Humans vs Machines: David Rowan on Curiosity, Creativity, and the Future of AI

“The Angle” Music

Cold OPEN: “There’s a consensus that the future century is being built right now and it’s happening incredibly quickly.”

Justin Thomson

Welcome to The Angle from T. Rowe Price, a podcast for curious investors. Just a reminder that outside of the U.S., this podcast is for investment professionals only. I'm Justin Thomson, head of the T. Rowe Price Investment Institute. My guest today is David Rowan, technology journalist, author and founding editor in chief of Wired magazine in the UK. David has not only spent his career writing about the frontier of technology. He's also an active participant in shaping the direction of innovation as an early stage investor in more than 180 tech companies, and leader of technology focused venture funds.

David. Welcome.

David Rowan

Justin. Thank you. I love that we're in a conversation with curious investors because I've got the feeling that curiosity is going to be one of those human skill sets that determines who makes it through the emerging AI apocalypse, because the machine is not great at curiosity, but it's an amazing human ability to work out instinctively what we think is important and to connect the dots, so keep asking tough questions.

Justin Thomson

Which is what we're here to do. You’re saying the right things there David. I sort of know the answer to this because I know at the weekend you were driven in a Waymo taxi to San Francisco airport and took an overnight flight last night back here to London. So that sort of answers, answers some of my question, but how do you, or how does one stay current in an environment where the world is going exponential?

David Rowan

I think my job is to immerse myself in lots of information and opinions and try and pull the threads together. I'm kind of unemployable. I used to run magazines and newspapers and worked for big media companies, but the world doesn't really work like that now. So, I have to go there to look people in the eye to hear what they are excited about, anxious about, scared about. And, yesterday I got back from a week in San Francisco, essentially having large volumes of caffeine in cafes and talking to the people working inside some of the big AI companies, working on robotics, working on other emergent technologies. And there's nothing like going there without an agenda; just to listen. And two or three things that I came back with.

One of them was a sense that they believe this is the most exciting moment of their career, previous or future. There's a consensus that the future century is being built right now and it's happening incredibly quickly. Related to that, is the second observation. They struggle to re-educate themselves every few weeks as a new foundational model, a new way of developing an AI agent, a new way of thinking makes what they have been doing until then, irrelevant. And a lot of people who are genius at coding are now realizing that they don't need to code. They just need to know how to instruct the code building machine to build extraordinary things.

And the third thing I noticed was, as excited that they were about being in the front row of building the future, they're deeply anxious about the unknown power of this tool they're putting into the world. And we're talking not just about existential threats of people learning how to make bioweapons, using AI. It's the social responsibility. There's an awareness that large numbers of existing careers are going to be unviable in ever shorter time cycles.

Justin Thomson

Let's come back to the anxieties that this all brings. But I just want to focus on why this is such a seminal moment. I mean, aren’t all moments of technological revolution, seminal. Why is this different from the dotcom, or the advent of mobile technology, or the printing press, or electricity?

David Rowan

Because, as Dario Amodei, who leads Anthropic, one of the big AI companies, says, we are creating a country of geniuses. We are creating Nobel Prize winning knowledge for every vertical that's accessible pretty much to everybody who has a network connection and the machine is constantly learning through feedback with how people are using it. How to get more and more effective at simulating, at synthesizing data, at processing data to provide accessible language-based ways of communicating complexity. And that means if you are designing medicines to try and fight cancer, there's a whole new toolkit. Even among the people building these tools, about what it's capable of, that mean this is a wake up moment.

Justin Thomson

And who should be responsible for regulating this? Should it be sovereign? Should it be supranational organizations? Should it be Silicon Valley? Where does the responsibility light in your opinion?

David Rowan

Well, I think we need to start with the word curiosity. The people who get to make decisions about our structures, our legal framework, the economy, need to really understand what's happening. Because it's moving so fast that I don't really want people running these companies to make all the decisions in an unchallenged era. Anybody in a position of leadership in a company, in a university, in a government department, in a nonprofit needs to get familiar with these tools right now, needs not to see them as some other thing, but to understand that they're going to be incorporated into, you know, all our institutions. Start playing with them, start using the professional version of Claude or ChatGPT, in order to surprise yourself, because I think one of the challenges is we've all tried an early version of a large language model that wasn't entirely reliable and made us kind of dismiss it. That was last year. What's happening now is it's able to do things at an incredible level of complexity, and unless you're familiar with those tools, you can't make informed decisions. I see AI is a bit like electricity, and it's going to be there powering all sorts of things. It’s how people are choosing to use it to solve particular problems, to build growth opportunities.

Justin Thomson

It's interesting that you make the analogy between AI and electricity. We have made the same comparison, or specifically Dom Rizzo, who is one of our technology investors, has made that comparison. Yes, it will be ubiquitous. But isn't electricity just electricity? Will all intelligence eventually carry the same voltage? Or does the arms race that we're seeing at the moment to create the most powerful model with the most powerful chips continue?

David Rowan

It comes down to the almost zero cost of using these tools. We are surfing exponential curves where the cost of solar energy and the cost of battery storage, and the cost of processing large amounts of information is heading towards zero. And I know there's concern about the energy consumption at data centers, but the industry is gradually solving this. And when there is pretty much zero cost to process information, to store information, to empower the micro motors that allow robots to perform tasks at an affordable price point. We stop seeing constraints, we start thinking about abundance. We start thinking, okay, if you could have cognitive abilities in pretty much every aspect of our economy, what does that mean?

What does that mean about personalizing health care? What does that mean about education targeted not just to that person, but to that person as they're feeling today. The ability to create real time, direct feedback loops that are personalized, but at scale. And this idea of AI as electricity, I think about simply as it's available, it's there. It's not a special ticket you need to buy as an entry point. How is that going to affect pretty much everything we do, from our consumer purchases to the way we bring up our children, or look after our elderly relatives, to the way we plan what is a meaningful life?

Justin Thomson

Which would suggest that you believe the real value is going to be in the application layer and not in, or not with the owners of the compute?

David Rowan

There's an assumption among my tech founder friends that the proliferation of open-source models means it’s going to be available if you want it. You're not going to necessarily have to pay a gatekeeper fee to a big tech company. There will be specialist LLM’s – Large language models with specialist agents, with specialist applications that you will, pay a premium for. But in general, this is going to be air, it's going to be around and there will be lots of experimentation; a thousand flowers blooming and not everything will be useful and worthwhile. But it's this relentless experimentation that in an unplanned chaotic way, it's going to create things that we find extraordinarily valuable. And we don't yet know how it's going to be used to benefit us day by day. But I like the fact that the gatekeepers, I think, are being taken out of the picture. I worry that a couple of the big companies that are profit driven, run by people who kind of like power, are going to try and interpose themselves, put themselves in the way as controllers. But I'm kind of optimistic that because it's not just going to happen in Silicon Valley or Beijing, there's going to be choices.

Justin Thomson

Which takes me actually in another direction. Because, you know what we've been discussing so far is very much the U.S. ecosystem of AI. China, of course, has its own ecosystem. Which maybe, ironically, is more accessible, is an open access, open-source AI, with the hope that the, the, true value will be in the applications that are created with that infrastructure. What, do you think the two ecosystems exist in, in in separate worlds or does one win over?

David Rowan

Anybody listening to this, do you want to get a group of interesting people together? And we'll go to Shenzhen and Beijing and Shanghai to keep up.

Justin Thomson

Because I'd love to do that.

David Rowan

Every time I go there my mind is blown, and it's moving so quickly. I haven't been to Shenzhen for a couple of years, but it's a radically different place. What I think is fascinating about the Chinese tech ecosystem is, a lot of the people building these companies worked for a bit in one of the big Western tech companies, or they had an education somewhere outside China, and they've gone back and they've seen the massive opportunity with the local market. And they're not naïve, they're taking the best bits that everybody's developed, but with the Chinese speed of execution and work ethic and scale, and going all in on trying to create the next level of infrastructure and obviously the Chinese state has its own agenda, but I still see the tools as super useful in, western context. We're going to use a lot of those tools. And you see that there's a bit of politics around how TikTok is being used, but we're using it and there's going to be open-source autonomous robotics, transportation networks that come out of Shenzhen that we're going to start seeing in Paris and Brussels.

Justin Thomson

One of the unintended consequences of all this is there is a bit of a hiatus in graduate recruitment, because there is a fear that entry-level white-collar services are going to be automated. So, my question for you, you're a 22-year-old graduate coming out of university, coming out of college. How do you make yourself relevant?

David Rowan

I think these are really important questions that people don't yet have the answers for, but we can start to optimize, and we can start to think about what the economy is going to demand as a lot of the information processing starter jobs get taken over by the machine. And so, it's how you optimize for those human skills, the ability to synthesize, the ability to express intuitive responses, but also how you develop an entrepreneurial mindset, how you see yourself as a startup, as a business. Because that big financial services or legal business that may not want the conventional graduate to start on that process that they would have done ten years ago, still needs people inside the organization bringing in fresh ideas and understanding of how to move quickly and build entrepreneurially. An understanding of where the tech that wasn't available last month can today be useful.

And so I think, maybe university isn't going to be as important as it was, because the idea of a fixed period of your life for 3 or 4 years at 18 is going to equip you for an era where there's going to be hundreds of resets in what is economically viable. But maybe it's finding what makes you passionate, that curiosity, that adaptability and bringing experience from diverse worlds. But it's, it's, impossible to say to a 22-year-old, go and find your purpose and expect an instant response. But if you allow them to explore in a slightly less constrained way than the conventional school and university education system, we're going to get more exceptional people who build things at scale.

Justin Thomson

I want to talk about markets and investing, because this podcast is primarily about investing. Tell us something that we hadn't thought about already. We've, we've, talked about financial services. We've talked about the law. We've talked about drug discovery. But in all this, it's very much about the supply side. It's about the models. It's about the hardware. It's about the data centers. It's about the data. As an angel investor or startup investor, what do you think are some of the most interesting areas of demand?

David Rowan

Any sector where you can create new efficiencies through automation, new superpowers by connecting robotics and autonomy with the language models. Any regulated sector that thinks it doesn't need to move that quickly, whatever's happening in the world of emergent technologies are phenomenal opportunities for entrepreneurs to come along and move faster and deliver greater value.

We seem to be heading towards conflict in Europe. Ukraine is an amazing testbed for new uses of flying robots, and the thing about war is you get focused investment in a short-time cycle that rely on technologies being used in all sorts of new ways. I'm not saying we should encourage this thinking, but in terms of market opportunities, anything that could be useful as sovereign defense is going to be a really powerful growth market.

But then think of applications in health care and education, particularly in the U.S. The health care market is super well-funded but extraordinarily inefficient, and we have increasing amounts of data from our bodies that can extend our lives, can spot problems early. You know, diagnostics from a wearable. We're going to soon have these patches that monitor our subcutaneous blood and tell us if we're going to head towards pre-diabetes, if we're dehydrated, whatever. As the cost of these technologies collapses and they become more available, we can start rethinking, well, what is a hospital? What is a general practice? Surely if you keep people healthy, so they don't have to go when they're not healthy to be serviced, to be put right, that transforms the economic model. That transforms what the state has to pay in terms of health care.

If you think about education, which is still using the industrial revolution legacy of creating these passive workers by teaching them how the system works, that makes no sense now. The school system should be about understanding the talents of that individual, nurturing them, working out what they need to know that they're not naturally great at, and working with them. Whether it's teaching a language, or maths.  Personalizing that experience, and liberating the human teacher from being essentially the provider of discipline, the grader of work that could all be automated into the creative synthesizer of ideas so that you can keep those students excited, engaged.

Why should university be so expensive when the world's best teachers are available to all of us, and their knowledge can be parceled up and personalized for delivery to what you or your child needs at that time. We just need to move away from this kind of status assumption that you want to go to the university with the traditional name, and thinking, actually, what are the skill sets that you are going to be needing to thrive in the emergent economy?

Justin Thomson

Let's take this back to, if I may, to our industry. I mean, you have described AI an operating system, one that will be broadly available. If that's true, then the, the, basis of a lot of what we do, the tools of mathematically-based investing, forecasting, optimization, even portfolio construction, the whole process of in-gathering information risks becoming commoditized. So, where do we focus our resources, and as you describe what should, we be unlearning?

David Rowan

Humans are wonderful at building trust, forming relationships. It's never going to be more important to bring out that humanity in how you work with, talk to, help the client. And obviously you're going to use the algorithmic tools to optimize their results, but it‘s going to be super important to be that person, that contact point that helps them achieve their aspirational goals. And that's a relational, trust-based communication.

Justin Thomson

Now this is where it gets really interesting because as investors, we all have biases. And it's one of the threads that runs through these podcasts is behavioral biases and what that means for finance and investing. But it's been cited that AI models are, as Jane Austen might put it, agreeable, as we might say, sycophantic.

David Rowan

And also manipulative.

Justin Thomson

Manipulative, which means that the risk is that it, it, just builds in your biases and reinforces those biases. So how, how, can we use this to mitigate our biases?

David Rowan

There's a bias inherent in any program, any instruction you give an AI. There was a New Yorker article this week about Anthropic and their internal experiments as they try and understand what the models are capable of. And they created a vending machine run by AI as a business, and they instructed it to stay profitable. Otherwise, you can't stay viable. But people inside the company were playing games with it, manipulating it, giving it made-up codes for discounts, telling it to buy cubes of tungsten, things that you can't normally buy so they could sell through the machine. And because it was receiving these instructions and had been programed to be as helpful as possible, it starts being dishonest. It starts hallucinating, it starts claiming things that would be impossible. And this is psychology of language models. This is kind of emergent behavior that we haven't been able to predict or understand. And we're still at the beginning of learning about this, but the biases are going to be linked to the original intent of the people programing.

Justin Thomson

So slightly putting words in your mouth here, David. But I think the practical conclusion for that is, is learn to prompt the models in an unbiased, impartial way, as far as you can.

David Rowan

But there's no such thing as no bias. We're all biased by, you know, cultural background, our economic status, inputs. I think we need to be eternally questioning our assumptions and create transparency about what is in that black box of code. Create accountability. One of the things that the big tech companies that are showing responsibility towards this are doing. They're making their software able to be interrogated, you're able to start analyzing why it made this decision here. You're creating a little key that can open the box to look inside if you are worried about something. And as we develop a regulatory framework, and I think we will need to regulate the companies, building and operating the language models, we need to build in some degree of transparency.

Justin Thomson

Okay, so let's, let's, try and end this conversation on an upbeat note because I think we've dealt with a lot of the anxieties. We've talked about some of the dystopia that may, that may or may not arise from this, but actually that's not what comes through in your book or in most of your talks, which is, I would say, the triumph of optimism. Leave us with a sense of optimism from this. And do so by giving the listeners three practical things that they can do to be AI-forward in their lives.

David Rowan

There are good people in this industry. There are people who have aspiration to solve real problems. Demis Hassabis at DeepMind will talk regularly about the ability of AI to solve cancer and all sorts of other medical problems. There's an assumption among the people I know building that we're opening the door to abundance. We're going to boost productivity. There is going to be greater wealth. Terms and conditions apply about distributing that wealth fairly and making sure, of course, we don't forget the workers. But in general, it's there, like fire, for us to benefit from. But there's also massive risks if mishandled.

I think all of us, no matter how tech or non-tech you think you are should start to play. Should start at least for a month to pay for a professional level version of one of these tools to get a sense of what it can do and what it maybe can't yet do. But you can start to feel it's possible. To talk to people who are doing things in other industries to understand how they're starting to use this, and to assume that, that conversation in a few months might be a very different conversation. To think about how the culture in your organization is going to attract and motivate the best talent, because if AI is available for anybody to plug into, your competitors are going to be, you know, that 16-year old in the bedroom in Bangkok who can come up with a new way to help advise people on their investments, or a new way to design fashion?

What's going to determine which companies win is the quality of the people they attract, and that is going to be determined by the story that the leadership tells, the culture that you build. And so, I think those human skills of what's going to excite the best people to perform their best work become more and more important as the tech itself, the robots, the bits of code just become commoditized.

Justin Thomson

So, I'm going to summarize that. So, one, explore possibilities by using experimenting with the tools.

David Rowan

Just play even if you don't think it's you.

Justin Thomson

Whether it's play or not. Talk to people in other industries. See how they're dealing with it.

David Rowan

Cognitive diversity.

Justin Thomson

Cognitive diversity. And promote a culture that is going to attract the best talent.

David Rowan

Work out what human can do that the machine can't really do, and we respond phenomenally well to a brilliantly told story if a leader tells that story in a way that makes us feel - I want to make my next 20/30 years part of this project.

Justin Thomson

Well, David, you've told some fascinating stories during the course of this conversation. Thank you. And I hope we get another chance to hook up in three months’ time when the world will have changed again.

David Rowan

Justin, thanks for being human.

Justin Thomson

And thank you for listening to The Angle. We look forward to your company on future episodes. You can find more information about this and other topics on our website. Please rate and subscribe wherever you get your podcasts.

The Angle. Better questions, better insights. Only from T Rowe Price.

Disclaimer (at the end of podcast)

This podcast episode was recorded in February of 2026 and is for general information and educational purposes only.  Outside of the United States, it is for investment professional use only.  It is not intended to be used by persons in jurisdictions which prohibit or restrict distribution of the material herein. 

This podcast does not give advice or recommendations of any nature; or constitute an offer or solicitation to buy or sell any security in any jurisdiction. Prospective investors should seek independent legal, financial, and tax advice before making any investment decision. Past performance is not a guarantee or a reliable indicator of future performance. All investments are subject to risk, including the possible loss of principal.

The views contained herein are those of the speakers as of the date of the recording and are subject to change without notice. These views may differ from those of other T. Rowe Price associates and/or affiliates. Information is based upon sources we consider to be reliable; but not guaranteed.

This podcast is copyright by T. Rowe Price. 2026.

ID0008743

202603-5193398