Introduction - David Pugh-Jones
This edition of AI, Software, & Wetware features an audio interview with David Pugh-Jones. David is the 🇬🇧 UK-based Chief Marketing Officer of Corpora.ai and a founding executive member & trustee of the charity Neurodiversity in Business. We discuss:
How neurodiversity inclusion is helping over a thousand corporations worldwide improve their business fundamentals
Why he believes both “AI won’t be a word that will be mentioned in a couple of years” and “AI can only look back, it can’t look forward”
His thoughts on AI tool hallucinations and seeking the source of truth
Why he sees acceleration and not degradation of human roles and jobs as the more likely outcome for society
How Corpora helps experts become polymaths and solve problems creatively
and more. Check it out, and let us know what you think!
This post is part of our AI6P interview series on “AI, Software, and Wetware”. Our guests share their experiences with using AI, and how they feel about AI using their data and works.
This interview is available as an audio recording (embedded here in the post, and later in our AI6P external podcasts). This post includes the full, human-edited transcript. (If it doesn’t fit in your email client, click HERE to read the whole post online.)
Note: In this article series, “AI” means artificial intelligence and spans classical statistical methods, data analytics, machine learning, generative AI, and other non-generative AI. See this Glossary and “AI Fundamentals #01: What is Artificial Intelligence?” for reference.
Interview - David Pugh-Jones
Karen: I’m delighted to welcome David Pugh-Jones from England as my guest for “AI, Software, and Wetware”. David, thank you so much for joining me for this interview. Please tell us about yourself, who you are, and what you do.
David: Thank you very much for having me. Hopefully there’s some value to be given here and I’m excited, actually. Thank you very much for allowing me to talk to you. I’m happy to give you a little bit of background of where I came from and what I’m up to. My name’s David Pugh-Jones, as Karen kindly mentioned. I am the CMO of a scientific frontier research platform called Corpora.AI, which is spelled C-O-R-P-O-R-A dot AI. I’m very lucky to say that I work for the English tech entrepreneur philanthropist and the ex-chairman of Candy Crush. They do call him the Candy Crush King. So he exited Candy Crush into King and then into Activision for, well, a lot of money. But he was one of the original investors in Candy Crush before people knew what mobile and casual gaming was. So yeah, I have the joy and pleasure of working for a philanthropist in Mel.
And a few other things about me. I am a Gartner Ambassador, so I help with advisory around marketing communities and leaders that are asking around AI technology growth, leadership, hiring. One of my passion points is neurodiversity. I helped to found a charity in the UK, launched in the UK Parliament, called Neurodiversity in Business. And over the course of the last four years, we’ve accumulated nearly a thousand global corporates that utilize our research to help nurture, empower, and educate companies from top down and bottom up to literally look after people, and look after talent. So yeah, that’s me.
Karen: That sounds great. And that’s an impressive initiative to get people actually taking neurodiversity seriously and understanding that it can bring a lot of benefits to have neurodiverse teams. I remember hearing in my last corporate role that people who are neurodivergent tend to be more represented in working in AI than in other areas, and I thought that was kind of interesting.
David: Yeah, the research is incredible and interesting enough, you can access the last two years of research for free on our site, which is neurodiversity in business.org. A lot of it is obviously taken from IT with a UK lens, but we are actually looking at global companies. And there’s companies like IBM and McDonald’s and JP Morgan and some incredible businesses that are now utilizing that to help with their staff. But the interesting fact is that if businesses actually do hire and nurture and empower neurodivergent individuals alongside neurotypicals, then the fundamentals kick in. You know, more opportunity, more creative prowess, more growth, more strategy, just everything seems to signal an up-weighting of all of those scenarios. So there is a value-add for having a very diverse workforce. And yeah, we just need to make sure that we’re empowering individuals to be the best version of themselves. And I do say that quite a lot.
Karen: That sounds great. So tell us a little bit about your level of experience with AI and machine learning and analytics, and how you’ve used it professionally or personally, or whether you’ve studied the technologies.
David: Yeah, so look, as a backdrop, I’ve been in the tech space for 30 years now. I remember in the late nineties when I was working in publishing and I got a call from a friend of a friend who asked me to go into this internet world, and effectively moving to advertising content and brand strategy. One, I was excited about that whole realm, that next phase of what the beginnings of the internet was. But on the other side of the coin is that I had to teach myself to code, to get up to speed on some of the things, because nobody else knew what to do. And so I was sort of thrown to the lions, so to speak, and thought, you know, you either sink or swim in that scenario. So to frame your question, I’ve seen every era of technology and then how it’s been embraced by consumers, and then how businesses have adopted it, and then all of a sudden, it becomes mainstream and then it becomes part of the vernacular.
Look, AI won’t be a word that will be mentioned in a couple of years. My prediction is that it will just be part of the fabric of society, right? Everything that we do and touch, it’ll be embedded into it. I use it on a daily basis. I use it with work. I use it with my life. I use it in scenarios from a content strategy point of view. And then I also help it to make me more efficient as an individual.
But I also recognize when not to use it. And that’s something that’s quite important is there are times and places where I don’t need it. I don’t need to ask AI what I’m going to have for breakfast tomorrow morning. I think I can make my own mind up on that one. But yeah, look from a business point of view, and let’s give you a great example. If you’re going on a holiday to a country you’ve never been before, the internet and search as we know it is becoming more and more fragmented, and it’s harder to find websites and advisory individuals. And unless you’re talking to friends and communities, to track down and have an agenda or an itinerary to go somewhere on holiday, that’s a great use of the technology. So it’s efficiency gains, productivity gains – about an opportunity to embrace it without letting it take over your life.
And it’s the same vein as what the internet was when it was a phenomenon. The same vein as when we all had a mobile phone in our hands and we realized we didn’t need to try and find a restaurant with a phone in it to phone home, or to call for an emergency. It’s another transition in the realms of humanity, should we call it that?
Karen: That’s fair. Yeah. A lot of people tend to think that AI is new. They conflate it with generative AI and the chatbots that started emerging in late 2022. I think a lot of people don’t realize that AI has already permeated our lives. It is everywhere. It is in our movie recommenders, it’s in our cell phones, it’s in the image recognitions and just everything. I’ve written a few articles about that. The analogy I use is an iceberg. A lot of it is under the waterline. It is there, but we don’t normally see it, but it’s affecting our lives. It can be in good ways, and in some cases, it’s not as good. It’s a mixture, like every technology.
David: Exactly. And look, there are scenarios where it just won’t do. And we’ll get into the weeds on what it’s brilliant at. But equally, there are lots of versions that are coming out and there are some of the biggest issues around sustainability and the demand and the exponential growth. Again, I think we as humans need to work out exactly how it benefits us, and where it doesn’t. And then we just need to make sure that we do the right thing in that sense.
But, yeah, you’re right. Look, three years ago, the world changed significantly when ChatGPT became this phenomenon that everyone had access to. Now AI’s been around for more than half a century. Those scientists and those developers and those individuals were playing around with code long before we even knew what was going to happen. It just happened to become consumerized and commercialized. And now it’s out there and we’ve all got our dirty hands on it, and we are trying to work out what to use and how to use it. It’s a strange place to be in right now because of what’s going on in the world. But we’ve just got to hope that humanity does the right thing in the sense of how we best utilize it, is my best advice.
Karen: So how did you go about learning about these technologies? Self-study, or what have you done that worked for you?
David: Yeah, look, my friends will say I talk too much and lots of them will say, I love the sound of my own voice – which is not true, by the way. When we hear this back, I’ll be cringing dreadfully! But I would say, I am very inquisitive, and I wouldn’t say I’m a geeky or a super tech individual, but I want to know how it’s going to help me individually and how it’s going to help anyone and everyone around me.
So that combo is that I talk to people that know more than I do. And then I take in fragments of that information. And then I go out and I do what I do: test and learn. Okay? I’m constantly looking for what it does, how it works, where can I find it? How easy is it to use? And then I drop off or push away all of those things that I can’t see it being, from a utility point of view, valuable to me. So I just jump in. I just think, what have you got to lose, right? We as individuals want to learn. Most of us are nosy, I would say, we want to learn more. And if there is a technology out there that helps us accelerate our knowledge and we learn more and we have some passion points about subject matters, then clearly just embrace it.
So it is definitely a combo. Yeah. Friends. Get involved. And I do a lot of listening and I go to a lot of events that are very much tech- and AI-focused. And then I look at it from, “What are the governments doing around the world with this? How are businesses and companies utilizing it? How does it help me, my day-to-day?” What can he give back to me from a time point of view, versus do I just need to say “Sorry, I don’t need it now?”
Karen: That’s a good overview. Thank you. Can you share a specific story with us on a way that you’ve used an AI or ML tool, those different features? I’d like to hear your thoughts about how the features worked for you, what worked well, and then what didn’t work so well.
David: Yeah. My biggest issue is this word called hallucinations. The fact is that a lot of these LLMs or large language models that are out there, are trained on data, okay? Some of them are trained on just the western world’s data. Who knows what China Deepseek’s trained on? It’s probably trained on scraping another LLM’s data. Who knows, right? But what we do know because of the hallucinations is, we’ve got a scenario where we’ve got to try and understand what is real and what isn’t. And that is a concern. Okay.
So for me, I’m always trying to find the source of the truth. What is it that is real versus something that isn’t? And currently these models are being trained on all this data that’s going in. But of course, if you start to publish content that has been outputted from an AI platform, and then you are passing that back into it, or someone else does that on your behalf and it continues to train on it. This regurgitation is just a constant flow of fallacy. Then this is where we get in the world of AI slop, okay, where nothing becomes real. It just becomes a perpetual cycle of just dirge and an entire rubbish. So the way I use it is, even with the technology and where I work, I want to see real substance, real humans that wrote stuff. So I’m always trying to get down into the weeds on verifying the context and the information that is in there.
Now, individuals can, if they find one that they’re comfortable with or one that they prefer to utilize from a user experience and a utility, they can get comfortable by training it themselves, so it starts to understand more about you, yet without giving it personal or private data or detail that you do not want to give away.
And you certainly don’t want that LLM to be passing that information to anyone else beyond the realms of what that LLM is. And so I use it sometimes for food, for travel, for understanding more about a business, about working out how I should approach a certain scenario. And look, I work in marketing and comms and PR and strategy, and I’m always trying to work out, is there a better way to tell a story? If AI can help me do that, of course I’m going to take it.
But my theory, and this is probably the fundamental piece of advice that I would give to anyone, is that AI can only look back, it can’t look forward. It’s trained on the data that is historical, okay? It can’t make predictions about when we are going to meet Martians on Mars, okay? Or when half the planet is going to leave on spaceships and live on another planet. Anyway, it can’t foresee that. So it’s very good on giving us context on what happened previously. It’s only as good as the data that you put in it. So slop in is slop out. Utilize it sensibly, wisely and ensure that you try and do your own research as to verifying the quality of the output or the context it gives you.
Karen: Is there any specific example that comes to mind of a time where it gave you great results that you were able to use, and another where it gave you garbage?
David: Yeah, we are very fortunate with our technologies. We’re looking at open-source intelligence, let’s call it, or the entirety of what the web exists. And I’ve seen many examples where I’ve asked it questions related to an event or something I was going to, or an individual. And then the context of that individual wasn’t necessarily true. So it’s taking information from all different sources, whether that’s true or not. So again, it goes back to doing your own research, right? If I hadn’t explored other sources to look at that, then I would’ve had a slightly different perception of those individuals that I was meeting from a business point of view. I won’t go into the details, but yeah. A lot of it was baloney, should we call it?
Karen: It’s funny, one of my earlier guests had called to my attention a series that I somehow missed called Person of Interest. I don’t know if you have heard of that series?
David: No.
Karen: It’s a neat show. It was, I think from 2011 to 15 or so, and somehow I completely missed it when it first came out. But the whole premise is that there is this machine which is this super-intelligent surveillance. It gets data from people everywhere, all over the world, and it’s always looking for threats. And supposedly it was designed to detect terrorist threats, but it also detects threats of violence to ordinary people. And so the people who built the machine and handed it over to the government are getting this information. The government doesn’t care about if this person’s about to get murdered, but they’re going to try to stop it. They did an episode where the person whose number came up was someone who ran this company which is basically a data broker. And they scraped all this data and they used it to provide information about people, and sometimes it was wrong. The episode was about basically people taking revenge on this founder and exposing his data and destroying business deals that were on the brink because of that data and exposure. And this was more than 10 years ago, and it’s still highly relevant. It’s so interesting to watch it now in the framework of what’s going on today.
David: Yeah. Look, at some point, okay, these models are going to run out of data. So let’s just imagine that the entirety of whatever’s available on the internet as we speak, is there. And these models are trying to get this information in there. Now, technically, those models have got a number of deals with publishers and content producers and individuals, and they’re obviously then training that. You saw Harvard releasing research and content and books that had never been published. They were then giving those to one specific model to then start training that one.
So when they run out – I think the stat I saw recently was between 400 and 500,000 pieces of content created every second, okay? Now, what I don’t know is the stats on how much of that had been reproduced with the help of AI. So let’s just imagine all of that is reproduced. It’s then published, and then as soon as it’s published, it’s then adapted and pulled into those models. And then they start to continue to train. But they’re only training on the information. They’re not then asking those publishers, “Hey, could you verify this and tell me that your source is true, or this actually existed?”
So again, this goes back to that ‘humans versus machines’ moment, right? I always say, brands and AI should talk to people the way that people talk to people. Because all of a sudden at some point, we are going to have this assistant, right? Assistant is no longer Siri on our mobile phone, but it’s an AI version of Siri, or it’s an AI in our earpiece that literally, we’re asking it everything and we are trusting the data and the information that’s coming into this ear, translating that in our own brains, and then we’re outputting that. This comes back to truth and trust and verification, all of these things. So that untraceable synthesis is very dangerous. And that’s a volatile place to be in. So again, it’s, this comes back down to us being the source of truth.
Karen: You mentioned that you used AI tools for a lot of different things in your life. Mostly it feels like we’ve been talking about words, but in marketing, obviously imaging is important. There could be other aspects like music and video. I’m curious, which of the tools have you tried for which purposes, or are there certain purposes that you avoid? Like for instance, have you tried to generate any music with the tools?
David: I’ve played around to try and do all manner of stuff with pretty much most of them, out of fascination. And each of them has a specific skillset, shall we say. One is better for research and one is better for content creation. But again, I think it comes down to you as an individual. So look, dabbling is a good place to start, okay. But just don’t feed it your life story and then hope it’s going to be able to help you navigate everything that goes forward from there onwards. Each one of them has a certain skillset that it does and it delivers. And they do become quite personal, because the more you put in, the more you feel like it’s more nuanced to you as an individual and the way that you talk.
And look, if you look at the lists of jobs of the future, and we don’t have to get into the weeds in this, but this is fascinating. All of a sudden, they’re talking about the degradation of human roles and jobs. But in some cases, actually, I think it’s probably the opposite, and it will accelerate them. Because there’s always going to be that human AI scenario. I don’t think AI is going to have a conversation with you, Karen, like I’m having a conversation with you right now. I just can’t see that. I think we’ve got to hold onto what’s intrinsically true is that, us talking about this and using it as an extension, and as a facilitator or something, that actually is a little bit more beneficial.
So pick the one that you think is the most friendly and intuitive towards what you put into it. And just go tentatively. I would say for starters, explore what it does and use it for a couple of things. And also, get into the communities, because there’s some fantastic communities out there on all manner of platforms, including the likes of Reddit, where you can actually see where people are getting true benefit from it. And yeah, don’t forget to talk to your friends about it either.
Karen: Yeah, the human element is important. I’ve had some guests make a distinction between the idea of humans being in the loop and humans being the loop with AI on the loop with them, more a sense of who’s in charge. And it’s important to keep your wetware, your brain, engaged when you’re working with these tools.
David: Definitely, definitely. When outcomes matter, provenance matters. And asking AI about provenance, you’ve got to get into the sources. You’ve got to get into the information on this. And yeah, I’m a firm believer, if we are going to accelerate R&D, we are going to find new drugs to solve cancers, and we are going to have students that are utilizing this to understand and find more information about a subject matter or use it as an assistant. But it always goes back to you and human provenance.
Karen: Some of the tools have claimed that they’re making strides in provenance towards never giving a hallucinated reference. Perplexity and Claude had both talked about that. You still have to verify regardless. I’m always curious to hear what people’s experiences have been with the tools and if some are working better for them and others in that regard.
David: Exactly. And until they come up with a strap line saying, this is 100% no hallucinations, then just take that with a pinch of salt. But the exponential growth of consumers and businesses adopting AI, it’s just a question of how they adopt it and what they use it for.
Karen: Are there specific scenarios where you would just not even consider using generative AI tools?
David: Yeah, it’s a tricky one. In some cases, if I’m organizing an event and I’m talking to the individuals that are going to be on a panel, and they have their own discipline and skillsets, I’m certainly not going to be utilizing AI to understand where the nuances on what they talk about, or what’s going to ignite their passion, and where are they going to light up about something that literally they’re obsessed about and they want to talk about and they want to share that. I’m never going to find that with AI. We’ve still got to be investigators in what we get out of other people.
And in some cases, if you think about a conversation in a pub or a bar or a club, or a gym, or even by the water, you’re making coffee in a kitchen in a work scenario, you’re certainly not going to say, “Hey, before you just tell me what you’re doing this weekend, do you mind if I ask AI to listen in so that we can then decide where else we’re going to go?” That’s just not going to happen.
At some point someone is going to turn around and go, “Yeah, maybe I do need AI to listen in to this.” But I’d be like, “If you can’t tell me yourself what you’re planning to do and you need AI to do it, then humanity’s going in reverse mode and we might as well just give up now.”
Karen: Yeah, so we’ve talked a little bit about these systems and where they get their data and their content and using data from online systems. And companies are not always very transparent about where they’ve gotten the data that they use. I’m wondering how you feel about companies that use people’s data and their works for training these AI and ML systems and tools.
There’s a concept called the 3Cs, which the Cultural Intellectual Property Rights Initiative has been espousing, that people deserve three things. They deserve to have the opportunity to consent. They deserve to be credited if their work is used. And they deserve to be compensated. And I’m wondering what your thoughts are about that. Whether it should be that way, and then whether it actually is that way.
David: I’ve been talking about this. I’d probably go back 20 years in talking about this. I remember in my heady days of being within the realms of Microsoft. I spent nearly a decade there and I was working in lots of different divisions, but always about innovations, technology and storytelling. And I always remember saying, we have an incredible amount of data. So you think about how much data we give on social media, every time we post, every time we comment, every time we share. And it’s no different in the realms of AI, right?
So if you extract all this, all the social media platforms that you live on, and then all of a sudden, you start sharing all of that information with AI. Then all of a sudden, they’re talking about some point, whether it’s $20 a month or if you want the faster version, it’s $200 a month. And at some point, it’s going to be even more than that. The business services are getting faster access to more information that opens up more data silos. We are still giving away our own data.
And I truly believe that at some point we have to say, “Do you know what? Isn’t our data worth something?” Isn’t it about time we flipped the coin and said, “Hey, you know so much about me, and right now, this is beyond the cookie and you sending me and sharing crappy ads with me that I was never even thinking about buying that particular product. But my data is quite valuable and I actually want to start getting paid for it.” And that means that if you want to send me stuff, yes, actually, welcome to do that, but start paying me intrinsically. So that platform gets paid by those brands and you personally get paid a percentage of that for them sharing that content or story with you beyond the subscription model.
Because look, consent is technically present, but it’s practically inaccessible in that sense. It’s buried in the terms and conditions about opt-out. So nobody tries to find the button that says “Do not train the model based on what I tell you”. It’s not the default in any of those models. So maybe we need to be a bit more tech-savvy or maybe we need to make sure that when you do use these platforms, what you do is that you just ensure that whatever information is only between you and them. Again, look, there’s been a scenario recently in the US with Anthropic and ChatGPT about who accesses that information, and that’s quite worrying, I would say. It’s quite concerning if using those platforms and your information could potentially be leaked or given away, certainly without your consent. That’s not meaningful, right?
So we do need much clearer disclosure. We need simpler control. We need AI infrastructure that respects jurisdictional sovereignty. And that’s interesting because if it starts traveling from one country to another, that’s the systems and the technologies, assuming entitlement to data, just because you’ve given access to that. But that accessibility is not consent, as far as I’m concerned. So there’s something to be concerned about and I think in the next 18 months we’re going to see this come to light as they start to put in more T’s and C’s, and more information about how your data’s being utilized.
Karen: Yeah, you’ve raised some really good points about informed consent and about the asymmetry in the power and whether people really have true consent. One of the lawyers I interviewed, she’s based in Ireland, and she put it this way: those terms are ridiculous for consent. It’s just not even practical because they have phrases like, “We can use this to improve our products.” Well, that’s wide open.
David: Yeah.
Karen: That’s big enough to drive a semi-truck through. And it doesn’t really tell people anything about what they’re really opting into or not. In some cases we don’t even have a choice. It’s a government system. You want a driver’s license or something, you have to give your data, and it’s not in many cases a choice to not have a driver’s license. So we don’t often have choices in some of these cases. But even in cases where we are given a choice, there are the foundation models, and then there are systems like Meta or LinkedIn or tools that people use very commonly. And how much choice do we really have, and do we really know what they’re using with our data? Both had some scandals in previous years.
David: Yeah. I’m not shocked, but I am kind of blown away by how normalized it has become for that data to be utilized in all manner of ways. And the other thing with more data, and this is probably the unspoken thing, is that sustainability seems to have fallen off a cliff. So nobody’s talking about how many lakes are drying up or how many data centers need to be built? And is Greenland required because it’s cold? Is it required because they can build this infrastructure that’s going to contain all of this data? You know, at some point, the demand is going to outstrip supply and all of a sudden, you’ve got cost hemorrhaging scenarios where access is limited. And then does it become limited in the capacity that is given to consumers. So there’s lots of fundamental questions that still go unanswered about the evolution of scale because if we are still in the adoption rate as we were with previous technologies that changed the world forever is. Yes, it’s great that it’s intrinsic and it’s embedded into the fabric of society, but only if it’s built in to do the right thing and actually make us as a world and a planet better at achieving things or solving scenarios that are for the greater good.
And when I ever say “greater good”, it always makes me laugh about a movie called Hot Fuzz. I don’t know if you’ve ever seen Hot Fuzz, where this cult always talks about the greater good. And as soon as someone says it, everyone says “the greater good”. Maybe my context is in a better situation than the cult from Hot Fuzz.
Karen: Oh, I’m going to have to look up that movie now! But yeah, sustainability is a big thing. I wrote about this a little bit in my book Everyday Ethical AI. The whole sustainability aspect is one of the five areas that I called out. But we’re hearing people here that are protesting data centers. They want to build it in these areas where it’s already suffering from droughts that are impacting agriculture, and they want to put a data center in there. And some people are protesting and fighting back, and in some cases, they are succeeding. In other cases, they just run them over, and just go ahead and do it. So there are people that are resisting.
The other thing that has come up more recently is talking about putting data centers in space, the regulations there about what you can do ,and whether you should do it, and what impact it’s going to have. I have a friend who’s a founder in the space situational awareness area. And the risks of satellites colliding and taking down our infrastructure is a whole ‘nother side path. Putting more data centers in space has an appeal, but it also could be very risky from other perspectives.
David: Yeah. There are some good stories out there and I’ll give you a great example of it. We are working with some scientists in the company, in little old England over here. And all of a sudden, our research that, you know, is AI-enabled, but we only use it to summarize, not hallucinate, right, because we already sit on that data. What we are doing is giving researchers and scientists the capability to look beyond their field of expertise. And they become polymaths. There was a piece around a Russian scientist in the forties that worked out that things could be solved by looking at another scenario where they’d solved it. So it was a bit about efficiencies and changing the mode of where you experience that.
And in one of those cases, with our research given to them, they were already going down the route of effectively ceramic-coating solar panels. And so what they’ve realized is by coating these solar panels, they’re getting more energy output and efficiencies from those solar panels to a point of two and a half to one. So all of a sudden, that research and that AI is doing some brilliance, because it’s reducing the demand for solar, and solar is creating more energy coming back. Proud, delighted, wondrous scenario, an example of putting AI-powered research into the hands of other scientists that are looking into all manner of scenarios to solve.
Or in clinical trials, or they’re doing research studies in biotech, or they’re working out drug discovery that they would never have done, in the same world that Viagra was found accidentally. These are the things that we should be focusing on, right? How can it help us as a society and as humanity, to become more efficient or actually to save lives. And that’s exactly the good bit. The data storage bit is concerning.
Karen: You mentioned the Russian scientist. I had some courses while I was still working in corporate in a method called TRIZ for Creative Problem Solving. And it sounds very similar, the idea of looking at a different domain and seeing how they solved a problem and then applying it to your domain. I always thought that was very interesting.
David: Absolutely. Yeah, absolutely. Look, sometimes it’s the left-field ideas, right? Approaching something with a completely different tack, and having others then look at that. One of the issues, what we are seeing now, is that we are saving time. We did some research, around a thousand businesses and researchers and scientists in America, and we saw the efficiencies of how they were saving time in the US, and then we did it with another 500 researchers in the UK. The adoption was interesting. But the trust in the output was only at 53%. They’re still utilizing the technology, but they’re not actually going down that route of verifying and ensuring that the content that they’re receiving is cited or comes from a source that they can trust. I think there’s a big education piece here, and for us as Joe Blogg‘s consumer, we just need to work out how best to use it.
Karen: You mentioned trust and transparency aspects a few times. I’m curious, in general, what you think is important for the tools that you use. And then I’d like to hear about what Corpora has done in that regard to help establish that kind of trust with your customers.
David: Yeah. Well, the way that we’ve approached it is very different. We are not an LLM. We’re talking about scientific frontier research, right? We’re talking about discovery. We’re talking about innovation. We’re talking about using it for the greater good. There you go. I said it again. We are looking to give it into the hands of individuals that are trying to solve something. We call the body of text the corpus, right? We’re trying to be the modern day library of Alexandria, but rather than it be limited to the elite, that only they could stroll through the avenues and look at the scrolls, effectively we are saying even with our accelerator program, is you put this in the hands of academics or students that are in that scenario – that are being tutored by professors, but they have a brilliant idea and yet they don’t know how to bubble it up or surface it. And they need research and scientific discovery to solve those scenarios. That’s the bit that I think is even more exciting. All of a sudden, we’re empowering the next generation to do and create and solve better things.
Karen: It sounds like you’ve taken steps towards addressing both transparency and perhaps explainability. When I worked with teaming up with domain experts and machine learning people to develop projects, those were usually getting the best results. But there was always this dimension of trust in terms of, where did the results come from? How did you build the model? What assumptions did you make? The experts want explainability. How do I know that this model you built for predicting when this transformer’s going to fail is better than my 30 years of experience in servicing them and maintaining them and understanding how they fail? There’s always that dimension to it. And especially when you get into the more technical fields, I would imagine the experts who would be looking at using your tools would raise some of the same concerns.
David: Yeah. And look, it’s access to information that is going to help accelerate discovery. And that could be in defense right now. That couldn’t be more prevalent if we tried. If you’re looking about drone technology and about protecting people, healthcare, we need all the help we can get in discovering new ways to cure things that are not great for us, as humanity. And then, the whole research and discovery piece is about us continuing to innovate, finding and understanding new things. And if technology is a good enabler for that, it still comes down to the quality of the traceability.
One piece that we are creating is this element of ideation to get people to think about approaching things in a different way. And that doesn’t mean making up scenarios. It means looking at the data, looking backwards to look forward. What is the quality of the information, and is there other ways to approach it, and what else has anyone else done?
And trying to compile that, based on where everything exists in the entirety of the web is complex, challenging, difficult, time consuming. If we can make it more efficient for individuals to have access to that and then also the capability. One of the programs we love is that we’ve got our professors in residence where we bring in individuals that are experts in their field. We’ve got a professor of AI and we’ve got a professor in medical research and one in aerospace and one in material science, and we bring all of those four together in a room, trust me. One, it is absolutely fascinating, and I am that fly on the wall just sitting there listening with my jaw slack as these individuals open up their minds to explore something in a very different manner. So maybe we should do that, right? We’re always looking for the norm or similarity or familiarity. Maybe we should actually sometimes just go out of our periphery and explore meeting and understanding other people that look at the world or approach things in a different manner. And it may open our minds too.
Karen: That’s a great perspective. I have a final question for you about public distrust of AI and tech companies. There’s a phrase that was noted in your document about how “Humanity and AI is a duet and not a duel”.
David: Yeah, look, I’m a firm believer that it is a facilitator and it helps incredibly, but I still believe in us, right? We’ve still got a chance here, right? It’s not over yet. AGI and autonomy and the world taking over everything, and then all of a sudden, we’re just sitting there eating chips, watching TV whilst AI just does everything for us? Come on, right? You know, we weren’t created for that reason and that reason alone. We look at it as, there’s provenance, right? Where do things originate from? There’s explaining it, right? And sometimes you can’t explain it, but you can understand the reasoning of why something was created. And then all of a sudden, there’s the sovereign boundaries of how actually does that data travel and that information travel. And that was the beauty of the internet, is it opened up our minds to seeing things that some people would never get to see. What’s exciting is that, with the realms of digital currency and stablecoins and mobile phones and satellites, is that kids in impoverished scenarios can learn in days or weeks. And they certainly don’t have to go to university to get a degree when the opportunity to learn, a rate of pace, is just incredibly transformational. So we’re certainly not entering an era where AI is going to replace us or the experts. But I think it is an opportunity where experts learn how to utilize AI for the betterment of the world.
Karen: That’s a great summary, so thank you for sharing that and for recapping it so concisely. I appreciate you making time for this interview. Is there anything else that you would like to share with our audience about AI, about using these technologies for good? A lot of the purposes you described certainly seem to fit in that area.
David: Yeah, exactly. I would say look, embrace them. Don’t fear them. Have fun, okay? Maybe start by having fun, and if you are already embedded into utilizing them, then see where it can help you, and to give you something back in life that you want to do, right? This isn’t “You must jump in and everyone must use it all of the time”. It becomes another facilitator, just as the mobile phone did, or just as the internet did, or all of these things are. There are efficiency gains, but no doubt, just like all the technologies that came before, there is the hype, okay? And there’s the excitement and there’s the frustration and there’s the nervousness around these. But I would say with all of the doom mongering that came about it invariably is, we did the right thing in the end, or we continue to do the right thing. There’s lots of bad things, by the way, you know. Social media, let’s not even go there. But there are lots of things that this has helped us to gain an edge. So, embrace it and have a little fun and play. Joy.
Karen: Yeah. You mentioned earlier about how the AI revolution is comparable to some of the other technology revolutions that you’ve seen in your earlier experiences. One comment I read recently got me thinking that, with the internet, and certainly with mobile devices, part of the hype was, “Oh, this is going to save you so much time! It will free you up from all this.” And that actually hasn’t happened in a lot of cases. Actually, the time that it’s freed up is quickly consumed by doing more things with that device. And so we don’t really end up having more free time. I’m hearing some people expressing concern that the same thing is going to happen with AI. It’s like, yes, it makes you more productive, you get more time at work, but you end up spending that time doing other things and so you’re not really getting any time back. And I’m wondering what your thoughts are about that from the historical perspective and what’s now.
David: Look, I have an addictive personality, so I get stuck into something quite quickly. I’ve even caught myself doom scrolling and thinking “How many miles of screen time have I flicked through with my thumb over the last 30 minutes whilst I was on a train journey?” Thinking “What an absolute waste of time where I could be focusing on something else.” And I suppose the other scenario is that if you are researching something or you are looking into learning more, then that’s a great opportunity to utilize it to understand and learn something about something that you’d never known before. Something that you are fascinated by anyway. But clearly there needs to be a brake pedal. Sometimes this is when you realize it becomes more of a hindrance and it’s actually alleviating you becoming better at something. Not everyone can do that, right?
But if these models are here, you can layer in elements of where you want it to give you information and be very specific, right? It’s the world of robots, right? At the moment we are telling them what we want and they’ll then give us a response. So if you only want to learn for 15 minutes a day on one piece of subject matter, people haven’t lost control just yet. Focus on staying in control is what I would say.
Karen: Yeah. You mentioned the addictiveness. A lot of the tools, they design them to be addictive, and they want to keep people on the site, keep their attention, keep them using the tool and providing more data as part of the use of their tool. So I know it can be a struggle for a lot of people.
David: Yeah, look, everyone should be in exploratory mode right now. You are in new discovery. It’s new technology, it’s exciting. It’s intriguing. Who do we put it in the hands of and what are they doing with our data? All of these questions will continue to be unanswered. And there’s some bad people out there that are utilizing it for the wrong reasons. But at least the majority of the adoption, if we can just work out how it can give us more value individually and then collectively, that’s the way to approach it right now. Because we are not going to solve this overnight. There’s lots of pieces of the puzzle that were still to be worked out, but whilst it’s in our hands, work out how it’s going to help you.
Karen: You’ve mentioned the different areas where your company is able to support really good purposes. Finding new drugs and improving energy breakthroughs, like with your solar panels. I worked in solar for a while, so I thought that was interesting. If you can share some of those links, we’ll put those into the article. [see footnotes 1]
David: We did a piece around innovators that were trying to patent their own technology or their own ideas. China and lots of technology companies in the US are not sat in patent offices just working out what’s coming through the machine. But if there’s a way for innovators to help them become more efficient, or even find other people that can help them bring something to life, because not everyone is a business person, right? Not everyone is a technologist. Not everyone is a storyteller. So if you can bring those individuals together to your idea, and that’s where I think AI could possibly help, is it collectively brings the best of all of these people to then accelerate something. If we discover scenarios faster because of the technology, then those are the things that we should get excited about. And we should make sure that we shout from the rooftops about where it’s enabled people to solve scenarios or bring something to the market that is making it better for us. Those are the bits that excite us and that’s what we’re doing now, finding players and partners that want to understand something faster than they ever could have done before.
Karen: Great summary and appreciate you making the time for the interview, David. I enjoyed our conversation and I’m looking forward to seeing what you do next.
David: Not at all. Thank you very much for having me. And thanks to your listeners. So, yeah, if they want to reach out, feel free: throw me questions or reach out to me on LinkedIn. I’m always interested in hearing from new people.
Interview References and Links
David Pugh-Jones (DPJ) | LinkedIn
David Pugh-Jones (@davidpj) / X
Corpora.AI
Neurodiversity In Business (UK)
About this interview series and newsletter
This post is part of our AI6P interview series on “AI, Software, and Wetware”. It showcases how real people around the world are using their wetware (brains and human intelligence) with AI-based software tools, or are being affected by AI.
And we’re all being affected by AI nowadays in our daily lives, perhaps more than we realize. For some examples, see post “But I Don’t Use AI”:
We want to hear from a diverse pool of people worldwide in a variety of roles. (No technical experience with AI is required.) If you’re interested in being a featured interview guest, anonymous or with credit, please check out our guest FAQ and get in touch!
6 'P's in AI Pods (AI6P) is a 100% human-authored, 100% reader-supported publication. (No ads, no affiliate links, no paywalls on new posts). All new posts are FREE to read and listen to. To automatically receive new AI6P posts and support our work, consider becoming a subscriber:
Series Credits and References
Disclaimer: This content is for informational purposes only and does not and should not be considered professional advice. Information is believed to be current at the time of publication but may become outdated. Please verify details before relying on it. All works, downloads, and services provided through 6 'P's in AI Pods (AI6P) publication are subject to the Publisher Terms available here. By using this content you agree to the Publisher Terms.
Audio Sound Effect from Pixabay
Microphone photo by Michal Czyz on Unsplash (contact Michal Czyz on LinkedIn)
Credit to CIPRI (Cultural Intellectual Property Rights Initiative®) for their “3Cs' Rule: Consent. Credit. Compensation©.”
Credit to Beth Spencer for the “Created With Human Intelligence” badge we use to reflect our commitment that content in these interviews will be human-created:
If you enjoyed this interview, my guest and I would love to have your support via a heart, share, restack, or Note! (One-time tips or voluntary donations via paid subscription are always welcome and appreciated, too 😊)
Our Frontier Accelerator program is designed to support researchers, founders, and spinouts emerging from academic institutions who are pushing into new scientific and technological frontiers but often lack the infrastructure, support, and research to fully realise their ideas. As a Scientific Frontier Research Platform, we combine open source intelligence, deep scientific insight, and access to funding pathways to help accelerate these initiatives from concept through to commercial and societal impact More info is here: https://corpora.ai/accelerator
The iTrend program, funded by the charity MSTART and supported by Corpora’s Scientific Frontier Research technology, is a collaborative initiative involving the University of Derby, the University of Nottingham, and the Royal Derby Hospital. Its primary aim is to improve long-term outcomes for patients with end-stage renal failure undergoing dialysis treatment. The program includes a synthetic human cardiovascular system designed to monitor blood pressure: Case study: iTrend: Intelligent technologies for renal dialysis - Engineering Professors Council
