Who Is Winning The AI Race? The Clarivate AI50 Report Has The Receipts

[00:00:05] - [Speaker 0]
How do you separate genuine technological progress from all the hype that surrounds everything at AI today? Every day, we hear about new AI models, new breakthroughs, new companies claiming leadership in the AI race. But behind all of those headlines sits a much quieter signal. One that often tells a more reliable story. Patent filings, research collaboration, and innovation data.

[00:00:34] - [Speaker 0]
These are the things that reveal who is really building the future and where global power is beginning to concentrate. So my guest today spends his time analyzing enormous volumes of intellectual property and innovation data to uncover exactly those signals. His teams track how technologies develop while organizations are pushing boundaries and what the data really tells us about that global race for AI leadership right now. So my guest today will share insights from their latest research, including the astonishing scale of AI invention activity and why organizations from just a handful of countries are currently dominating the landscape. And we'll also talk about how the collaboration between academia and industry is accelerating breakthroughs and explain why safety, governance, and trust might become the deciding factors in the next phase of AI adoption.

[00:01:35] - [Speaker 0]
So if you wanna understand where AI innovation is really happening, what the signals inside patent data reveal about the future of technology, you're

[00:01:45] - [Speaker 1]
in

[00:01:45] - [Speaker 0]
the right place. But enough from me. Let me introduce you to my guest right now. So thank you for joining me on the podcast today. Can you tell everyone listening a little about who you are and what you do?

[00:01:58] - [Speaker 1]
Yeah. Thanks very much, Neil. It's a huge pleasure to be here. I head up the Clarivate Center for IP and Innovation Research. Our job is is kind of conceptually quite simple, maybe a little bit harder in practice, which is to take enormous quantities of of data, innovation and intellectual property data, and understand what it means.

[00:02:19] - [Speaker 1]
And that means things like where is technology headed, and who's behind that, who's building it, and then overlay on top of it measures that speak to performance benchmarks, essentially. Right? So what we're doing is we're we're attempting to find the signal in these in in a very large amount of evidence. Right now, as you can imagine, the very much top of mind question is what we started calling the mass ambiguity that is AI technology when and the difficulty that that's creating, and that's why we're really focusing in on it. And that's where our AI 50 comes in.

[00:02:54] - [Speaker 1]
It's really part of sorting through that ambiguity.

[00:02:58] - [Speaker 0]
And I'm glad you mentioned that because before you joined me today, I was thumbing through your latest research at Clarivay, which shows AI patent filings crossed over 1,000,000 in 2025 and have doubled repeatedly since 2019. So what does that pace of growth tell us about how seriously organizations and governments are about treating AI innovation right now? It feels almost like a a gold rush of sorts, doesn't it?

[00:03:24] - [Speaker 1]
Yeah. It does. I mean, it's telling us a few things. It's telling us that AI isn't just chatbots that we're all pretty pretty used to by now. It means that when this is cropping up like this in invention data, it means that it's rapidly being embedded in industries at scale.

[00:03:40] - [Speaker 1]
And it tells us that it's shifting along kind of waves of generations of use cases and applications really quickly. You know, you mentioned the doubling. Another way of describing repeated doubling is that's exponential growth. Right? And Yeah.

[00:03:52] - [Speaker 1]
That's what we're seeing in AI invention levels. That is probably mirroring the exponential increase in capability that these extraordinary reasoning models are producing. Tell you one way of how we deal with that pace. It's moving so fast that probably for the first time in my career, I've actually had to change the time axis that I plot stuff on. Right?

[00:04:12] - [Speaker 1]
I'm now plotting data weekly when I've never had to do that before in my career. You said AI is more than a million inventions. It's actually more than 1,100,000 inventions so far. It's so big. And one of the ways that I wanted to maybe crystallize this, that's about 8,000 brand new AI inventions a week right now.

[00:04:31] - [Speaker 0]
Wow.

[00:04:32] - [Speaker 1]
A year ago, it was half that. Two years ago, it was one or 2,000 a week. Ten years ago, this kind of ages me. Twenty sixteen doesn't feel that long ago. It was a 100 a week.

[00:04:44] - [Speaker 1]
So this has really come from nowhere. The other way that you kinda need to look at it is, well, what's 8,000 inventions a week? How does that compare to something else that we can kind of get our get our teeth into? It is now twice the activity each week than communications technology. It's three times what pharmaceutical activity, patent activity is.

[00:05:04] - [Speaker 1]
So so, you know, you were talking about how seriously do we need to take this. I'm not sure that being around this is really an option anymore

[00:05:12] - [Speaker 0]
Yeah.

[00:05:12] - [Speaker 1]
Yeah. Because of that scale. I'm also not really, hopefully, one for too much hyperbole because it it can make you a little bit nervous, but it's really difficult to avoid describing what's going on as anything other than a revolution. This is a fundamental change in the foundation of what modern innovation actually is.

[00:05:34] - [Speaker 0]
So if we take that 1,100,000 AI inventions, if we take a massive figure like that, and then something else that stood out to me here, we've took this to the next level. Your report also highlights that organizations are from just four countries that are accounting for 82% of leading AI innovation, which, again, not a breathtaking stat, but where is global power currently concentrated? And what explains that level of dominance that we're seeing now?

[00:06:03] - [Speaker 1]
Yeah. So four countries account for the location of, like you said, just over 80% of this AI 50. They are Mainland China, United States, South Korea, and Japan. If you were to group Europe as one, because, obviously, we break them out into individual countries, it would place at number three. So you can maybe add a sort of fifth one in there after China and The US.

[00:06:24] - [Speaker 1]
There's a couple of ways of explaining that. Right? One is and this isn't unique particularly unique to AI. These are the leading knowledge in r and d economies in the world. They have the networks.

[00:06:33] - [Speaker 1]
They have the education systems. They have the research ecosystems. They have the funding ecosystems to do what they're doing. Another way of looking at it is this kinda reflects the tech stack of AI itself. Right?

[00:06:45] - [Speaker 1]
So chips manufacturing, chip design, the mathematical research base that you need, the software powerhouses that you need, and then, of course, the industries that particularly lean on cutting edge innovation in order to create their next product. All of that, I think, accounts for why we see those four slash five places being being dominant. The other point that I would make is really around China. China is is really interesting because it has a lot of diversity. It has loads of organizations.

[00:07:17] - [Speaker 1]
That's really a feature of the Chinese innovation ecosystem. It probably also accounts for why they have quite a large number of organizations in this list.

[00:07:27] - [Speaker 0]
A question I've got to ask here. Here in The UK and across Europe, they they often get accused of being too focused on overregulation and red tape rather than innovation. And I know it's a a fine balance, but do you think sometimes that UK and Europe maybe overregulate? It's incredibly important, and wouldn't knock anyone for doing this. But do we think we focus too much on that side of things, or or is it not that at all?

[00:07:53] - [Speaker 1]
Difficult to say. So there's a couple of things that I would unpack from that. We can maybe get into this a little bit later when we talk about deployment of AI technologies. Regulation is kind of interesting. So one of the reasons that you would look at what makes up a knowledge economy, what makes up an economy that that has a high level of research commercialization going on.

[00:08:12] - [Speaker 1]
One is you wanna be able to have the infrastructure within there, the the the legal systems, the court systems, the contracting systems, the commercial ecosystems that actually enable that to actually happen. It's not just enough to kinda go write a scientific paper. These ideas need to kinda flourish into an economy itself. The other thing I would say is that specific to AI, it's looking like the safety aspect of this is actually pretty important in terms of uptake. And we you know, there's there's some signal that we have within there that that that speaks to those kind of points.

[00:08:44] - [Speaker 1]
So I wouldn't say disadvantage. I would say, when, you know, regulation as a whole, absolutely, it can go too far. It can stifle, but we're also talking maybe a little bit around enrichment.

[00:08:56] - [Speaker 0]
Yeah. I completely get that. And going back to the findings in your report, I'm also curious, were there any unexpected companies or countries or regions that are emerging as leaders in this latest dataset? And what might the market be overlooking about them? Because it's very easy just to focus on the big four there that you mentioned.

[00:09:15] - [Speaker 0]
But anything else that you that you've found there?

[00:09:18] - [Speaker 1]
So first, I'd say the firms that you expect to see are here. Secondly, I would say that some of the companies that we've that have become so habitual to us in the discussion of AI models aren't here, and that's because either they're kinda too new, which in is interesting. They're brand new companies. Or that they're leaning on open source and trade secret approaches to technology protection. And and that that's a really important thing that we need to say here, that invention data alone, which is what we're relying on here, is not all of AI innovation.

[00:09:51] - [Speaker 1]
But it does provide an amazing proxy and sample, and that's because of its size. There's a lot of it here. It's the fact that it's consistent from country to country, and it has a lot of structure. It has super technical details. I don't know, Neil, if you've ever gone and read a patent, but it's a 20 page technical document.

[00:10:06] - [Speaker 1]
Right? So there's a lot of detail inside there. One of the things that I think is probably quite surprising is that we saw a lot of diversity. Now, you know, we see organizations that are creating the fundamentals of AI, so that, you know, the chip hardware, the models. But we also see those that are deploying AI technology into their industry.

[00:10:24] - [Speaker 1]
So think medtech, energy, retail, social media, industrial technology that filters through to the rest of the economy. Maybe two surprises that I would call out. One is Asian academic institutions, and the other is the very familiar big automotive companies that we all know. When we talk about AI in general terms, we don't tend to think of them, but they are there.

[00:10:51] - [Speaker 0]
Yeah. And looking at that list, I think the top 50 includes both large tech corporations and academic institutions. So how does that mix between commercial players and universities? How's that helping shape the direction and indeed speed of AI development? Because, again, quite a fascinating point too.

[00:11:11] - [Speaker 1]
Yeah. We've looked at this crossover between academia and corporations a lot over the last couple of years. It probably makes it faster, the first thing I would say. And I would also say it measurably makes it better. Yeah.

[00:11:23] - [Speaker 1]
The reason behind that is that AI, it's a huge melting pot of technologies and disciplines. It's probably one of the causes of that kind of omni ambiguity that we've I was mentioning earlier. Right? This is really, really difficult. It's really tough technology.

[00:11:39] - [Speaker 1]
It's speaking to a a a diversity of technical need, expertise that you need to be able to go into this stuff. So you're talking about that mix of super advanced mathematics, computer science, data engineering, ethics. This is really broad. Add into that the difficulty around deployment, how these technologies perform, how they need to be engineered to work in these different environments that we're talking about, some of which are very heavily regulated medical technologies, automotive technologies. We've been measuring this for a while.

[00:12:11] - [Speaker 1]
You can actually measure that kind of people need, the the expertise density that you need on an invention. If I go a little bit broader, look at everything. Right? So for decades and decades, patents had on average two inventors on them across any technology. So think material science, telecoms, electronics, and everything else.

[00:12:28] - [Speaker 1]
That started to bend up in the February. It started to sort of start to increase. By 2010, it was three people, not two people. Last year, it's four. In AI, it's over four.

[00:12:41] - [Speaker 1]
And in in in really high strength AI inventions, it's actually edging closer to five. Couple of ways to pack that. It's getting harder to innovate is one of the things that I take about that. It's actually more sort of discipline intensive. Another way of looking at it, going back to the sort of academic angle that when corporations and academia are kinda cropping up together, it's usually because there's a really tough problem to solve.

[00:13:03] - [Speaker 1]
Tougher solved problems tend to be more valuable solutions. Let's put it like that. We know that when ideas when research teams are crossing that corporate academic boundary, the ideas produced, the solutions that are produced are more often part of a kind of super high strength invention tier. They are stronger inventions. So the mix that we're seeing in this list, it doesn't just sort of speak to a need.

[00:13:32] - [Speaker 1]
It it speaks to a kind of conclusion, which is that this collaboration is not just nice. It doesn't just make great marketing copy or a good press release. It's actually performance advantage.

[00:13:42] - [Speaker 0]
And just to throw something else into the mix here from a geopolitical standpoint, which is a whole episode on its own at the moment. Yeah. How should business leaders interpret the concentration of those AI patents? And does patent leadership translate directly into economical strategic advantage, do you think?

[00:14:02] - [Speaker 1]
So, yeah, we we need to be a little bit clear. Right? If you file a patent tomorrow, it doesn't immediately become GDP. It doesn't immediately become revenue. Yeah.

[00:14:09] - [Speaker 1]
At least not yet. So one of the signals that we have when we look at really strong ideas, very, very impactful ideas, and we tend to look at that as, like, a right at the top or top half a percent of the strongest ideas that we have. That slice of activity, it maps to advantage over the longer term, medium to longer term. And that means that this data is a really good forward indicator of future capability. It's also a current indicator of intent.

[00:14:38] - [Speaker 1]
Right? Commercial technical direction. The geopolitical piece. So the geographic view of this data is really interesting because I think it maps to where we know AI is kinda coming from. We talked earlier about the list and where everybody's based.

[00:14:50] - [Speaker 1]
Quite a lot of this comes down to what we mean by leadership. We talked earlier on about, you know, the AI 50 and where they are. For a macro view, you can't just limit to 50 data points. You have to go to a higher level. The 50 companies alone isn't enough.

[00:15:05] - [Speaker 1]
And then the other piece that we need to sort of name what we're talking about, this is a question that really surrounds Mainland China versus Europe and The US. Right? That's where a lot of this is coming from. And the idea of a scoreboard and the idea of of of who's ahead. Right?

[00:15:19] - [Speaker 1]
We did that a couple of different ways. Right? So one of the ways you can do is count how many AI patents you've got. Fundamentally, that's a risky way of approaching this because not one it doesn't differentiate. One idea is not the same as another one.

[00:15:32] - [Speaker 1]
But if you do that, what you see is China has more than Europe and The US put together. So just to be clear, counting them, they've got more. But switch to that super high strength piece, the top half a percent layer and the balance shift back. Europe and The US are ahead. They've got more of this leading technology in their in their back pocket.

[00:15:54] - [Speaker 1]
But there's a lot of nuance to this. Right? AI, it's not a single slab of tech. It's not a box marked AI. Right?

[00:16:02] - [Speaker 1]
It's got layers of technology within it. It has phases within those layers of technology. So if we look at China, it's relatively newer to AI. Maybe it doesn't have the heritage that goes back to the seventies that maybe The US has. It maybe has a little bit less less depth of capability in the hardware layer.

[00:16:19] - [Speaker 1]
We can see that within our data. It has a little bit less of the enrichment technologies around AI, and I'm I by that, I mean, kinda safety and usability related technologies. But it it perhaps has a better than expected situation in the generative model space. We can see that within our data as well. But maybe I would go back to the collaborative point, advantage and leadership.

[00:16:44] - [Speaker 1]
We are tending to frame this question of technology stopping a a country border. Right? And that's not how this works. Ideas don't see borders. If we, between us, come up with something today, somebody else is gonna take that and go, well, that's now state of the art, and I'm gonna work from there onwards at it and go over the top of it.

[00:17:05] - [Speaker 1]
And one of the ways that we kind of look at this is is, well, let's look at when they work together. Let's look at an idea that contains an inventor team with both people based in China and people based in The US. When you do that and you look at them, more often they produce one of these high strength inventions. And so I think it means for leaders that are grappling with this, I think it actually comes back to the basics. What teams, what backgrounds, what specialisms, what research programs do you need to build the technology that you want?

[00:17:36] - [Speaker 1]
And then maybe the where are people based piece kinda dropped away a little bit.

[00:17:41] - [Speaker 0]
And I'm not sure if you're able to answer this or name anyone, but I'm curious if you look at all the data that you've got there and all the insights, are there any signs or signals that certain companies are indeed quietly building a long term advantage without generating the the same public attention, the big headlines around that we often see in our news feeds. Anybody under the radar there that you see?

[00:18:04] - [Speaker 1]
So it's interesting. Right? So one of my the way that I would maybe approach that, Neil, is that that's kind of what this AI 50 list is. Right? It is companies that are building long term advantage and largely just getting on with it.

[00:18:17] - [Speaker 1]
I think it's useful for us to it's not just a a list of companies, right, and and and research institutions. What we wanted to do was make this a kinda useful list for anybody that's grappling with AI as a strategic question. Clearly, it's one of the most common conversation topics in many, many meeting rooms. And one of the ways that we could do that is kinda profile this 50 and profile it in a way that summarizes, distills down approaches to developing this. And so what we did is we took our 50, and we broke it into how much do they contribute to the core, so the hardware, the models, essentially, versus how much of their total research effort is AI focused.

[00:18:56] - [Speaker 1]
So pick an automotive company, 90% plus of their stuff is gonna be door handles and engines and and entertainment systems, and then there's this AI layer within it. That doesn't mean that they're not a major AI innovator. They're in our list, but there's a different level of focus within it. If you do that, you get a fairly nice two by two grid, standard two by Yeah. Those that are more focused on AI enablement and those that are more focused on AI deployment with sort of two flavors within those each.

[00:19:27] - [Speaker 1]
What that does is that gives you a really good set of lenses to to look at how they're operating. What does AI innovation leadership actually look like using just these 50 organizations as a sort of micro case study? What that does is it kind of informs you who you could or should be working with. Could you source technology? Could you partner?

[00:19:46] - [Speaker 1]
It informs your kind of research direction. Are you translating this stuff? Are you scaling it? Do you need to make it more specific? And it's doing that without needing to or maybe pretending to forecast what happens next, you know, the business cycle that's still to come.

[00:20:02] - [Speaker 1]
So my answer to your question is I think that this is what the AI 50 actually is. It's the quietly getting on with it. They are producing very high quality technologies. They're protecting them very judiciously. They're building these repeatable systems.

[00:20:16] - [Speaker 1]
That means that they're building repeatable value for the years to come, and that kind of means that they're building a compounding advantage. Right?

[00:20:24] - [Speaker 0]
And everything we're talking about today, just for people listening, I will be posting a link to the report so people can have a little look through that. And for investors and enterprise leaders listening, any signals in pattern activity that they should be watching to maybe understand where the next wave of AI breakthroughs may be emerging? Anything you'd say to those people listening?

[00:20:45] - [Speaker 1]
Yeah. There's a lot. So I appreciate, you know, the spot for the link. There's maybe a couple that I would fill out. So the first one, one of the one of the interesting signals we pulled out is LLMs.

[00:20:54] - [Speaker 1]
Right? LLMs are still at the cutting edge of this technology spectrum, but they're they're being joined by RAG technology, retrieval augmentation technology. That is right at the frontier of of a spectrum of maturity that that we place these technologies on. What that's telling us is that the generative wave, it looks like it requires that ground truth layer as assurance, and it's not yet solved. The second thing I would say is and it's very related to that, which is around explainability and safety in general.

[00:21:27] - [Speaker 1]
So this is about usable systems, testable systems, trusted systems. I mentioned it a little bit earlier on. This looks like a prerequisite to mass deployment, right, of AI. When we look at these deployment waves, the most high risk applications of AI, we kinda define that as kind of finance people laws. They are yet to be as mature as some of the other use cases.

[00:21:55] - [Speaker 1]
And so unpack that a little bit. What does that mean? That means that that guardrails, we talked a little bit about regulation earlier on, they aren't just there because governments want it or because it makes for good PR copy, although both of those things are true. It looks like it's strategically necessary. There's a hump that we kinda need to get over.

[00:22:15] - [Speaker 1]
And that's particularly true if you think about sort of agentic deployment of AI where it's acting and not just thinking. So that's my second point. The third one that I would point to is that we are bang in the middle of the second generation of AI deployment. And the way that I would summarize that is that it is as we talked about earlier, it's on from chatbots. This is segment specific use of AI.

[00:22:37] - [Speaker 1]
So think life sciences, think industrial technologies, think simulation, think software coding, which we kind of already know. It's the one that is has come to the fore in terms of everybody's knowledge. You see that segment specificity in terms of these maturity measures that I was talking about, but you also see it within the AI 50 list itself. That's the diversity we were talking about.

[00:23:00] - [Speaker 0]
And I was also reading earlier that because of algorithms and things, we we've stopped trusting in facts or going with facts, following facts. We've actually going into echo chambers, getting confirmation bias, and having our own worldview fed back to us, and facts are where it's at. And one of the reasons I wanted to bring this up today is one of the things I try and do for my guests is give them an opportunity to boost any myths and misconceptions that they may see around AI or your area of expertise. And you're someone right in the heart that you've done this report. You've got hard facts in front of you, but then you may retreat to LinkedIn and Reddit and see a lot of maybe exaggerated claims, untruths, or myths.

[00:23:39] - [Speaker 0]
Any myths or misconceptions that that you see that frustrate you that we can maybe lay to rest today?

[00:23:45] - [Speaker 1]
So I there's maybe two truths that I would focus in on that our data is actually talking about. They speak a little bit to risk and opportunity that that is that that only ambiguity that that we've been talking about. Evidentially, when we measured the maturity of these technologies, and we're doing that by looking at kind of the shape of the development curve, how much academia is in here, how much they look commercialized or they look lack of commercialization activity. When we did that, we actually saw a little bit of a paradox. We saw in the numbers that the usage of AI in deployment kind of in in being used for a specific purpose, that slab of data looked to be more mature than the core technologies of AI itself.

[00:24:31] - [Speaker 1]
Wow. And it so to summarize that, it looks like deployment is out running development in places.

[00:24:38] - [Speaker 0]
Yeah.

[00:24:38] - [Speaker 1]
And I think that this is a kind of quantification of something that we can kind of feel in AI discussions, and that's in the difficulty in getting AI to actually create revenue, to actually create profitability. It surprised us when we saw it in the data because what it's speaking to is what we dubbed the risk bubble. And this is something that I think a lot of people are grappling with, which is why you spend a lot of money on a cutting edge AI model. You put it into your product. You put it into your service only to find two months later that there's a better and more powerful model available.

[00:25:09] - [Speaker 1]
And there's kind of a systemic and inherent tech debt in AI. All the while, there's this hardware model deployment loop that keeps compounding. The hardware gets better. The models can you know, there's more compute. The models can therefore do more.

[00:25:25] - [Speaker 1]
That means that they can go into more areas of of industry and the economy. How you interpret that, how you act on that is quite difficult. Right? But what it could mean is that the winners in this process and feel free to define a winner. Let's look at sort of three years.

[00:25:42] - [Speaker 1]
Right? But it looks like it won't just be those that are moving really fast. It'll be those that have a re they know why they're doing this. They have a really strong set of principles and kind of infrastructure and governance rules that accounts for that speed. You're gonna have to think about modularity.

[00:25:59] - [Speaker 1]
And then the other truth is what I I think I've been kind of alluding to this within this data all the way through, which is the need for assurance. And I would call this safety is strategy. We kinda sort of discovered this conclusion within our datasets. Adoption in many, many industries and segments is going to be based on trust and usability and risk mitigation and humans in, humans on the decision lead. And that goes back to the same point.

[00:26:31] - [Speaker 1]
Speed is you know, it's gonna get you a prototype and demo up and running. You're gonna learn really, really quickly. But AI assurance is probably what's gonna enable you to win a market.

[00:26:43] - [Speaker 0]
And just expanding on that, if we look further ahead at those three to five years that you mentioned there, that race for AI leadership, that's gonna continue to intensify. Anything else you'd leave the listeners with on how global competition might evolve in that time and and the kind of risks of opportunities that governments and corporations alike should prepare for right now. Any anything that you would pass on to them as they they leave this podcast?

[00:27:09] - [Speaker 1]
I would really reiterate that point around governance and speed. You're gonna have to balance those two things. I would really reiterate that safety is strategy piece because it is speaking to does anyone I I always go back to you know, I'm an engineer by sort of academic background, and I go back to something that one of my professors once said, which is engineering is where science meets risk. And that risk can be economic. It can be safety.

[00:27:36] - [Speaker 1]
Does it work? Right? And look at AI within that lens of of economics, safety, and does it work? And that's what we mean by safety and assurance within AI. Are you willing to pay for it because it does work and it's not gonna cause you really big problems downstream?

[00:27:54] - [Speaker 1]
So assurance is gonna be so important when AI moves from that chatbot layer into those parts of our economy where money and people are at risk. That's basically the way that I would summarize it. There's another way of looking at this, which is the you know, we see it somebody cheating at an essay in high school is bad. It's very bad. You should not do it, but it's not gonna cause a bank to collapse.

[00:28:24] - [Speaker 1]
There are two different layers to this, and so that's where I think it's really important. Speed, absolutely. You need to be testing and learning about what works, but assurance is what's gonna get you over the line.

[00:28:34] - [Speaker 0]
So much to take away from listening to you there today. And for people listening, again, I'll add a link to the report that we've referenced throughout. I urge people to check that out, but you do have a lot of events around the world. You'd have a lot of big announcements this year as well, I would imagine. For people listening, wanna keep up to speed with everything that you're doing, where would you like me to point them?

[00:28:54] - [Speaker 1]
Clarivate.com is where you find everything that we produce, the Clarivate AI 50, the top 100 global innovators report that also contains a lot of this information and much, much more that we've done. Also, follow us on LinkedIn, Clarivate for intellectual property. That's where you can get everything.

[00:29:08] - [Speaker 0]
Excellent. Well, I will add links to everything there. As I said, so many big takeaways there. Organizations from just four countries accounting for 82% of the list bought the variety in that top 50 with organizations ranging from big tech corporations to academic institutions. Really is a great read, and there's so many different trends in the data there showing which companies are quietly taking the lead in the innovation race.

[00:29:34] - [Speaker 0]
So please go check that out, but more than anything, just thank you for coming on and bringing all these figures and insights from the report to life. Thank you.

[00:29:43] - [Speaker 1]
Thanks so much, Neil. It's a massive pleasure.

[00:29:45] - [Speaker 0]
What I found fascinating in the conversation today is how the numbers behind innovation often tell a very different story from the headlines that we see in our news feeds every day. I mean, more than a million AI inventions recorded so far, thousands more appear every week, and the majority concentrated in a very small number of research powerhouses. And as my guest explained, leadership in AI is much more about speed or scale. It actually is collaboration, trust, safety, and the ability to turn breakthrough ideas into technologies that people and businesses are willing to rely on. So if you'd like to explore the data for yourself, I'll include a link to that research report in the show notes so you can dig a little bit deeper into that.

[00:30:33] - [Speaker 0]
And as always, if you enjoy conversations that break down complex technology trends into something we can all understand, make sure you subscribe to Tech Talks Daily and head over to Tech Talks Network. I host eight other podcasts over there, 4,000 interviews. So many great conversations in there and fascinating insights. So please check those out. But for today, that's it.

[00:30:57] - [Speaker 0]
Thank you to Ed for joining me, sharing his perspective on what that global AI innovation really looks like behind the scenes, and thank you to each and every one of you for listening. Speak with you again tomorrow. Bye for now.