3169: From Candy Crush to AI Research: Mel Morris on His Latest Innovation

[00:00:04] What happens when you take one of the visionary minds behind global gaming phenomenon Candy Crush and apply it to revolutionising research with cutting-edge AI? Well, today I'm going to be joined by Mel Morris, the renowned tech innovator, entrepreneur and now the founder of Corpore, which is an AI-powered research engine. Now, earlier this year many people were asking me what the big tech trends of 2025 were.

[00:00:30] And I've wrote a few articles on this subject and for me it's all about search. I no longer get the information that I need from Google. It no longer gets me the information that I need as a journalist. And we have seen the rise of new services such as perplexity out there that make life a little bit easier. And even the Google CEO recently said search will change profoundly in 2025. So I think there's big things happening in search this year.

[00:00:59] And that's one of the many reasons I invited Mel onto the podcast today. And while many will know Mel for his success in gaming and his time as the Derby County owner, something very close to my heart as a lifelong Rams fan. Today we're going to be diving into the transformative power of technology.

[00:01:17] So in this episode, we're going to explore how Corpore is tackling the growing complexities of modern research by delivering unparalleled depth, clarity and speed. With the ability to process millions of documents per seconds and immediately distill insights with full source attribution. I want to learn about how this new platform could disrupt fields from academia to medicine to journalism and even law.

[00:01:47] So on today's podcast, Mel is going to share his vision for the future and the digital reimagination of the Library of Alexandria by offering real time actionable knowledge in an era of information overload. So how will this game changing technology impact the way that we work, the way we learn and innovate? And what inspired Mel to take on this ambitious new challenge?

[00:02:16] Well, let's get him on the podcast now and find out more. So thank you so much for joining me on the podcast today, Mel. Can you tell everyone listening a little about who you are and what you do? Well, I'm into my now the sixth decade in technology. I started literally before I left high school at John Port. They actually had a computer donated to them there. And they're sort of playing around with that as a part time thing at a very tender age. So six decades now coming into technology.

[00:02:44] Most of that time has been actually as an entrepreneur creating and building tech businesses. There's a litany of those and some really big successes. Prevex, Udate, WebRoot. All these are all companies that I've been intensely involved in. And of course, King with Candy Crush was another one which I provided the seed funding for and chaired the company from inception right the way through to it going public in 2014. And I'm glad you mentioned Candy Crush there.

[00:03:12] Because I suspect for many people listening all around the world, you'll be forever known as that web entrepreneur who gained a large part of your fortune through that successful startup investment in King, the firm behind Candy Crush. But can you tell me a bit more about your origin story? What ignited your passion for technology and would put you on this road? Well, it started really, I suppose, as a sort of, I started working in a firm in Burton-on-Trent, Roe-Burtshire, my firm, the pork pie people.

[00:03:39] And I wanted to get into technology and I was actually doing a decolating job, which was really boring. And one of the ladies there said to me, you look pretty miserable. I said, well, I'm really looking into breaking into writing computer software. There's no opportunities here. And her husband happened to be a senior player at EW Bliss in Derby. And so I went for an interview there and he said, well, you know, why don't you take a couple of weeks holiday and come and see how you fare with us? And so I'd love to, but I've used all my holiday allowance up.

[00:04:08] But I'm happy to sort of put in my resignation. I'll come for two weeks and I understand that at the end of it, there's no job. That's the way it is. So I started there and I was fortunate that the guy I was working for, a guy called Jed Dunkle, was a really, really super hot kid in terms of computer technology. And I was able just literally to learn so much from him in my early career. And then I ended up having a knack for writing computer software. And that stayed with me for quite a while.

[00:04:36] I moved to the States in 1981 for four years working for Wang Computers. And I became a troubleshooter over there for performance issues with Wang technology. And that sort of obsession, I suppose, for performance optimization set the pattern for the rest of my tech career, basically. What a great story. And I think for many other people listening, including myself, as I was born and raised in Derby,

[00:05:01] and my dad gave me that mantle of becoming a Derby County fan for the rest of my life, was the time that you spent at Derby County. And when I think about your time now, I always think back to an interview with Noel Gallagher. And I remember him saying that fans were always saying to him, Oasis fans were saying, why don't you invest in Man City? And he said, I tell them every time I've worked hard my entire life, not for you lot to start throwing crap at my windows after a nil-nil draw at York.

[00:05:27] So as a Derby County fan, I know you've had a lot of stick, but I think one aspect of your story that doesn't get mentioned as much is the fact that you invested and lost 200 million of your own money trying to take Derby County into the Premier League. So looking back as a Derby fan yourself, what do you take away from that entire experience? Yeah. Fine margins. You know, we got to the playoffs, as you know, several times during my time at the club, and we failed on the last hurdle.

[00:05:55] And, you know, the thing that I don't think people quite realise in investing or owning any form of club of Derby size in the championship, and that's that you've pretty much got a right to check for a million pounds every month just to keep the lights on. Yeah. Obviously, that would be a podcast episode on its own, but this is a tech podcast. I want to talk about technology with you today, and you've entered a new chapter with Corpora AI.

[00:06:22] So can you tell me a little bit more about what it is, what sets itself apart from traditional search engines and other AI tools that are being delivered out there? Because I think the world of search is dramatically changing right now. But what is it that you're doing and what are you doing differently here? Well, search is changing. And I've long said the following phrase that search doesn't scale. And everyone looks at me as I'm crazy.

[00:06:46] But what I mean by the fact that search doesn't scale is it doesn't scale for the human, doesn't scale for the user. Because, you know, there's so much information out there. The amount of information you can glean from the top 10 links on Google is pretty limited. Yeah. And of course, now we've seen ChatGPT coming along and indeed other models as well. And these technologies are pretty good at giving you an overview.

[00:07:11] If you look at that overview and you know it's correct, you think, wow, that's incredible. But if you look at that overview that it gives you and you know that it's not correct, you think, what is this? And you really never know. So if you're using it on things that you're well-versed in, that's fine. If you use it on things that you don't know much about, you have to take it at face value. And so our particular role here was that we wanted to go a slightly different way about this.

[00:07:39] We wanted to be able to use technology like the sort of models behind ChatGPT and Google Gemini and Lama and all these other things. We wanted to drive those with real data. So we built a system that was capable of actually looking at huge volumes of unstructured information, text, audio, those sorts of things.

[00:08:01] And the system we built literally can ingest on a single system 2 million documents a second. It's a colossal capability. So we've actually taken a massive amount of open source intelligence, web pages, articles, government reports, you name it. And we've ingested that information, well over 100 petabytes of it. And we put that into our own system, which we call a language graph.

[00:08:28] So it's a bit like a graph database technology and a bit like a vector database, but it's actually superior than both of those in many ways. What it lets us do is it lets us query literally tens of thousands of documents in a blink of an eye. So 4PORAI might look initially a little bit like ChatGPT. You put in a prompt, you ask it a question. But what happens is that it takes that question, it breaks it down.

[00:08:57] It thinks about what's all of the information that you would like to know about that subject. And it runs literally thousands of natural language queries against our own data set. What it then does is it pulls that information together. It deduplicates the information, netting it down to the net unique content, and then effectively uses that content to drive the process of summarizing and creating a research document for you. We call it a research engine.

[00:09:28] And it's really incredible that you can literally ask questions on just about any topic you could think about. And it will produce either a one-page synopsis, it'll pay a four-page exact summary, or a full eight-page comprehensive report. And the difference is this. If you run ChatGPT, you probably won't get any usable citations as to where the information came from. If you use Google Deep Research, you might get 25 or 30. Perplexity, maybe 15.

[00:09:57] You do a large report with Corpora, you'll typically get four or 500 cited articles. It's completely game-changing. Wow, that's some phenomenal stats there. I mean, because I use perplexity, or I have used perplexity as a journalist when I'm researching to get that added sources for the content that I've got there. So your ability to process millions of documents per second is phenomenal.

[00:10:21] So can you tell me a bit more about how people use it, whether they be in medicine, law, or journalism? Because there's so many different opportunities here and a real need to get those citations, those sources when performing research. Well, it's interesting. I mean, firstly, we call it a research engine. So in terms of use cases, let's sort of separate them out for a second. So we have professors at universities using it for research.

[00:10:49] Now, in their use cases, quite often, they want to limit what's searched to be against academic materials, either peer-reviewed documents or cited materials elsewhere. And again, with our product, you can do that. But if you want to really go into research, cited materials quite often won't tell you what is, what I'll say, the leading edge of that subject. Because all things that are new start out unsighted. We take Einstein.

[00:11:19] He went for years with no one acknowledging his theory of relativity. Eventually, of course, it becomes de facto. So if you want to do real research to see what is new and exciting, you have to look beyond just academic articles. So for the sort of academics, they can find a whole raft of articles. And I've had a professor come and see me literally two or three weeks just before Christmas. And he said, the problem is that as researchers, we think we've got it completely nailed.

[00:11:48] But having used Corpora for a few hours, it dawns on me that there's probably 10 or 20 times as much information out there than we uncover. And with Corpora, it's so easy to find that. So it changes your view of a subject that you thought you knew because you can now get access to all this additional rich information. And if you think about deep research, and Google's got a product called Deep Research now on the back of Gemini Pro. They call it deep research.

[00:12:18] But if you run the product, it looks at maybe 200 or 300 web pages. Now, if what we're saying is that 200 or 300 web pages in the world is enough to define any subject, we may as well pack up and go home. When we run a research project, it's typically going to pull up between 10,000 and 15,000 articles that it's going to look at before it produces the report. So there's a complete difference in the breadth and depth of what we're doing.

[00:12:48] And the reason is because we actually have this ultra-fast technology that sits underneath the technology that can literally read through documents in the blink of an eye. So we have the ability to be able to map all this data and still return a very fast response. And other people don't do that. So use cases are people wanting to produce new technology. Maybe they're looking to file patents or to understand how technology works in a particular field.

[00:13:15] Or they want to look at maybe a market opportunity and say, well, what is the market opportunity for this somewhere? And how should I approach that? So just about any question that you could ask typically of a model like ChatGPT, you can ask Corpora. The difference is the response you're getting, you're going to know exactly where that response came from. And you made a good point there about the lack of information on traditional search.

[00:13:42] And I think anybody searching for any topic and you get, I don't know, 150,000 results, finding anything beyond page five, there seems to be nothing there. So it is so important to get that more information on any topic at all. And when I was doing a little research on Corpora before you came on the podcast today, one of the comparisons that stood out to me was with the Library of Alexandria.

[00:14:06] So can you expand on how it is empowering users to explore real complex topics and undercover, ultimately groundbreaking insights that you wouldn't get through those traditional search methods? Yeah. Well, I mean, the parallel with the Library of Alexandria was that, you know, you can imagine now going back at the date, this is going back thousands of years now.

[00:14:26] And, you know, someone would turn up at the Library of Alexandria and there'd be all this information in there, but no index really, no way of really being able to even scratch the surface of what was there. And we've reached that same point now with the internet today, that whether we use Google or Perplexity or these tools, we're still actually only looking at literally the tip of the iceberg. And so with Corpora, we're actually taking a belief to the surface with an ability to really understand the information.

[00:14:56] So we have one company who's been looking to help develop their technology, and they've got a rather exotic piece of electronic technology. And they wanted to understand, well, how different are we compared with what's out there?

[00:15:09] So they run a real series of searches, and some of these case studies are on LinkedIn today, and they understood straight away just how different what they were doing was, but also the touch points with other technology that might be, let's say, parallel to what they're doing and could actually help them actually further where they go. So you can do some incredible things with this. Examples are that Corpora will let you research articles in a foreign language, but produce the output for you in English.

[00:15:40] So you can literally do things like the Russia-Ukraine war, and you can say, well, okay, I want to look at what the English source documents say about that, and I want to look at what the Ukrainian-written documents say about that, or the Russian ones. And you can do all of those things with the output in whatever language you choose to pick. So in this scenario, what we're trying to do is provide access to information that you might otherwise not have seen.

[00:16:05] So if you're a medical practitioner, and you know pretty well what's available in the UK and what's been certified over here, but it might be useful sometimes to look in Italy and look at just all the research in Italian documents, because if it's certified over there, the chances are that it's probably going to be certification for the Hover here at some point soon.

[00:16:25] So those sorts of abilities to be able to look under the surface, to be able to look at things that traditional search and these other products don't let you do is a really phenomenal part of what Corpora is about. And for every academic out there, there's what I call the casual, almost curious inquirer.

[00:16:47] When you know that you can ask questions about anything, if you go into a meeting in, let's say, 10 minutes' time with Corpora, you could go to that meeting having known nothing about the subject, and actually probably go there in a pretty informed position, literally in the space of one or two minutes. And I think finding information online is desperate for a change.

[00:17:10] It's been arguably to blame for so much polarisation we see there, because nuance and context is often lost. I recently read Meta have announced that they're going to be pushing political content that they haven't done before. And Google have been accused in the past of determining what information or narratives get featured on that page one when you're searching for information. So transparency and credibility is so important in any research process.

[00:17:37] So how do you approach this, particularly with features like source attribution, verifiable content, and ensuring that all sides of any information is present there? How do you approach something like that? Well, we try and have an unbiased view in terms of how our technology works. So our technology is effectively trying to understand your question, trying to functionally decompose that into its constituent components,

[00:18:03] and then goes out and uses natural language queries against each of those, and then compiles that, and uses a weighted model, therefore, to give you not only one citation link, but several that will allow you then to verify the authenticity or the veracity of the information that's presented to you. We're not making an opinion on this thing. We're not trying to, you know, bring a bias.

[00:18:27] And unlike search engines, where they have to basically bring back documents that they think are the best match, the problem with things being the best match is that it brings a lot of repetition. So you're not going to see the diversity of messaging out there. Because we can look at literally 50 to 100 times as many source documents. We're always taking it in a broader perspective.

[00:18:52] So you will get more of the fringe discussion points and the counterbalancing views than you would do with other technology. But ultimately, it has to be the user or the viewer's choice as to whether they want to believe that source of the material. If the product doesn't give you the source to look at, you have no idea where it came from. But if you have the ability to follow the citation, to read the source material,

[00:19:18] you will at least have an ability to be able to say, well, I trust this source. It's credible. And therefore, I'm going to believe what's being said here. But as we all know, as the internet grows, AI is going to be both a blessing and a curse. Because whereas AI is going to help us pick through mountains of information, it's also going to be the biggest generator of mountains of information.

[00:19:45] The catch is, it's not going to be generating much new information. It's going to basically regurgitating different views. In fact, if you've used ChatGPT, get two people side by side. As ChatGPT, exactly the same question, you'll get pretty much the same answer. But how it's written will be completely different. It will create two different articles that probably say much the same thing. Imagine that now on a search engine. And all of a sudden, the top 100 articles all say the same. They look different.

[00:20:15] They say it differently. But they say exactly the same message. And some of the features that really stood out for me, because I'm a bit of an impatient user, that wants information, I need access to those citations very quickly, is how you provide real-time updates. And most importantly of all, near-zero latency. These feel like unique features to me. So how do these capabilities impact high-stakes research and decision-making for enterprise users? Because there's so much hype around technology at the moment.

[00:20:43] But this feels like a real ROI, a real big value moment. Well, it is. I mean, so if you take, let's start with ChatGPT and move forward to where we are. So ChatGPT is typically updated about every three to six months. So they top up or retrain the model. So ChatGPT is always going to be looking at data that is, on average, three months or more out of date. So if you ask it questions about what happened in January 2025, it really doesn't know much about it.

[00:21:12] To bridge that gap, then you'll find now that ChatGPT and other products now will go out and look at the web to bring newer content into play. But the amount of information you can look at in that model is relatively limited. So you're never going to get a full view of what's out there. What we have is a system that is constantly ingesting new information.

[00:21:36] So if you're on the highest level of capability with a product, you're literally talking about real-time data coming in, stock prices coming in, news articles coming in, you name it. So we have an ability to process information faster than other people. And therefore, the latency between that information emerging and it being available on the platform is that much faster.

[00:21:58] And that makes a massive difference to people looking to see whether or not they should be buying or selling a stock or whether this particular news article is going to move a market or whatever. Those things can be huge. But it can also go as far as to say, well, there may be other times when people want to see things in, let's say, an updated view. Maybe someone's looking for new treatments for a particular disease or concern that they have.

[00:22:26] And so you always want to have an ability to look at up-to-date information. But you also need to look at the historical view as well, because if you need to learn a subject from scratch, you need both. You need the ability to look at the breadth and depth of what's there before, and you need the ability to look at what's new and has happened in the recent days, months, even seconds. So if we zoom out for a moment, I think ultimately your mission is to democratize access to advanced research.

[00:22:55] So to bring to life everything that we've talked about today in our conversation, how do you see features like adaptive categorization, sophisticated summarization? How do you see all these things transforming the research experience, whether we have an expert listening or a non-specialist listening in performing research tasks? How do you see it all coming together and making that big difference? I think the difference is this, that, you know, firstly, the way things are today, and we only have to look back at what happened during COVID.

[00:23:25] During COVID, we had a lot of things that were going very badly wrong in the world. But our reaction to it meant that we suddenly started to do things quicker. We started to develop new methods of developing vaccines and treatments, and that accelerated at quite a pace. So I think today, when people start to use Corpoor, they will realize that often, even people that know a subject don't really know it from the inside out.

[00:23:54] There's always more information or maybe taking a newer view of things. So in this model, we're trying to not just democratize this thing, but we're trying to provide access to that breadth and depth of a subject. So people can actually then start to learn that much faster. It's about accelerating people's ability to learn something, to know something.

[00:24:16] And I think that goes right the way across the board, whether that's someone looking to, let's say, do research for, I want to think about doing a PhD. What should I do the PhD on? Well, Corpoor is a great tool to be able to go out and say, well, I could look at this. I could look at that. I could maybe do this. And it will probably change the way that they would think about what they wanted to do their PhD on.

[00:24:39] It's not going to write their PhD for them, but it's going to help them look around at the different subjects that they might want to be interested in actually doing something with. So those things are huge. But let's take some of the sort of, let's say, less well-off areas in the world. A product like Corpoor could be transformational for them, where they literally could access to information and knowledge that they're probably never going to get within their existing school and educational systems,

[00:25:07] that this could change their lives, literally. And we started the podcast today talking about your origin story, what put you on this path. And throughout our conversation, we have been very forward-looking, looking into the future. And, of course, none of us are able to achieve any degree of success without some help along the way. So a question I'd love to ask you here, you've experienced so much success. Is there a particular person that you're grateful towards who maybe saw something in you, invested a little time in you, helped you get you where you are,

[00:25:37] that we can give a little thank you all shout-out today? Is there any person or people that you'd like to give that shout-out to? There are lots, but there's two in particular. So the first one, Sally passed away a couple of years ago, having had COVID. And one of my best friends was a guy called Jeff Shingles, CBE. Jeff Shingles was the founding, if you like, employee of digital equipment in the UK.

[00:26:01] And he ran that business from, I think, something like about eight employees to about 4,000 people and many billions of pounds of turnover. And he had such a following. Anyone who worked at Digital would know his name instantly. And he just was a wealth of common sense that was fantastic. And I remember one meeting that we had as a board meeting of a company I had that was in really bad trouble because of the property crash of 91, 92.

[00:26:29] And he turned to the meeting and he said to the CFO, what do the banks normally do when you get to this position? And the CFO said, well, they normally take you to the wall to make sure they've got the best deal they can. And Jeff just turned to the meeting and said, in that case, then let's start building walls. And it was lost on many of the other board members. But you can understand exactly how profound that was because you change the way you look at things.

[00:26:59] And the other person who is probably the greatest marketing mind that I've had the pleasure to know happened to be the person who ran Wang Equipment, the competitor to Digital Equipment in the UK, is currently a guy called Sir Ken Olyssa. And Sir Ken Olyssa is also the Lord Lieutenant of London. So you'll see him quite often appearing with what was then the prince but now the king. But he's a really, really close friend and has given me so much support and help over the years.

[00:27:29] Absolutely incredible. So those two stand out. There's many others, but those two in particular. Well, I think that's a powerful moment to end the podcast today. But before I do let you go, for anybody listening wanting to find out more information about Corpora, how do they get up and running? Is there pricing involved? All that kind of stuff. Where would you like to point everyone listening and any advice on how to get up and running? So ever so easy.

[00:27:54] You go along to corpora.ai, C-O-R-P-O-R-A dot A-I. You register on the site. It's free to register. And you can then run as many queries as you want. And for the time being, we're going to leave this thing open. It's going to be free for quite a while to come. We're more keen on getting people using it and getting people to understand what it is. So really enjoy it. And it is so easy to use, just like using a search engine, but the results are far more powerful.

[00:28:22] Well, having listened to your statistics there, I would be leaving perplexity behind and giving this a try. The idea of real-time information, adaptive categorization, content normalization, document discovery, big long list of things there. So I'm going to take that away and have a play myself. I urge everyone listening to do the same. But more than anything, thanks for starting this conversation today, man. Thank you, Neil. Thank you for your time. Really appreciate it.

[00:28:47] As we wrap up this conversation with Mel Morris today, I think it's clear that Corpora isn't just another research tool. It feels like a fundamental shift in how we approach knowledge discovery. From the ability to deliver real-time sourced insights to its potential to transform entire industries, I think Corpora could have a lasting impact on how individuals and organizations navigate today's complex information landscape.

[00:29:15] There's a lot of digital noise out there. Mel's journey from the success of Candy Crush to the latest venture that we're talking about today shows a remarkable ability to spot opportunities where technology can make a real difference. But over to you. I'm going to ask everybody listening a question here, and I would love for you to get straight back to me. Has your search habits changed? No matter what your job, what your age, if you need an answer to a question, or you need

[00:29:43] information, or you're performing research, are you still going into Google and simply Googling it? Or do you feel the need for more? Is ChatGPT, for example, is that allowing you to see a bigger picture and taking you away from things like sponsored ads and giving you more in-depth information? And has that led you to other platforms? Claude, Google, Perplexity. There's so many more emerging ways of searching.

[00:30:10] And I think even our smartphones are now going more conversational search through AI rather than typing into a box to ask Google a question. So much going on here. So many big changes. So how is your searching methods evolving? And will you be checking out Corpora? Me personally, I love just going out there and having a play with these things. People always say with AI, oh, what shall I do? It's all a bit overwhelming. It's daunting. For me, I say, if it's free, go out there. Have a play with it.

[00:30:40] Get using it. What do you notice difference? So I'm going to go check out Corpora. I urge you to do the same. It is free at the moment. It's out there for everyone to use. Let me know your thoughts. Good, bad, and indifferent. I'm not here to promote it. I just want to hear how this world is evolving and how it impacts you, how it could impact your work. So please email me, techblogwriteroutlook.com. You can connect with me on LinkedIn, just at Neil C. Hughes. Equally, Twitter, Instagram, just at Neil C. Hughes.

[00:31:10] I'd love to hear your perspectives. This isn't a monologue. It is a dialogue. So please join the conversation. And if search isn't your bag, don't worry. I'm going to be returning again tomorrow with another topic around how technology is changing the way that we live, work, and even changing our world in some aspects. But I've taken up far too much of your time. So meet me here, same time, same place tomorrow, and we'll do it all again with a different guest. Speak to you then. Bye for now.

[00:31:43] Bye for now.