. If you do those two things,'cause it's not enough to just innovate, you then have an incumbency to make people aware of it because you've gotta go from ideation to implementation. Yeah. So if you can get a million people to know about it and you made something that's good, you can do almost everything else wrong. And you can still have an impact, a positive impact on society. Thanks so much. This has been a, such a great conversation and I, I feel like we have to have a part two at some point.'cause there's, I feel like there's still so much more to dig into and as your product evolves, I mean, I'm sure like in, in, you know, a month or two it's gonna be something entirely different. But thanks so much for being on, it's been a really interesting conversation and uh, it gives me hope for the future. So Appreciate it, Chris. Always a pleasure. Always happy to come back. Yeah. Awesome. Thanks so much.