Kordel K France is the founder and lead engineer of pioneering machine sensing startup Theta Diagnostics.
For the better part of the last decade, he has been working with his incredible team to build a new type of nanotechnology sensor and multi-modal AI platform that enables machines’ sense of smell.
His driving mission is to give machines the sense of hyper-sensitive smell through machine olfaction and accelerate a new paradigm in robotics and AI.
Kordel is a technologist, inventor, and multi-disciplined engineer whose reward function is maximized by teaching machines to solve complex problems through AI.
While most of his recent work is contained in the bootstrapping of Theta and its acquirers, he has also had the privilege of working with and learning from incredible engineers at some of the world’s best companies in solving challenging technology problems.
Although his focus has been on the engineering of sensor systems and multi-modal AI, he has deep manufacturing and hardware development experience that enables him as a comprehensive product developer and technical leader.
His experience spans computer vision, autopilot development, aircraft design & simulation, natural language processing, swarm intelligence, and optics.
What you will learn
- How AI can give machines a sense of smell.
- The importance of integrating all five human senses into AI for achieving human-level intelligence.
- The role of machine olfaction in detecting medical conditions through breath analysis.
- The potential of AI in agriculture, starting with self-driving tractors.
- Kordel’s journey into AI, inspired by his childhood experiences on a farm.
- The evolution and current state of AI education, and the importance of math and statistics in AI development.
- How AI models need to be explainable, especially in critical fields like healthcare.
Transcript
Jeff Bullas
00:00:04 – 00:00:34
Hi! Welcome to the Jeff Bullas Show. Today, I have with me Kordel K, France. Kordel is the founder and lead engineer of pioneering machine sensing startup Theta Diagnostics. For the better part of the last decade, he has been working with his incredible team to build a new type of nanotechnology sensor and multi-modal AI platform that enables machines’ sense of smell.
His driving mission is to give machines the sense of hyper-sensitive smell through machine olfaction and accelerate a new paradigm in robotics and AI.
Kordel is a technologist, inventor, and multi-disciplined engineer whose reward function is maximized by teaching machines to solve complex problems through AI.
Jeff Bullas
00:01:14 – 00:02:13
While most of his recent work is contained in the bootstrapping of Theta and its acquirers, he has also had the privilege of working with and learning from incredible engineers at some of the world’s best companies in solving challenging technology problems. Although his focus has been on the engineering of sensor systems and multi-modal AI, he has deep manufacturing and hardware development experience that enables him as a comprehensive product developer and technical leader. His experience spans computer vision, autopilot development, aircraft design & simulation, natural language processing, swarm intelligence, and optics.
So, Kordel, welcome to the show. I’m really pleased to have you here and we’re going to have a chat about uh what different types of A I, how you think where A I is going, we’re going to have a lot of fun chatting about that. So, Kordel,
Jeff Bullas
00:02:13 – 00:02:34
What got you into AI? Did you watch the Jetsons? Did you watch movies? Did mum and dad engineers? I, how did this all start? What was this call to the Adventure of artificial intelligence research? Because you’re currently also in the middle of doing a phd in A I as well, aren’t you?
Kordel K France
00:02:35 – 00:03:05
Yeah, that’s correct. Um And yeah, just to preface, thank you for having me on the show and thank you to your listeners as well. Uh But yeah, so my original genesis for getting into A I uh actually started when I was a toddler. I grew up on a farm and my father was a big proponent of using new technologies to increase efficiency and to increase yield uh on the farm. And he brought home some self-driving software, uh self-driving software package for just one of these tractors one day and retrofitted it
Kordel K France
00:03:05 – 00:03:31
to the tractor to love to drive itself. And uh as a child, I’m watching this machine drive itself without any controls, without any interaction. And it was something straight out of a sci fi movie for me, something like I would imagine coming out of Star Wars. And so it was really, it was a really defining moment, really surreal for me as a toddler. And as I grew older, I asked my father, I said, how do I work on this? How do I contribute to this? I wanna be able to like, contribute to this field. And he said,
Kordel K France
00:03:32 – 00:03:50
Well, you gotta be good at math, you gotta be good at robotics, you gotta go into uh sensors and software development, et cetera. And so that was kind of the accelerant for my work and my North Star if you will for the rest of my life and my life’s work in artificial intelligence. Uh So believe it or not, it actually started with agriculture,
Jeff Bullas
00:03:50 – 00:04:34
right? OK. And that’s, and we’ve got some big farms here in Australia and um we do uh there’s a lot of machinery and development going on and like cropping is done very accurately. Now, isn’t it with GPS, for example. So, um, and then you’ve got drones checking things out to make sure the crop is ready or there’s, you know, all sorts of things going on. And, um, I’ve got a really good mate of mine who actually runs a 7000 acre, uh, sheep farmer and he realized that he wasn’t, uh, he didn’t, actually wasn’t growing sheep. His, what he really needs to do is be a grass grower because that’s what feeds the sheep. So he had to get very good at keeping optimizing grass growth
Jeff Bullas
00:04:35 – 00:04:55
because we created what I think is 200 paddocks, small size, they could have access to water in the corner and they grazed every corner of a small paddock rather than eating in the corner of a large paddock. So the tech is actually really quite fascinating. So, yeah, it’s, and it’s getting better, isn’t it?
Kordel K France
00:04:56 – 00:05:36
Yeah. Yeah, it’s interesting because there’s, there’s a, there’s a lot of room for innovation within agriculture. But b it’s, it’s always been funny to me because uh self-driving technology kind of was more proliferate in agriculture before it became uh more of a consumer tech, uh you know, automobile industry type thing, granted self-driving in a field with a tractor is a much simpler problem than self driving on a highway. But the technology was kind of there and was uh I guess it was a little more familiar in agriculture before it became mainstream on, uh, you know, Tesla and some of these other, uh, electric or, or uh, auto autonomous vehicle companies that we see today. But, yeah, there’s, there’s a lot of interesting things going on in ag, not only with A I, but just technology evolution in general.
Jeff Bullas
00:05:36 – 00:06:25
Yeah. Yeah. As we were talking before, I’ve been involved in, um, technology industry since the mid nineties when the personal computer started to become, um, much bigger. Well, it actually took over from mainframes. Then it took over from the mini computers which weren’t that mini and then the personal computer which was good, but then it got networked and then it got better and then the internet got connected to that and that got better. So we’re in the middle of watching the world become a global network, aren’t we? Really? And that has so much data that A I helps us make sense of it. So let me ask you this question then. So you, you, you’re inspired by this self driving tractor and
Jeff Bullas
00:06:26 – 00:06:37
they’re really good because they can’t, well, they can kill people but they’ll only keep, kill a sheep or destroy a fence if they actually go a little bit mad, mad, don’t they? But what was the
Kordel K France
00:06:38 – 00:06:41
playground? It’s a stand, a bike’s environment.
Jeff Bullas
00:06:41 – 00:06:58
That’s right. Yeah. Fences senses. So your dad said you need to be good at physics and math, um, and robotics and so on. So, is that what you started to focus on when you got to high school? What were the next steps?
Kordel K France
00:06:59 – 00:07:29
Yeah. So I was always really good at math and always had a passion for it. And uh so I continued that path throughout high school and in college and um I went into applied mathematics uh because theoretical mathematics just wasn’t quite there, I didn’t have the real world applicability that I wanted. Um But physics and applied math seemed to get me the right formula of robotics and material science. And um everything that I needed to basically handle the hardware side of robotics and then
Kordel K France
00:07:29 – 00:07:58
uh the software side kind of came later. But um if you, if you peel A I back and all the layers and all the frameworks that wrap a lot of these artificial intelligence models at its foundation is really mathematics and statistics. And so everything’s kind of built on top of that. So if you understand math and statistics and probability pretty well, you can kind of build up from there. And then if you become a decent, you know, software engineer, you can start building A I and then, you know, the hardware component comes in with merging A I into actual robotics.
Kordel K France
00:07:59 – 00:08:27
So I really wanted to kind of have this holistic uh this holistic suite of skills that allowed me to excel in a mechanical uh engineering field, software engineering field, electrical engineering field um and understand the theory as well as being able to be a practitioner to actually build some of these software or these uh artificial intelligence and robotics systems. So, yeah, math was kind of my North Star to guide me throughout the way. And then uh I picked up software and um everything else down the road.
Jeff Bullas
00:08:28 – 00:08:47
So you, you, you’re doing, you’re going to high school, you’re looking at that, doing good at math. Then you go to UNI and you did, you study, was there an A I specific um degree that you did or was it just math? Physics and different buckets of subjects?
Kordel K France
00:08:47 – 00:09:27
Yeah, back then, there were no A I courses, there were no A I courses. There were not even any machine learning courses. So it was just all math and statistics. The closest I got to machine learning was a course uh called advanced Time series. And now that’s kind of, it’s still, it’s kind of considered machine learning in some buckets. I guess if you look online, you’ll find some courses where advanced time series is now machine learning and all that. But back then it was basically the components of A I but not defined as A I if you will. Uh and then I took a bunch of programming courses that allowed me to apply those mathematics and statistics principles into real world applications through software. And
Kordel K France
00:09:27 – 00:10:08
uh but yeah, there were, there were no machine learning or A I courses back then. There weren’t even like any certificates that I could do to actually uh get that, get that as an achievement or get that on my resume. Um Whereas now I think they’re a little more proliferated uh within colleges and universities. So, uh it’s been interesting to see how the curriculum changed. Um Some of the titles of the courses have changed from what I feel are applied statistics to intro to machine learning or something like that. So, a lot of the building blocks are still fundamentally machine learning or for our statistics and math, but they just kind of have, have uh changed, it changed a little bit in their titles or uh thesis. But it’s been interesting to see how the fields progress in that regard,
Jeff Bullas
00:10:08 – 00:10:27
right? So you mentioned that uh you know, generative A R and what we see as jet G BT is really just, it’s, I suppose it’s the popular well known face of, of what A I is. What are the, what are the other types of A I that we need to know about?
Kordel K France
00:10:30 – 00:11:11
So, yeah, the, the proliferation of generative A I has really given a lot of folks um given the field more attention and there’s been a lot of business value and a lot of just a lot more interest that’s entered the field because of the use of generative A I. And this is fantastic because we’re getting more research, we’re getting more focus and um there’s a lot of tension on large language models. But I do like to remind people that there is a whole other realm of artificial intelligence out there. Beyond large language models that folks are still studying, folks are still doing a lot of research on and that we still have the opportunity to capitalize on in a similar manner that we did with large language models. Um you know, just a couple of years ago,
Kordel K France
00:11:11 – 00:11:50
One thing that’s particularly exciting to me as we move towards the artificial general intelligence realm is the development of human level intelligence and enveloping that with all five senses. So what I mean by that is um we have, we have vision, right? We humans can see with their eyes, we can hear, we can taste, we can smell. There’s a camp of folks that believe human level intelligence can’t be achieved unless all five senses are incorporated into an A I model, you might get, you might get something very intelligent. But if it is, if we’re defining human level intelligence as a G I, then you have to have all five senses now. Uh
Kordel K France
00:11:51 – 00:12:21
So going back to your question, one thing that’s interesting to me is trying to develop these senses that aren’t necessarily getting a whole lot of attention right now. And uh machine or faction or machine smell is one of those that I don’t is one of the uh realms of A I that I don’t feel is quite as mature as machine vision or machine audio. Uh So there’s a lot of opportunity there. I feel like to kind of level the playing field with the rest of the sensing modalities to try to bring out this legitimate form of human level intelligence if you will.
Jeff Bullas
00:12:21 – 00:12:27
Right. So what would the benefit be to have a machine that has a sense of smell?
Kordel K France
00:12:30 – 00:13:05
So one could argue that if I’m sitting outside a cafe in Paris and I’m eating a baguette, I don’t get the same experience by just observing the passers by and hearing this, the uh chatter of people speaking French. Um If, if I don’t have the sense of smell, I’m not smelling the baguette, I’m not smelling what else is being baked in the back of the bakery. Um I still get an experience but I do lose an element of that experience, right? So being able to build that into machines is something that’s attractive. Now, that’s, that’s kind of the human psychology portion of it.
Kordel K France
00:13:06 – 00:13:32
But there are a lot of other use cases that can be built or integrated with the use of these artificial uh smell sensors. So um I’m 11 area that I’m particularly focused on with my company at THETA diagnostics is we’re trying to build these sensors to detect different medical conditions over people’s breath. Now, this is one application of many but uh you know, several, I think it was a few 100 years ago,
Kordel K France
00:13:32 – 00:14:14
a physician would actually smell your breath as a proxy of how healthy you were. Right. That’s kind of gross and not sanitary now. But they would actually smell your breath to see like, ok, what there’s an indication there of what ailment you might have. People don’t do that anymore because it’s not super sanitary. But the principle still applies. And there’s a lot of research backing, um, the ability to detect the chemicals that are emitted when folks exhale. And so that’s one application of these artificial fact, these machine or faction sensors is that being if we can uh is being able to program them to detect the number of chemicals and the level of chemicals coming out of your breath as you breathe and using that as a proxy of your well being going your general human well being,
Jeff Bullas
00:14:15 – 00:14:47
right. Yeah, it is. And this race, some research being done regarding using dogs to actually well detect drugs is one thing but then also disease. And then there’s one lady who said she could actually smell dementia because she picked it up first. I think it was the New York Times article I read yesterday actually, so she actually could smell dementia. So that was um so I didn’t dive into the article too much, but so the sense of smell is made like you said, it’s almost like an unexplored frontier, isn’t it?
Kordel K France
00:14:48 – 00:15:27
It is. Yeah, if you look at dogs right. You mentioned dogs and uh canines are very good at, I mean, a, a large part of their brain is attributed to the sense of smell, much larger part of the brain than humans have. And they can track scents that are uh kilometers away. Some studies suggest multiple, many kilometers away and they can detect at a resolution of a part per quadrillion, which is extremely small as a level of comparison, a smoke detector can only detect at a part per 1000 or maybe a very, very low part per million uh resolution. So there are several orders of magnitude, dogs are several orders of magnitude more sensitive than a smoke detector.
Kordel K France
00:15:27 – 00:15:55
Now, if we can build that capability into machines, then we can get a lot of the same digital replicas in machines that we can with dogs. So there’s a host of applications that can be applied there where uh you know, dogs can track, uh they can smell drugs, they can smell explosives, we can build the same uh uh the same functionality into machines. If we can replicate their noses, the dog’s nose to a high degree. Um but there’s, there’s a lot that we can do in medicine as well because
Kordel K France
00:15:56 – 00:16:25
um and if you follow some of the literature during COVID, they were training dogs to detect COVID by smelling it. And you, we, you can do the same thing with the, with the machine or faction sensors as well. So uh if you are able to really replicate all the use cases that a dog can do is kind of a general proxy for what we can do with machine or faction sensors. But that’s not the, that’s not whole, that’s not, you know, everything we can do. It’s just a good proxy for what we can do.
Jeff Bullas
00:16:25 – 00:16:46
Yeah. So for diagnostics, you basically try to add a layer of software on top of machines to actually improve it to any machine that has sensors. Is that correct? Is that part of your overall mission is trying to build more intelligence into any sensing machine?
Kordel K France
00:16:48 – 00:17:15
Partially. So we actually had to build our own sensors. Um So I mentioned a moment ago how dogs have a part per quadrillion resolution? Our sensors have a part per trillion resolution which is very much more sensitive than other sensors on the market, but we’re still not quite as good as dogs. But so we’ve had to build our own sensors. And with that, we’ve had to build out a uh a ca a host of other capabilities um towards something called multimodal machine learning.
Kordel K France
00:17:15 – 00:17:41
So that means basically we’re pairing our smell sensors with vision sensors and audio sensors so that we can have multiple scents, uh multiple senses if you will uh while we’re doing the mach machine of faction detection. So as a frame of reference, when a dog is navigating on a scent. It’s primarily using the modality of smell, but it also uses its eyes to avoid obstacles and its ears to uh have situational awareness about what’s going on. So it’s using three modalities,
Kordel K France
00:17:42 – 00:18:06
We use vision and other uh sensing modalities to try to get a better idea of the entire environment that we’re, that we’re, that we’re detecting or that we’re smelling. And so that’s a large part of what we’re building into our device. It’s not only work with our machine or faction sensors but to work with other sensing technologies as well. So um we had to actually invent the sensors and build them from scratch and we’re building that to integrate with other technologies,
Jeff Bullas
00:18:07 – 00:18:33
right? So where did the inspiration come from to sort of look at the sense of smell? Um Where did there was an aha moment for you regarding that? Like when you’re doing research and going hang on the smell’s been forgotten by the A I community, for example or I don’t know, but where did that inspiration to go? OK. We need to look more closely at doing and also your phd is in the same area as well, isn’t it?
Kordel K France
00:18:34 – 00:18:35
Yes, that’s correct.
Jeff Bullas
00:18:35 – 00:18:46
So where did the inspiration come from? For the sense of smell is to be integrated into A I um and into machines to complete the picture if you like of the modality of A I?
Kordel K France
00:18:48 – 00:19:32
So before theta was started. Uh I, I was, I started a small start up called Secret Technologies. And we had built a machine learning platform that specifically enabled device learning. Uh So you didn’t need to be connected to the cloud now, a few years ago, 10 years ago, that was a much more impressive thing than it was now because doing A I models on the edge or on uh edge devices, which was a much harder thing. And we started looking at the general trend and we started to see that, you know, the success we’ve had so far might not be the same moving forward. So we might, we might uh lose our moat if you will because um machine learning models are getting smaller uh by just by definition. So our compress
Kordel K France
00:19:32 – 00:20:16
algorithm might not be super attractive in the future. And uh the amount of compute being put into edge devices is just growing. So we were looking around and trying to figure out what’s something that uh we can work on, that’s not necessarily getting as much attention as some of these other fields in A I. And uh since I was in college, I’ve been always fascinated by a lot of the abilities, a lot of sensing techniques out there that can uh monitor the environment and break those, break the environment down the air particles into its constituent uh compounds. And we looked to run out of the technologies and a lot of them were functional and they worked, but there hadn’t been a lot of innovation in the field for several years. And uh some of the,
Kordel K France
00:20:16 – 00:20:44
Some of the uh critiques of the, of the, of the mach machine faction were that the sensors were very slow, they weren’t sensitive enough and that they couldn’t detect multiple compounds at once and those weren’t bound by the laws of physics. Right. So it’s not that it couldn’t be done. It’s just that there hadn’t been such a huge, large amount of innovation in the field. So we try to basically look at that a little bit deeper and build out the hardware and software for it. And we built uh we, we, we always try to make data driven decisions.
Kordel K France
00:20:44 – 00:21:13
So we survived a bunch of data and the data actually pointed towards being able to detect a much finer resolution towards like that part per trillion resolution I was talking about and being able to detect multiple chemicals simultaneously, which is another critique of existing sensors. So um we did a bunch of hardware development and you know, it was a lot of pain and a lot of trial and error. And eventually we came out with a sensing technique that worked. And um from then forward, we were able to kind of carry them on into several different industries.
Kordel K France
00:21:14 – 00:21:50
So the seeker eventually started with something completely different and became FTA diagnostics. But um we became or we, we, we, we, we started by building out the machine learning platform that eventually brought forth the mach machine or faction sensors. So it’s kind of a chicken and egg problem. Um But uh yeah, it’s, it’s, we still even today, don’t see a whole lot of attention on mach machine or faction. Uh There’s a lot of attention being developed into machine vision and audio and being able to identify patterns and sounds and images. But I’m still, it would be great to see some more research in the field to try to move it forward because there’s a lot of opportunity there.
Jeff Bullas
00:21:52 – 00:22:37
Yeah, I suppose the challenge is it’s also almost invisible. You can’t see the smell, you can’t hear the smell of it. It’s a look. And for me, I appreciate smelling in a different way because I had some real sinus issues and there’s a new uh sinus spray came out called the mi I would go for a ride with my mate and my, my nose would be running and it would just, and then on top of that, then my nose would be blocked. So my sense of smell started to disappear, which means that affects your sense of taste. Yep. So for me, I felt like I was having part of me removed because of that issue. So for me, when I use this new spray,
Jeff Bullas
00:22:38 – 00:23:17
uh which happened to be an accident discovery, combining two that were used individually and put them together and the spray became like this magic drug almost, um, an, a bit over the counter in Australia. But, um, suddenly I could taste food again. I could walk down the street and smell flowers or women’s perfume and it just adds to the whole dimension of walking the street going part on the right, you know, spring flowers, you know, the fall sense, smelling the ocean, you know. So these are all adding to our, you know, senses as humans, which actually makes us more human, doesn’t it?
Kordel K France
00:23:19 – 00:23:59
Yeah, I fully agree when you don’t appreciate one sense uh as much until you lose it, right? And I think that’s part of the driving mantra behind human level intelligence is if we really want to develop human level intelligence, we’ve got to incorporate all those sensing modalities into A I with large language models, there’s a huge amount of attention on the psychology part I feel like which is can it think? Can it actually construct a thought and can it build, can it do critical thinking on a problem? But that’s really just one fundamental aspect of human level intelligence, right? When you incorporate all these different sensing, mental sensing modalities with vision, smell, audio taste, et cetera. To me, that’s what really constructs
Kordel K France
00:23:59 – 00:24:25
human level of intelligence and a body, right? I mean, if, if we just have the brain, if we just were floating heads that could have all these senses, but we didn’t have a body, right? We lose something uh as, as, you know, as part of that human level of intelligence. So, um there’s a lot of interesting fields uh or a lot of interesting work that could be done with humanoid robotics in that regard. But I fully agree with you. Once you lose one sensing modality, you kind of uh you realize just how important it was.
Jeff Bullas
00:24:25 – 00:25:08
Yeah. Yeah, exactly. So, you know, part of going for a bike ride is you go, there’s all sorts of sensors going on, right? It is. And I’m talking about a bike that you actually have to push with your legs, right? So um but for me, it’s such a sensory experience and I don’t understand people for me that just want to sit on a bike inside and just peddle for me that is missing, actually almost the entire point of being feeling like a teenager again and going fast down hills and being on the edge of danger and having that adventure and the sense of smell, speed, everything that makes us human. Now, I have a question arising out of humanoid robotics, for example. So
Jeff Bullas
00:25:09 – 00:25:54
and there’s something quite interesting at the moment in the investment community in the A I area, which is that Elon Musk said you need to pay me another $56 billion to keep me interested. Um So I can lean into A I because Tesla’s much more than a car and he believes that the car’s self-driving is gonna open up a whole new world of opportunity. On top of that, he’s actually developing humanoid robots and that’s where he believes that they will be worth a trillion dollars a year. I’m interested in your view on, you know, we all look at A I and robots and we’re human, not robots. That’s where it is, that’s where A I is. And we look at the Jetsons and we look at, you know, Star Wars,
Jeff Bullas
00:25:54 – 00:26:30
What do you think is, is? I think the opportunity actually lies elsewhere and human aerobics are just a sexy front end for something that maybe will never take off just like, you know, Apple’s new, you know, head goggles, right provision, whatever they call them. So, which has been a failure by the way, so far. But maybe they’re just a good idea, but 10 years too early, I don’t know. So I’m intrigued by what you think about the potential of human-like robots and you are building, you know, stuff for robots. And I think you’ve worked at Toyota who builds robots for production as well.
Jeff Bullas
00:26:30 – 00:26:49
So I’m, I’m interested in your thoughts on the humanoid robot potential that is sort of there. Um And it’s the big, you know, holy grail for A I for Tesla maybe and maybe uh Elon Musk, I’m interested in your thoughts.
Kordel K France
00:26:50 – 00:27:13
Sure. So there’s several different hats I can wear to kind of approach this uh answer and uh some of them are conflicting, but from one perspective, I, so a few years ago, I used to think that we needed to kind of adapt the world to robotics in order to prepare for them. Uh So what I mean by that is we needed to put certain markings on the road for self driving cars and
Kordel K France
00:27:13 – 00:27:42
we need to uh make it so that um there were no steps for, you know, home robots so that they could roll up the stairs they could roll up the ramp or uh they weren’t gonna be able to use stairs. So we’d have to build a way for them to get up the stairs. And I think that was the wrong method of thinking because if we’re going to truly realize robotics, I think we’re gonna have to basically make robots adapt to the world instead of the other way around. So now I am able. So having humanoid robotics is effectively exercising that mantra
Kordel K France
00:27:42 – 00:28:13
because the world is designed for humans, effectively, everything we do is everything we interact with is literally designed for is designed for us of our species. So by building artificial intelligence that can embody humans, we enable the entire world for robotics. So I think it’s, I think it’s necessary to some degree that we build humanoid robotics and really enable them so that we can have the full effect of artificial intelligence and uh the robotics industry in general. But I don’t think that that’s entirely, um,
Kordel K France
00:28:14 – 00:28:42
I don’t, I don’t think that’s entirely conclusive. I just, I think that’s, that’s one angle uh, of thought that I address it with. Um, the other thing that’s been interesting to observe is how is the adoption of people to A I, because as, as A I is moving forward, you see this growing fork of thought where folks are very appreciative and they say bring on the robots and others are very against it and they think uh Terminator is about to rise up tomorrow and
Kordel K France
00:28:42 – 00:29:14
That division seems to be getting bigger and bigger. And so I’m curious on how robotics companies are going to harmonize that, right? Because uh you know, Tesla was very much appreciated for its electric cars and how much they did for the environment and everything. And uh people have different thoughts on that, but as they start building out A G I, right, their shareholders and their appeal is gonna start to change. And I mean, Elon Musk said several times that he wants to contribute to A G I and that Tesla is one of the closest companies to building it. So uh you’re kind of,
Kordel K France
00:29:14 – 00:29:27
you’re kind of changing the company’s direction a bit and I don’t know, I don’t know what to think about that yet. I think he’ll be successful no matter what he does, but the public perception of it might change. Um So, yeah, what are your thoughts on it, to be honest.
Jeff Bullas
00:29:29 – 00:30:05
Uh, I think it’s going to be a lot like a self-driving car. Um, it’s been a promise for what, about two decades? And, uh we still haven’t cracked that nut, have we? Um, and I think, do we need a humanoid robot? Um, so for me, I think it’s maybe a good idea, but leaning into the research behind it produces what we call most side effects that actually produces other opportunities, I think. And I’m so for me, I think it’s, he’s selling a dream and a vision to the shareholders.
Jeff Bullas
00:30:05 – 00:30:41
I think the reality is maybe two or three decades away. Maybe the singularity as Ray Kwell talked about is 2050. Maybe that’s the time frame. Um self driving car. I reckon that is simple compared to a humanoid robot, right? That’s my take on it. Look, we’ve all had great ideas, but it’s sometimes that idea is, you know, a decade or two early. Um I read an article recently that the second movers are the ones that make the money and you only have to look at apple versus nokia to realize that’s true.
Jeff Bullas
00:30:42 – 00:31:10
But anyway, I am fascinated by what the future holds. I remember watching the Jetsons, you know, cartoon series on TV when I was a kid growing up. Um So for me, I’m fascinated, um I think Elon Musk is a great salesperson and can sell the dream and I think we do need those people in the world to sell dreams, to bring everyone forward and to evolve and grow because, you know,
Jeff Bullas
00:31:11 – 00:31:52
shit, it’s a lot of fun at the moment isn’t what’s happening in the world. It’s both good and bad but there’s a paradox of any invention that’s good and bad. Like, you know, gunpowder, you know, nuclear energy. The list goes on and on social media. It connects us, but it also disconnects us. And so, yeah, any technology is gonna be used for good or evil. But, um, I think, um he sold a good story to the shareholders and uh he sold the fact that he’s gonna increase the, you know, that’s gonna be a trillion dollar company um, within a short time frame. But I think we’re talking a decade or two, at least that’s my guess.
Kordel K France
00:31:54 – 00:32:17
Yeah. Yeah, it’s, it’s still a ways off for sure. Um, you had uh Nick Bostrom on the podcast a few episodes ago and uh his book on Super Intelligence is one of my, my favorite books. I’ve, I’ve read it several times and in there, he mentions the fact that there’s gonna be a day where the thought of being able to produce a G I is not necessarily going to be a thought and it’s gonna be a capability. And I feel like
Kordel K France
00:32:18 – 00:33:00
Until like this last year or two, we’ve kind of been passing the ball like we’ve been on the zero or we’ve been on the, you know, you know, 8090 yard line and we’re just passing the ball, but a touchdown is not really feasible. Right. We can’t, we’ve never really, we never really thought we’d get there and now we’re at the point where it’s like the next play might be a touchdown. We should reconsider what we’re doing. Right. Everything we did all the research that’s being done in A I is now one degree closer, very, very close to enabling A G I and uh or maybe not A G I but very, very capable artificial intelligence systems. And so there’s a different level of sobriety because it’s like, well, you might just build something that’s
Kordel K France
00:33:00 – 00:33:36
pretty damn powerful with this next move you make so be careful. Uh Because before it was just like, oh, this is a much better image recognition model. Oh This is a much better sound recognition model. Oh This is, we can do protein folding now, but now it’s like we’re very, very close to that touchdown. And you know, it’s, it’s, it’s uh cause for pause just to say, should we, should we throw the ball or not? And uh so I, I like a lot of Nick Bostrom s thinking um I still think we’re a ways off from actual true A G I but it’s uh we’re getting closer to, we’re getting closer to demanding the level of sobriety that needs to kind of lock everyone down and say, should we build this thing or not?
Jeff Bullas
00:33:37 – 00:34:37
Yeah. And any useful technology is the intersection of multiple technologies. You couldn’t build a smartphone unless you had small processors. You didn’t have, you know, really good actual material technology as well. Um Then you there, it’s just so many layers of technology in a smartphone. It’s mind boggling. But um that’s the, so the smartphone, the Apple phone is the intersection of many technologies um including the internet. So, you know, if you just had a phone, you couldn’t call anyone that’d be completely useless, but it might look nice. So um but they heard but interviewing uh Nick was really interesting because I think what we’re seeing too is the intersection of philosophy and science, which actually has just about always been the case because uh physicists, for example, are trying to find out,
Jeff Bullas
00:34:38 – 00:35:28
you know, the secrets behind the universe. How did it start? Right? In other words, physicists actually are asking really big damn questions. Where do we come from? Where are we going? And what do I find? Fascinating. But for me with A I is and for me it’s a bit of an aha moment. It’s going OK. ChatGPT, you can write a damn fine article. Sure. It’s losing a bit of soul. It’s not good at storytelling. It’s not as good as a human writer. And I only have to go and read some, you know, fantastic articles on the New York Times, um, from these incredible writers, ah, that do these connections of dots and bring together a story that just blows me away. And then the writing is creative. It’s storytelling. So for me,
Jeff Bullas
00:35:29 – 00:35:43
when the machine ChatGPT wrote an article for me, I went, wow. Ok. So, you know, we publish articles I write. So I’m going to the machines to do the writing. Why am I here?
Jeff Bullas
00:35:46 – 00:36:48
What’s the purpose and meaning of life? And that’s what Nick does discuss in basically deep utopia’s latest book and Superintelligence. He wrote in 2014. Um And having a chat with him was almost uh I was slightly intimidated because he’s got such a big brain and where it goes is fascinating. Um He even goes with things like, um well, what’s man’s purpose and meaning? Well, we can artificially create that as well. And like what? OK. So that’s just a snippet of Nick Bostrom’s brain. So I was really privileged to actually interview him. But yeah, for me, as humans now, we actually have to ask some questions before it escapes from us. And we should have asked these questions of other technologies like social media before it escaped from us and started splintering the world and became a force for, you know,
Jeff Bullas
00:36:48 – 00:37:19
to separate us rather than connect us. So, um yeah, I’m, I’m fascinated by that whole area of, of and you’ve mentioned an AGI . I think people need to know what that means because a lot of people don’t. So we have generative A I, which is generated out of, you know, lots of data learning neuroscience, large language models, but artificial general intelligence where the intelligence becomes as smart as humans, isn’t it?
Kordel K France
00:37:21 – 00:37:35
Yeah, that’s one camp of thought. Uh some say it has to be, well, not just human level intelligence, it’s got to be consciousness and that’s what constitutes human level intelligence. And so then, well, then there’s different definitions of consciousness, right? I mean,
Jeff Bullas
00:37:35 – 00:37:39
we don’t even know what cons this is. Humans do not know what it is.
Kordel K France
00:37:39 – 00:38:14
Right. Right. I mean, I, I’m personally not gonna push my nose up at if, if, if someone showed me an artificial intelligence that was just as smart as a dog and was just as conscious as a dog tomorrow. I would not, I’d be pretty freaked out, I’d be excited and freaked out. Right. I mean, it’s not human level intelligence. So I’m not gonna balk at that. But I mean, that’s, that’s very impressive, right? Just to even have that pedigree of, of, of intelligence and consciousness. So uh yeah, but for the sake of conversation in, in most realms, I think A G I artificial general intelligence is reflective of human level intelligence or better.
Jeff Bullas
00:38:14 – 00:38:52
Yeah, it’s um like for me, the A I, we, we just started this journey this, you know, I look back at, you know, the rise of the personal computer, then I’d see the rise of the phone, I’d see the rise of social media. I’d see the internet, the web. Um What I’ve found fascinating with ChatGPT is it gave humans the general, you know, population. Um It gave me a friendly, simple user face and that’s what browsers did to the internet back in the 19 nineties.
Kordel K France
00:38:54 – 00:38:55
That’s a great analogy.
Jeff Bullas
00:38:56 – 00:39:36
So that’s, and if you have a look at different technologies, that’s what happens. Look at the PC. Back in the 19 eighties, the PC was given a nice friendly user face when a mouse was brought into the equation and you could click and go, it ended up giving it a nice, easy to use. But the general population you use rather than, you know, the geeks um you know, that sit in dark labs and uh invent things and uh you know, have changed the world. Um I love the phrase um There’s a phrase in the Bible called the Me Shall inherit the Earth. I think the real truth now is that geeks will inherit the earth. Really? So yeah,
Kordel K France
00:39:36 – 00:40:03
go back to uh Bostrom. He argues in his book, Super Intelligence that um those who helped bring A I forth will be saved to some degree. I mean, I’m, I’m using that loosely but it’s like the robots will look back and be like did you help me, come about if so, you know, I’m not going to, I’m not gonna do bad things to you. Otherwise, uh, you’re gone. Um, so it’s kind of like a very pessimistic way to look at things but I, it’s just kind of funny. Yeah. To your point.
Jeff Bullas
00:40:05 – 00:40:07
But, yeah, you know. Yeah. Sorry. Keep going.
Kordel K France
00:40:07 – 00:40:45
Oh, no, I just, I was gonna mention I was, I was talking with a university student the other day and uh she goes, well, what do I work on? I mean, I am just getting so good. Like I don’t know what I’m, whatever I study at university, it’s just gonna be able to do better than me. So, what do I do? Right. And you mentioned a few moments ago, well, what, what role do humans play? And I told the student, I said you can’t have that mentality because if everyone starts thinking that way there’s gonna be no validation on artificial intelligence, nobody’s going to check it. Right. If you have, if you have a subject matter expertise and you go to school and you, and you become very, very good at psychology, right? You can correct
Kordel K France
00:40:45 – 00:41:03
chat GP t when it’s wrong or the A I, when it’s wrong about psychological psychology related questions or psychology related matters. If we all just kind of sit back and just say, well, the A I is gonna do it. The A I is smarter than me then. I mean, it’s, it’s not going to be perfect all the time. It’s going to be wrong and, but it might always believe it’s right.
Kordel K France
00:41:03 – 00:41:22
So if we all sit back and just kind of let it take the front seat, we can paint a very, very bad picture of the world in which we lose total control. So it’s important for us to always be ii . I think it’s important for humans to still try to have that subject matter expertise because even the way I can do it very well, many things very well. It’s gonna need to be checked always in the future.
Jeff Bullas
00:41:23 – 00:42:09
Yeah. And we only have to have a little look at uh Google adding um A I to its search called Google Reviews, right? So, I read someone who had done some research on some of the results and used it and shit. It was funny. It was so funny. It’s like medical research. We’ll eat a hot rock, that’ll fix things, right? So is this his checks and balances? And for me, I really see A I as being an assistant to us, as humans. We are already hybrid humans. We have our smartphone, we don’t even remember phone numbers anymore, right? We wanna find something out. We’ll just tap into the universe into the intelligence of the planet via the web. This is a, it is already here. We are already hybrid humans.
Jeff Bullas
00:42:11 – 00:42:49
So for me what I’m really excited about with, you know, generative intelligence is that with large language models, it’s tapped into the super intelligence and the super consciousness and brain of the planet of humans. And what I do when I get a result is the result of a collective organization distillation of the intelligence of the planet right in front of me within that document or whatever question they asked or whatever prompt I do. So for me, I’m going to my God. I actually have the intelligence of the planet at my fingertips right here right now. And that becomes,
Jeff Bullas
00:42:49 – 00:43:39
we’re gonna become more hybrid machines versus inhuman. And what we can do then is use it as an assistant to make us better humans to free us up to do things that I think make us even more human, such as having more adventures, doing more creative things, creating functional beauty, making, you know, beautiful architecture. I think it’s, but it’s got to be done in the right way. And I think this is where philosophy becomes really, really important in terms of why are we here? Right. So yeah, can I just pause something here. Go on. We’re back. I had to stop that because we’ve got some construction going on. Sorry guys all listening in, we have some construction going on at our place. So it’s all right. Um Yeah. Where were we? I just got disrupted by a jackhammer.
Kordel K France
00:43:40 – 00:44:04
No worries. Oh yeah, you were, you were I, I was, I was thinking about the framework that you kind of just, uh, positioned a minute ago and I really like that. Right. Like, you have all the intelligence at your fingertips and as a matter of what you do with it. Right. That’s a really interesting way to think about it. I mean, part of me thinks that I, I kind of almost the, the state of A I right now is almost analogous to me to like,
Kordel K France
00:44:04 – 00:44:25
uh the state of gasoline engines, like, or, or, or, or, or automobiles back in like the early 19 hundreds, if you were to show an automobile to Caesar or somebody else from like 200 BC, I mean, they probably think that’s a I, we would look at that now and be like, that’s not A I, that’s what are you talking about, but it’s a machine that automates something, right? And
Kordel K France
00:44:25 – 00:45:01
Every, every level of automation I feel like becomes the next frontier of A I. And it’s just a matter of like, like computers back in the eighties were extremely intelligent to someone in 18 hundreds. Right? Like that would be artificial intelligence. Now we look back and be like you had a kilobyte of raM, what do you? That’s not, it’s nothing. Right. So, I, it’s, it’s just an interesting shift like the frontier for what A I is and auto and, and how we automate more and more and more uh, it is kind of, I don’t know, there’s an interesting dance there. I don’t know what to think of it yet, but it’s, it’s, it’s definitely, there’s definitely, uh, some analogy there.
Jeff Bullas
00:45:01 – 00:45:47
Yeah. Yeah. And I think looking for analogies of the past and the present to help us make sense of the future I think is really, really important because the thing is the future, we’re all making predictions of where it’s going to go. Um You know, and for me, part of that question is, well, how do I get a sense of where it’s going? And so I wrote an article which was about the three pillars of artificial intelligence and to see where the money, the venture capital money is going. Who, what are the big, you know, research labs like a Apple spends 20 billion a year on R and D
Jeff Bullas
00:45:49 – 00:46:30
I believe it. Yeah. The reality is that if you, for me, wanted to look at where the money was going and who was behind the money and then you start to get a sense of where it’s potentially going. And because I think the challenge for me with artificial intelligence and you both, you and I, you’re on the technical part, I’m more on the business human side of what A I means and how we can use it, whether it’s a business or our personal lives for the benefit of mankind. And I’m going OK. So how can we, what is really going on. So for me standing back and looking at the whole ecosystem
Jeff Bullas
00:46:30 – 00:47:17
gives me a poor perspective because the trouble is, it’s just like going to Manhattan. You get lost in the canyons and you can’t see, you know, New York for the trees, the skyscrapers. So it’s actually getting above the game and looking down on it as if you’re a God and I, I sort of have this thing about that. We are all gods really. We don’t have to go looking for God. I think that um what’s within us is just magical any rate. So, um but for me, I just want to get to think as humans, we’ve got to stand back a bit and take a pause and look at it in all its perspective. And um that’s happened to me a few times and, and I’ve tried to invoke that. Um for example, social media
Jeff Bullas
00:47:18 – 00:47:56
transformed my life. I started blogging about it, and wrote about it. Then I ended up having been invited to speak all around the world. And for me, social media connected me to the world and to my tribes. And so I was in, I was in Egypt, for example, and I was um invited to the World Youth Forum and there were 10,000 young people hosted by the president of Egypt. And he hosted a roundtable with five people for social media and five people against social media. In other words, the pros and cons and for me that was an aha moment to actually realize that
Jeff Bullas
00:47:57 – 00:48:36
technology can run both ways and the truth is gonna lie somewhere in the middle. So we need to get our heads around how we use A I for example, now and get ahead of the game because social media is we’ve lost that, that gigs out of the bag, right? So trying to control it now is almost impossible. But so for me, trying to get context and having a conversation with you is part of that context. And what I love about what you’re doing is that you’re going um we’re not fully human unless we can smell and then OK, let’s create a machine and I’m going, wow, that is fucking great. So
Kordel K France
00:48:36 – 00:48:37
well. Thank you.
Jeff Bullas
00:48:38 – 00:48:51
Yeah. Yeah. And I just um and I’m never for me, you know, Kordel, I’m just fascinated by human creativity and I think I’ve got a long way to go. I’d be interested in your thoughts on that.
Kordel K France
00:48:52 – 00:49:17
Yeah, I, I fully agree. I think uh first off, I really like that way to think about it, right? If we can look, social media is more or less uncontainable now, right? We can’t really go back and if we can study that trajectory and how we got to this point, we can try to make sure that we don’t get the same way with A I right to a point where we can never go back. Um that’s a really, that’s a really insightful way to think about things. Um,
Kordel K France
00:49:19 – 00:49:56
for one thing that bothers me about A I. All right. Or this current state of A I is these big companies right now are focused on making bigger and more intelligent artificial intelligence models. And they’re, everyone’s kind of in the shock and awe stage right now because they look at, they use chat GP T and they get a really intelligible result back. They get an essay written for them. Um they get their emails composed, they get a paper written whatever. Um and they get all these answers and it’s very, it’s very amusing and we’re still kind of in the honeymoon phase but it where there’s not. And because of this, we’re kind of overlooking transparency and explainability. So
Kordel K France
00:49:57 – 00:50:26
what if, so we give, so we work a lot with healthcare, right? And they’re very recalcitrant to using A I because if an A I declares uh someone, you know, makes a life changing decision that’s extremely, extremely bad. So we have to kind of follow the mantra that A I cannot be the last link in the chain. We have to give as much evidence as possible to the physician and then they make the final diagnosis. And we also were being held to the level of scrutiny where we have to make our A I explain itself. So
Kordel K France
00:50:26 – 00:50:56
and we feed it an X ray image or something and it says, oh, this, you know, this person has a probability of lung cancer with X percent. That’s not good enough. We have to say, well, we think this person has lung cancer with X percent because there’s some patterns in the lower left knowledge of the lung. And they have this history and, and et cetera, et cetera, et cetera. So we’re making this explainability report that allows A I to explain itself just as a human can. And I think we’re gonna get to the point where that’s where I hope we get to the point as humans where that’s the most
Kordel K France
00:50:57 – 00:51:28
uh that’s the most important thing is making sure A I can explain itself because that will allow us to kind of keep the train from running down the tracks without us and uh just put humanism back into A I. Um because right now it’s very smart. But if you, if you keep asking why, why, why, why did you do this? It’s not good at explaining itself and understanding what’s going on inside and to some degree, humans aren’t either, right? Like when we make decisions, being able to thoroughly explain what we do, um we need to be able to do that better, but we can still do it way better than A I.
Jeff Bullas
00:51:29 – 00:52:06
Yeah. And this is where I think A I is a human assistant that’s by our side that we have control over. Um for me, that’s me, I’m just trying to think of how we frame it and how we use it. And it’s really fascinating to actually keep exploring that and keep asking the questions, what guardrails do we need? And you’ve just mentioned putting guard rails in place just for medical reasons because the challenge with A I is, it’s so broad, it touches every part of what it means to be human, right? And you’ve just added another sense to it. Um So
Kordel K France
00:52:06 – 00:52:20
yeah, yeah. So uh in, in order to try to make it more human, we’re uh we’re actually going against the problem, I guess to some degree. But uh um yeah, hopefully an interesting dichotomy there. But uh
Jeff Bullas
00:52:21 – 00:52:48
yeah, but um yeah, no, it’s just and, and Nick Bostrom very much with his dystopian look at it i’m back in his superintelligence book. And what was funny when they reached out to me and went, oh, Nick Bostrom, you sound familiar and I went, oh, I, you know, I did a search of my kindle in other word, in other words, another part of A I extension of my brain kindle searched, found book, read it in 2014. So it’s like,
Jeff Bullas
00:52:49 – 00:53:51
OK, and this happens all the time now, right? So um I forgot something in Google docs. Do a search there. It is OK. Great, fantastic. Um So humans are beautifully imperfect but they have some superpowers. That is, I find it just mind boggling and the creativity, you know, behind humans is just um something and from entrepreneurs to scientists, creativity lies in every corner of being human. Um but guardrails, I think um are really, really important because we didn’t put them in place for social media and it’s now becoming a challenge even to democracy. So, and then it’s being used and applied to autocracy where the government can actually control the people and, and different governments have different philosophies. Chinese philosophy is very different from American philosophy. Um So America’s all about individualism.
Jeff Bullas
00:53:52 – 00:54:02
Chinese philosophy is very much about OK, you work for the good of the state. So that produces a totally different attitude to almost every part of humanity.
Kordel K France
00:54:03 – 00:54:04
That’s true.
Jeff Bullas
00:54:04 – 00:54:36
You know, and so we’ve got to be very careful that we judge the Chinese and the Chinese judge the Americans realize where we are coming from as humans? What’s your philosophy, how do you live and work? So anyway, I, yeah, and right. I um so just wrap it up and we’re having a good time. So corner, where do you think IT’s going, do you think we’re gonna reach singularity in 2050 according to Ray Kwell?
Kordel K France
00:54:37 – 00:55:06
Uh I mean, I, I don’t, I don’t know, I struggle with those predictions just because um A I surprises me everyday and where it goes, uh the things that the things I think are possible uh end up not being possible now. And the things that I thought were 10 years away will happen tomorrow. So, um I mean, there’s definitely that it’s not, it’s in the set of all possible outcomes for sure. It can definitely happen. But I don’t, I’m not sure that we’ll get there by 2050. Uh I do think that
Kordel K France
00:55:07 – 00:55:30
we need to take a more sobering look at a, I, I don’t think we’re at the point now where we’re about to build terminator runaway with the runaway train. But I think to, to really compliment what you’re saying about using social media as a smoke signal for what’s about to come. We should definitely study that and definitely look at other technologies that we can’t go away without right now that we can rid ourselves of right now to try to
Kordel K France
00:55:30 – 00:56:06
put those guard rails in place to make sure that doesn’t happen with A I because once it does get to the level of A G I, whether it’s next year or in 1000 years, we wanna make sure that we have the right framework of thinking so that uh we can control it and we’re not caught flatfooted. So I hope there’s a lot more interest and explainability in making these large models explainable. And uh um I guess mirroring humans to that degree. But I, I also hope that there’s a lot more investment in other aspects of A I besides large language models because there’s a whole other series of industries out there that, uh, need the talent and attention and funding, to be honest. But,
Kordel K France
00:56:07 – 00:56:21
uh, tomorrow is going to, it will never go, like, uh, I think it will with a, I, so I’ll be, I’ll be just as surprised in the news tomorrow as I was today when I read it. So, um, I hope that we can control that course though and paint the future that we want as humans.
Jeff Bullas
00:56:22 – 00:57:03
Yeah. And I think that, uh crossed my mind as I was listening to you, um, is that I’m going and you, you, what you said was something about? Ok. Can you, can there, can the a I explain itself? In other words, how did you come to this decision? In fact, maybe that is just one guard rail we need amongst many. Is that, please explain as, um, in other words, how did you get here? Tell us what evidence you use to do this and it actually has to be built into every prompt almost as, um, you know, a little note that you click and go. Please explain. Tell me the steps, what information you use, what research you use to make it,
Jeff Bullas
00:57:04 – 00:57:12
I suppose to reveal its um soul and you know, where did it get this from? I don’t know your thoughts on something like that.
Kordel K France
00:57:13 – 00:57:43
No, I, I fully agree. It’s funny because have you ever had the serial Lucky Charms? No, I haven’t. No. Oh, well, it’s it’s, it’s a cereal with a bunch of marshmallows in it. And when I was a little kid, I wanted to eat cereal. My mother would give it to me and I would only eat the marshmallows and she would just always get mad and tell me you can’t do that. I’m not giving the cereal anymore. And I always say why. And she said, because I’m the mom, that’s why she’d always say that with everything. I’m a mom, that’s why. And I was like, I want an explanation like I, I fully respect your decision but just explain why. Right?
Kordel K France
00:57:43 – 00:58:08
And right now I feel like I am the same thing like it gives you an answer and you say, well, why? And it says, well, because I’m a I, that’s why don’t worry about it. And you know, we need to take the ego out of A I and have it, make sure it can give the appropriate responses and explanations for its decisions. Uh Otherwise you’re gonna end up with a bunch of frustrated teens. You like I was as humans and we’re just gonna, you know, we’re gonna rebel against the A I mothers because they can’t explain their decisions.
Jeff Bullas
00:58:09 – 00:58:13
I love that analogy. In other words, she’s going to trust me, I’m your mum.
Kordel K France
00:58:14 – 00:58:15
Exactly.
Jeff Bullas
00:58:16 – 00:58:21
Instead of saying because the sugar’s gonna kill you, son. Um OK, so we give you,
Kordel K France
00:58:21 – 00:58:23
We appreciated that. Yeah, just tell me that.
Jeff Bullas
00:58:24 – 00:58:42
And when the A, I tell us, ok, this is a result of going, why? Because I’m a, I, well, fuck off. Right. You’re the machine. Actually, you’ve got to answer to me. Um I, I’m still the oracle and my God. I brought you forth and I can actually take you away as well. So that’s the old kill switch thing, isn’t it? Really? So.
Kordel K France
00:58:43 – 00:58:45
Exactly. Exactly.
Jeff Bullas
00:58:46 – 00:59:14
Kordel. It’s been an absolute pleasure to talk to your buddy Ed String. Really fun. Um We’re well over an hour now and I’m, I’m aware that you might need to go and create more sense processes and uh turn on a robot somewhere. Um This is one question I always ask my guest: if you had all the money in the world, wouldn’t it work again? What would you do every day? That brings you deep joy?
Kordel K France
00:59:21 – 00:59:42
Honestly, I would probably be doing the same thing I’m doing now contributing to what I hope is the safe development of artificial general intelligence. I would just be doing it um, with a much better wardrobe and with a lot more money and probably a lot more progress because I have the money so well, probably the same thing I’m doing now.
Jeff Bullas
00:59:42 – 01:00:07
In fact, you do more of what you’re doing now. Exactly. Yeah. And I, I think for me, um and I go back to Joseph Campbell again because he said he was asked a big question. He was raised as a very strict Catholic. Um And he studied the myths and stories of humanity and religion and culture. And he said, I don’t believe there’s any purpose to us as humans except to follow your bliss.
Jeff Bullas
01:00:09 – 01:00:56
And what that means is following your curiosity that calls you. And that is the intersection of both what you’re good at and what you love doing and how to make a difference in the world. And for me, it resonates a lot because it’s, and following your bliss is not a destination because you’ll never arrive. It’s actually, it’s following your adventure. What is your call to adventure? And for me listening to you today is your call to adventure. It is a combination and intersection of multiple experiences of your life. And for me listening to you and hearing your answer, um You are indeed following your bliss. And I think we need more people to do that because that’s where the magic really happens.
Kordel K France
01:00:57 – 01:01:05
Thank you. I appreciate that. That’s a, I love that question. That’s uh not something I’ve actually thought about before, but uh that’s the honest answer. So I love
Jeff Bullas
01:01:05 – 01:01:27
That’s good. Well, it’s um a question I ask a lot for myself and I think, um, I started following my bliss after I turned 50 because I know I was just following everyone’s path before that. And, um, sometimes it takes, uh, you know, something to happen in your life. That’s not necessarily what you want. My marriage broke up, lost the family home. And, um,
Kordel K France
01:01:27 – 01:01:29
oh, I’m sorry to hear that.
Jeff Bullas
01:01:29 – 01:02:09
Well, it’s life. Right. But what, be careful what you wish for because out of that became the best part of my life and continues to be and I wouldn’t be doing this today and having a chat with you unless that happens. So, um but um, yeah, but thank you for your honest answer. And um, Kordel, I’m looking forward to seeing you know, your journey and how it unfolds. Um And because I think you’re right, I think the sense of smell has been ignored for too long. We need to be fully human and you’re helping to do that by actually asking the questions. How do we make a machine more human because that does raise the big questions, doesn’t it?
Kordel K France
01:02:10 – 01:02:24
It is. Yeah, there’s uh and, and I think those questions need to be asked more and more often as these artificial systems get more and more capable. So if we can contribute to a small part of that, then uh I think I’ve done my job here. So uh hopefully that’s the case.
Jeff Bullas
01:02:24 – 01:02:40
Ok. Well, thank you for following your bliss. Kordel. It’s um but for me that is what it means to be fully human. So, thank you for leaning into your curiosity and it must be almost a compulsion. Now, do you feel slightly obsessed?
Kordel K France
01:02:41 – 01:03:15
Very obsessed. Yeah, it’s, it’s definitely, it’s definitely obsessive. Uh, and I think, yeah, maybe that’s the sign you are following, following your bliss. Right? As if you are obsessive, almost possessive over it. Uh You know, so, um, yeah, I thank you so much for the uh enlightening conversation. I really like the framework that you think about a lot of these things and I think that uh the world needs to hear more of the way you’re thinking about some of these things because it’s gonna foreshadow how we put some of these guard rails in place for more uh ethical and more um intelligent and safe A I systems in the future. So thank you,
Jeff Bullas
01:03:16 – 01:03:43
Thanks Kordel for that. Um That is very much appreciated and um I’ve just loved this conversation and um to have a chat with, you know, such intelligent, creative people. Um I feel grateful to have this honor of doing this all the time. So part of what I keep doing is what I’m doing today. So um basically having chats with really fun, smart people is great. Thank you.
Kordel K France
01:03:44 – 01:03:46
Of course, I love it. Let’s do it again soon.
Jeff Bullas
01:03:46 – 01:04:01
We will look, we’ll keep in touch actually on this and um I wanna see how this unfolds Um, because I, yeah, it’s the full modality of being human. You’re actually adding, you know, some more magic to the bucket of IO I really, so that’s fantastic.
Kordel K France
01:04:03 – 01:04:09
Terrific. Well, I’ll keep you posted and hopefully, uh, next time I’ll have some, uh, some interesting progress report.
Jeff Bullas
01:04:09 – 01:04:10
Thank you.