Arthur: [00:00:00] welcome everybody to today's episode of Pitching Passionate. with great pleasure we've got Daniel Dolsteinberg today, who is a tech entrepreneur, investor, and expert in everything quantum
Just to give you a bit of a summary on
Daniel.
After a career in the early days of global derivative technology, Daniel specialized in disruptive technologies for almost three decades, founding several companies in the space. He created one of the first global standards for digital rights management, which helped transform software and gaming industries, and in 2017 was the first co-founder of the Atari Token Project, a groundbreaking future of work metaverse project with Atari, an Ethereum founder, and several gaming legends.
He is now co-founder of Eden Base an Ecosystem and Fund, investing in quantum technology, as well as being the author of Unsupervised, which is all around technology which is ranked one of the best books by Bloomberg , in the year of 2023. Daniel, [00:01:00] you've invested in Quantinium
one of the most exciting players in quantum computing space in November 2025
What is quantum computing?
Well-
And why on earth are people talking about it?
Daniel: Expert's a really interesting term. I wouldn't class myself as an expert in any technology. I understand how quantum fits into the landscape and, you talked about my early career, and my early career is actually dotted with doing lots and lots of things. ... And I think that's one of the most important things that we have to accept at the moment, which is the world is changing very fast, and being siloed in areas or siloed in technologies, I think
means that you miss the entire picture. And one of the reasons I wrote the book Unsupervised was actually because in 2020 when we were in lockdown, one of my co-founders of Eden Bay, said, "Will you write a book?" And I was like, "I don't wanna write a book." But I was
seized the idea that the world was on the cusp of enormous [00:02:00] change.
And change had been coming on for a long time, but the exponential nature of that change meant that we were in for a completely different future
Quantum is a kind of computing. It is based on quantum physics. The history quantum physics is about a hundred years ago a bunch of very smart guys came up with this new observation in how subatomic particles interacted, and they created quantum physics.
And they built all very equations that explained deeper than Newtonian physics, how the universe and how , the world operated. And up until
the 1970s,
they carried on building this. And then in the 1970s, a very smart physicist called Richard Feynman suggested that you could build a computer based on these subatomic particles.
So rather than being binary, they were actually particles that were probabilistic in nature.
And in the '80s, IBM [00:03:00] started trying to build these machines, and nothing really happened until around 2023, 2024, which was they solved a lot of the science, but they have anything you you could- recently as that
that. They didn't have a work, a you could classify in any phase of working. But in 2023, people started, like Quantinuum, were able to build the first generations of these compute. And so that is
Daniel: difference. Instead of working on a binary system, ones and zeros, which is what computers do, they work on probabilistic natures.
So they at all combinations of the answer at once, and they're two things only. One is data matching, and traditional computers are very bad at matching because it's a factorial problem. If you have five people sitting around a table, it's a hundred and twenty-five combinations. If you have ten, it's three million combinations.
If you have twenty-five, it's one with twenty-five zeros. So for a traditional computer to compute one with - different options takes longer than the [00:04:00] universe has existed. The thing because universe is, Probabilistic in nature.
To map a probabilistic system like biology or materials
onto ones and zeros is like taking a vinyl and putting it on a CD.
It sounds okay, but it's not the original. And when you are very complicated like how does a drug interact with the human body? How does a material operate? How does photosynthesis work? With a binary system, we can only approximate it. With a quantum computer, in theory, 'cause no one's actually done it yet- we can model those sort of things. And very important because that is frontier for humanity. The frontier for humanity is understanding the things we don't understand. AI is very at looking at finding matches in things that we do understand, and in odd case it finds something [00:05:00] that we understand by fluke.
But this is able to compute things that we can't do. Most people when they say, "What's a quantum computer?" They say, "Is it a faster computer?" It's not a faster computer. It's just a different computer. And then they say will it reduce the number of computers we've got? Will it reduce the number of data centers we've got?"
it'll do the opposite. And think a very good analogy might be imagine lived in a world where you had refrigerated containers, trucks. So within UK we could transfer anything up and down the motorways, and across Europe we could do it, but couldn't go from Europe to the UK, and we couldn't go from Europe to China.
And then today someone invents the refrigerated container ship. Suddenly can something new, which is you use a container ship from going from London to Liverpool. You'd use a truck still. But you could go from Liverpool to York, or you go from New to Beijing. What does that mean?
It [00:06:00] means you've now got a bigger operating sphere-
therefore you need more trucks
because the more container ships, you're gonna need more trucks. So you've got now a global industry. And whilst quantum computer is not a container ship, you've gotta look at it like that, which is it does something can't do.
The results from that are then to be computed by traditional computers. If you wanna do anything that's process-oriented, anything that's in the current sphere of what we do, calculations computers are gonna be much more effective than a quantum
Fascinating.
when, everyone's always talking about AI these days, and with good reason.
It's crawling and reaching into so many pockets of life with dull things, but also fun and exciting things. What maybe people don't know, . we've been working on AI s- since the '70s, '
Arthur: 70s, right? '
50s probably.
Daniel: Probably. Okay. Now with quantum, where are we at, and w- what are the hurdles to get to actually attaining value from this
Arthur: this
Daniel: technology, this concept
Firstly, let's go to AI. AI is a misleading term because what we're building today is not in- artificial, it's not intelligent. It's something else. [00:07:00] It's very real. And it's very sophisticated and effective at doing things mainly based on the data that we feed it.
So could it be a better lawyer than us? Could it be a better accountant than us? Almost certainly. Could it be a better management consultant? Could it be
better at
anyth- a banker- ...
trader?
Yes. Could it be a better human? No, because it's not a human. It's like a tractor. When they created the tractor, it was a better horse doing what we needed it to do in a farm.
But it a better horse? No, it's absolutely not a better horse. like a horse. Did mean horses disappeared in farming? Almost exclusively because it was better at what we created it for. So whilst an LLM, and let's say what AI is at the moment, it's large language models is that a better accountant as a person?
No. But is it a better person who can do [00:08:00] accounting? It's a better, it's better at doing accounting, yes. Absolutely. And does mean it doesn't make mistakes? No. But then humans make mistakes, too. So opportunity for LLMs to rewrite the world is enormous. But is it is there route I for what we've built today to become intelligent?
Almost certainly not. So that why when you say rightly, we've been building these machines for a long time, they are an exponential of we've been building for a very long time. When we had Microsoft Word, we were able to write documents, and it allowed people to put their ideas down more people than on typewriters and create many more things and collaborate and so human knowledge moved on.
With Excel, the same for numbers. I think you probably look at LLMs in same way, it's a continuation of that process. And did Word allow that weren't traditional authors to become authors and [00:09:00] write better books than traditional authors who were good typing?
Probably, yes. So they change the framework. Then onto your quantum question, which is where are we now? So- To
To build a
quantum computer, you need to have something called a qubit. A qubit is a bit that works in a quantum state. To have an effective quantum computer, you have to have two of them, we have reached that stage. To do things are very on quantum, you minimum 100, more like 400,000, et cetera. Where are we today? As I said between the 1970s or 1980s and 2023,
2023,
we moved from zero. There's a physical qubit which works, and there's a logical qubit which is we're able to use it in effect.
Okay. In
2023,
Arthur: 2024, we reached the threshold where people were able to build a single, I would call it workable qubit rather than logical, 'cause it works - much of the time, but not all of the time.
Daniel: [00:10:00] By the end
Arthur: last year, we'd reached 50. So we, again, always asymptotic. So most companies that
Are professional and
are in the space predict that somewhere between '27 and '29 we will reach 100. Many of them predict that by 2030 to 2032 we will be reaching the early That's when it becomes extremely meaningful, the early thousands. -
.
Sorry, can you just run through the numbers
So
Daniel: let's say
Arthur: beginning of '24, one.
Yeah. Middle, end of '24 10. End of '25, 50. So on that basis- ... they're talking about 100. Now, there are many ways of creating a quantum computer. There's a neutral atom, there's trapped ion, there's topological, there's supercooling. A little like electricity. So in electricity, we have different ways of making electricity.
We have solar, we have water, we
have
natural [00:11:00] gas, we have oil
All of those are different ways of doing it, but in effect they do the same thing, which is they create qubits.
Daniel: And then once got the qubit, you have to connect them to the, a grid and start using them. little like electricity, it doesn't matter how your electricity reaches your house, you still have to have the wires- and have the switches, and you have to have transformers. And sometimes that electricity might be coming in from the gas, it might be coming from wind, it might be coming in from solar- Yeah ... it might be
Arthur: in from the waves. That's a,
such a, a great analogy there.
Daniel: And
so what we have to build next.
we have to build next is once people have working m- quantum machines, how do connect together? How do you connect them the grid of compute? How connect
them to
Arthur: GPUs, which is what we use to run AI? How do you connect them to CPUs? How do you store data that comes out of a quantum computer?
We don't know the answers to those yet, which is why it's such an interesting area, because we have to build an entirely [00:12:00] new
Ecosystem to be able to operate these systems on. And so that is why it's such a fascinating business in terms of investing, 'cause it's like the early days of electricity. It's not just the production that's important, it's every component in the chain.
And we're still working on it, I saw a company the other day that I use the word claim because I haven't, we haven't due diligenced it, but the claims that they m- they're moving towards superconductivity. We haven't got to superconductivity yet for electricity. Yeah.
Daniel: Yeah. That will be, that will move
humanity on
incomparably. we're still working on those things, as we likely to be working on quantum computing and GPUs in 40, 50 years and discovering new things that are gonna make them more interesting.
Arthur: And,
in investing in this space, obviously, you it's about making money,
Returning investors, cetera, cetera.
How do you see the timeline
playing out in terms of the,
Financial rewards back to the investors space? [00:13:00] And, o- obviously new technologies, you have all sorts of people investing in different at different stages and early players, people moving markets and capturing value such different ways.
How do you see the coming back and, or, to which particular parties of the technology do see reaping the most rewards from that investment?
Look, that's a really fascinating question, Arthur, and I'd probably like to widen it a little bit, which is, as you mentioned in my bio I created two companies, one in the internet era and one in the blockchain I didn't create anything in the AI era, though I did invest, and I wrote my book that was my work in the AI era And AI forms a tiny part of where I see These technologies going. One of the reasons I think it was a, a Bloomberg book was that it was very focused on AI at the right time.
But the other technologies are equally important. So that's the first thing. The second thing is I have seen so far three [00:14:00] technological shifts. Technological shifts are gonna be very different from now on and the book talks about it and
we can talk about that later if you want. But what I've noticed in those shifts, the first one was pre '94. I was working in banking, a young kid. And really fascinated by the internet, and in '97 I left banking and formed an internet company. But if you were investing in the base layers of the internet, which was the routers and the fibers and the browsers, when was the best time to invest?
Pre '94. When was the best time to exit? 1999. . Five, years later. I got involved in blockchain in 2016. When I exited my first company I started investing. And in 2014, a lot of people came to me and offered me these Bitcoin businesses, and I didn't get it. I understood the technology, but I didn't understand where it was gonna go.
By 2016, I understood where it was going to go.
So I invested heavily in blockchain, , in crypto in 2016, and I [00:15:00] founded a company at the end of 2016, which was a metaverse company. And the best time to hold crypto was 2016 to 2022. I exited almost all my holdings in 2022. Six Internet AI and tech. 2019, I became fascinated by AI. Not because I hadn't seen it.
I wanted to invest in AI business in 2012. I just- ... couldn't get involved though. Couldn't see the outcome. 2019 it became obvious. When's the best time to exit internet AI businesses? I would argue today Not for the long term, but as an investor. OpenAI was worth, what? A billion dollars in 2019.
Today it's, what? A trillion? 800? It's gonna be a long time before you get an 800x again- yeah ... in this space. You will get it but not that. So exactly the same quantum.
Daniel: In
Arthur: 2020, one of my partners, Eric, came to me and said, "Do you wanna invest in this cool company, Cambridge Quantum?
I'm raising money for them." And I was like, "Cool. I understand technology, but I can't see a horizon." $400 million [00:16:00] I was like, "I can't see a horizon on this." I passed on it. By 2022, I had written a financial model for one of the quantum companies when it went onto Nasdaq, and I was starting to see timelines by '23.
By the time I invested in Cambridge Quantum, which became Quantinuum, it was nine billion. It'd gone from four. Yeah.
Daniel: Yeah.
Arthur: But that's the same thing. So- By 2025, it was clear to me what the time horiz-- that I could see that six-year window. It's 2026 to 2032. That is the time horizon, and it tends to follow those same cycles.
So the way that we invest in Edenbase is we have a six-year fund. Why do we have a six-year fund? Because we wanna be disciplined on finding the companies that have
...low value today
and high value in 2030, 2031, 2032, maybe a couple will go into 2033. So it's that discipline rather than looking for companies that may be valuable in 2038, 2039.
I talk to
Daniel: a lot
Arthur: [00:17:00] people that have AI businesses now that are looking at getting into quantum. I said, "That's the right thing to do because in 2028, '29, there may be no money for AI 'cause it's the AI winter or whatever. You wanna be where the next big thing is." So that's very much how we look at it and where we tend to invest.
And why
Daniel: Eden
Arthur: Edenbase is really useful for me is, I mentioned to you I'm very good at seeing signals and understanding where they go. I'm less excited by staying in fields for a long period, and that's what Edenbase does. We have a quantum team. We have a phenomenal partner in the quantum fund, introduced to us by one of the founders of Quantinuum.
And that team
Daniel: has mapped out
Arthur: where the value lies in the quantum ecosystem very effectively. In fact, the National Wealth Fund of the UK uses our thesis map. And then what we do is we look for where the value is going to accrue within that time period, and where we mainly see it is in bottlenecks.
So what will happen, it was very [00:18:00] similar to the internet, I remember there was a company called Vermeer, that became Front Page. They developed a product called Front Page, and Microsoft bought it for $100 million. They'd sold 23 copies or 40 copies. They bought it for a six-month lead Because they needed a tool to be able to write websites that they wanted to have Microsoft.
That's a bottleneck. Bottlenecks are very valuable because they allow the bigger companies to accelerate to the next step. And so what we do here is we invest in some modalities. Modalities are the way that you create the qubit. But mainly we invest in things like - quantum memory, because we know when you start building these quantum machines, you're gonna need quantum memory, but no one's really focused on quantum memory at the moment because most people are trying to build the machines at the moment.
So that's what we do. The other thing that I think is really important now, and makes , the world very complex, is - these system technologies [00:19:00] we're building today cannot be looked in isolation. So I, as I mentioned to you, I was in blockchain in
2016 through
2019.
I left the industry in 2019 and exited everything in '22. I was on a panel in Palo Alto, Silicon Valley, in February 2022,
Daniel: and
Arthur: said to the blockchain people in the room... a very serious crowd. It wasn't the sort of, you know, the people you were getting in 2016. Very serious. And I said to them, maybe the blockchain industry should be looking at data protocols for the AI industry."
I was shouted off the panel. This was six months before the launch of ChatGPT. Because people aren't seeing... They're siloed in their domain, so they don't wanna accept that AI might be coming, and you can't do that anymore. Why do I think... If you ever read my book, you'll see that I was very bullish on AI in 2020.
I said, by 2025, we will definitely have LLMs. Definitely And 2023 was the big year.... But quantum, I couldn't put a date on. I said, [00:20:00] "You need 100 logical qubits," but I had- couldn't put a date. The reason it came so quickly is because the LLMs we built have enabled us to solve these problems that were impossible to solve.
What they're also doing is they are bringing forward massively the date at which they're usable, because programming for quantum computers is incredibly hard. But LLMs are
Daniel: proving really good
Arthur: writing algorithms and programming. So suddenly you're increasing the speed at which the qubits go up, and you're reducing massively the n- number of
Daniel: qubits that you need to operate e- problems, because AI is scrunching that space.
And then you s- all the technologies we're building, biocomputing, putting compute onto biomass, spatial technologies, which is clear what I think Mark Zuckerberg is probably really building, which is the interface layer between compute and humans. These things are gonna... [00:21:00] Each one of them gets added to the other one and compresses what you've got.
And so this makes investing very hard at the moment, because if
you
that you
Arthur: be a quantum specialist investing in quantum for the next 20 years, I think that's probably an error, because there will be something new that comes along which is a better opportunity to be investing in. Plus other technologies will reduce the value of the things that we've got.
..
Arthur: And the final thing to talk about this the whole investing and the Edenbase thing is, we are building a new future. And so the wealth and power of the world is moving
Daniel: at a
a hugely fast rate from an industrial economy into a tech economy.
The, industrial economy is finished. We have to accept It doesn't exist. Like the agricultural economy was finished by the [00:22:00] 1850s, it was the industrial economy. So you still needed it, it just had no value. That is happening now. And so y-
you can be either
in an industrial economy or you can be investing in a tech economy.
Quantum technology,
how will the world change
Over the next 10 years?
Arthur: Look, that's a really hard question. It's a really good question, Arthur, but it's a really hard one because we don't know what they're gonna do. . When I wrote the book Unsupervised, it was pretty clear what LLMs could do by that stage.
And
I think if you ever read the book
If you read it today it feels like it was written today. It's very up-to-date, even though it was written five years ago,. the problem with quantum is we don't know what the two things that I think we can be confident in is a hyper-optimized world, because if you can do matching, you can optimize the world.
So optimization is really important to food production those areas where you need [00:23:00] hyper-optimization will be their portfolio management. I don't know what that does. If everybody has a hyper-optimized portfolio, does that mean everyone's investment goes up at the growth rate of the economy?
Things like that might happen. The second thing is, It opens up for us the world of discovery, so new materials, how photosynthesis works, how weather patterns work, what temperatures are going to be like in the future, 'cause these are all
Daniel: problematic systems.
Weather, for example,
Arthur: whether- Weather ... in the UK, I think we've got
Daniel: we- in the UK, I think we've got one of some of the best weather forecasts in the world. But yet it's so hard to predict, right?
Arthur: And it's important because, farming, mining, all these things, logistics all depend on weather. But more importantly, imagine we can start discovering materials that can sequester, carbon out of the atmosphere or, can create super light airplanes or, all these things are open.
The biology is probably the, the really big one. Being able to discover drugs that are [00:24:00]individualized rather than generic. Your DNA makeup and mine are probably relatively similar, but they're different.
If you were to take a drug saying to you, " for you taking an eighth of an aspirin eight times a day every three hours is better, whereas for me, I should be taking one a day.
That type of thing, that sounds really ridiculous, but those sort of things, or discovering the, that drug in your body will mean you'll never have a heart attack. Again, these are examples. I think we are just about to start opening up that world, and that's really important.
Daniel: Sometimes
one can, fall into conversations or in energy or over someone's absolute fear with
technology and wh- and
where we're at with the world, what, people talk about a conflict's going on, or maybe people talk about,
L- living standards or c- cost
Arthur: of living or employment,
Daniel: What do you say to the people who are like, "Oh, the world's getting worse"?
two answers
Arthur: to that. On the positives, technology [00:25:00] has given us the best decade ever for humanity.
We've got the greatest life expectancy, the best health. It's not a straight line across the whole world. Some places have got worse, some have got better. But it's also made some things worse. Social media certainly has made a lot of things worse, even though it's made a lot of things better.
depression. There's a lot of the anxious generation. We have all this going on. The, the other side of it, I think, is really worrying. And one of my roles between, in the sort of 2007, I think, to 2012, is I was the second representative from the British government to the European Commission with the chairman of Rolls-Royce.
And I was very fascinated by a few things. One was education, and we were talking about that earlier. But the other was future of work. What is work gonna look like? One of the arguments I would always make to the European Union is, you want the average salary here to be less than 30 euros an hour, but there are only 500 million [00:26:00] Europeans.
There are eight billion people in the world. So unless you are creating real value, eventually, if there are eight billion people, , the average salary is much lower. So I'm fascinated by future of work, but I've always found lawmakers have been totally disinterested in future of work because it always extends beyond their time in office.
It's always a problem for someone else. And solving these things require really hard work and dedicated work, and you have to be honest with people. And I think we're falling into a huge trap. And we earlier discussed LLMs are not intelligent. I don't see there's a path to intelligence.
But are they a better lawyer? Are they a better accountant, management consultant, trader, doctor? Probably already. Two years' time, almost certainly better. By twenty thirty, [00:27:00] multiple times which means where do you go to get your best advice from a lawyer or an accountant or a doctor?
You'll go to an AI. And it won't even be that. These technology companies are d- are building really cool stuff. They will be agentic AI. So they will be monitoring you already. So they won't say, "Arthur." You won't say to it I got my heart feels a bit funny, what should I do?"
It'll say, "Wait a minute. I'm monitoring you, and you've got a problem. Stop eating potatoes or whatever, or stop drinking so much, or go and see a doctor." It's going to
Daniel: be very
Arthur: reactive, not proactive for humans. So what does that mean? If you look at a country like the UK, and I don't wanna be scary, but y- look at what we're in.
We've got the top one percent of taxpayers pay what? Twenty-five percent?
Daniel: Many
Arthur: them have left already. The next twenty-five percent pay, I think, seventy-five percent of tax. That's the accountants, the lawyers, the management consultants. What [00:28:00] happens in twenty thirty, twenty twenty-eight when they don't exist?
The tax take just
Daniel: collapses.
Arthur: And yes, we have hundreds of thousands, if not millions of people not having a job. I was at an event - this week on geopolitics. Volkswagen are considering laying off 100,000 people. Microsoft have
Daniel: 100,000 people. Microsoft have
Arthur: already laid
Daniel: off 30,000
Arthur: off 30,000 people.
Meta are laying people off. These aren't companies that are losing money, Meta and Microsoft. . Meta's profit's got 23% a year, are laying people off. Really? Yeah. So just to finish off we do have to consider this very real question about what is the job of the future? And we can't leave it till after it happens because there won't be money to pay for the services to keep the country running while we go through that phase.
And so what... where these people are right- wrong, I think, is saying technology will come up and kill us," 'cause I don't think that's likely. [00:29:00] Where they're saying the technology companies are taking all the value out of the equation, and okay, so we can regulate them. Zuckerberg and Musk don't care what Burnham thinks on regulation.
They only care what Trump thinks on regulation. And the Chinese companies care what President Xi thinks. But China and US are in a battle for survival against each other on building the best technology. So they're not gonna stop because someone says, "Please stop building this."
So there is a real element of worry that we have to start to consider. What is an education system of the future? What are we going to have people doing?
Daniel: And we need to do it today. Yeah. But if it, a- again, I think the challenge when you talk about technology and, even quantum, what...
you can easily fall into that trap of just looking down this isolated channel, right? And forgetting what's happening outside of that. But
Jumping into, for example, education, you've got
Arthur: all these
Daniel: academic institutions that are, are funded through people studying these programs that...
the [00:30:00] value... A l- a lot of programs that, people are questioning the value of studying some
Arthur: of them.
Daniel: But yet, - the universities are set up, their cost basis, they need the, this volume. But there's tremendous excitement with, how education...
already it's been distributed in an incredible way, but, the ways in which people can learn in a personalized way and,
You can
connect in a far more social way. Like, how do you... let's take education for a field. y- o- optimistic view on education with all of this, happening, how do you see structures changing?
again,
another g-
Arthur: g-really good question, but it's another question that's very difficult to answer because of what you've just said about the complications of the industry. In the UK, you've got teachers unions, you've got an entire ... public sector environment, which is very loathe to change.
You've got an ideological education system often. And even though these technologies are really good, implementing them is very difficult. And then you've got what you've just said, which is [00:31:00]these technologies can help with education and need to help with education, but that just makes the system more efficient, and more efficiency normally means less people in it.
And then you've got the whole social side of education. I read a piece that I was very disappointed by because it was written as a sort of such a positive, which is China has got rid of a load of their subjects they teach at university to focus on AI and robotics. And all the things they were getting rid of is the humanities.
I didn't study humanities, as it happens, but the humanities are what makes us human
Daniel: And if you're assuming that
Arthur: robotics, great, start to be an AI engineer or be a robotics engineer. But in the next thr-three or four years, the AI will be a better AI engineer and be able to build better robots than you.
What do we have to fall back to? We have to be, what is it to be human? And so I think that's the wrong [00:32:00] course. Is it better that we actually say to the UK, "Yes, we want a lot of engineers, and we want people to create cool companies and whatever," but we should actually be teaching people more humanities- to accept that being human is very important.
Daniel: It's, not an easy one to figure out how the world is gonna look like in, in 30
years.
h- who
do you think are the brightest minds that really see that
Arthur: changing? Look, a lot of people write a lot of stuff about it. It's normally biased. It's normally with, a point of view. I haven't come across many that are thinking about the problem as holistically as they should be thinking.
They're either pro-AI or anti-AI, and then they're pro,
Daniel: or
Arthur: Socialism, or they're pro-capitalism,
Daniel: or they're pro capitalism, they're
Arthur: or they're that or they are the other. They have a background in this or that, and that's going to make it very difficult. The other thing is we're not homo- homogeneous.
I think one of the things that we've fallen into a [00:33:00] trap of,
As humans, and we always do, humans always do this, they assume that the environment and location they're living in has always been there, and therefore they speak for everybody else. That's just not the case, and we can see that particularly now, where historical figures are being judged ...
by the time and place of the person making the observation in the 21st century.
Daniel: just not a
Arthur: thing to do. But we have that today, which is, you know- To, for me to build a, an LLM now, I have to train it on some data. I have to give it access to some data. I have to give it my rules for how to deal with that data, and then I have to give it guardrails around to say, "You can answer that and you can't answer this."
Daniel: That means that it's
an embodiment of me,
Arthur: in effect. And an embodiment of me doesn't work so well for a woman, and an
Daniel: embodiment of me
Arthur: work so well for a Frenchman [00:34:00] or a German or a Saudi Arabian or an Indian. So we have to be... We have to look at it not just, as a leading thinker for a certain time and place.
We have to be looking at it, how
Daniel: do we
Arthur: allow it to be a much broader solution? One of the
Daniel: reasons that Eden
Arthur: existed, and we have this hub here called Q Base my, one of my partners created Level 39 and Tech Nation, which was a very porous ecosystems that allowed a lot of knowledge and great wealth to be created.
Daniel: Because
Arthur: we want
Daniel: more people
Arthur: come in and understand what's going on and have their own viewpoint on what to deliver. Because otherwise you get into that state where just 'cause I think it's good for me, it's gonna be good for someone
Daniel: In the, most basic terms, quantum gives us answers that we currently
Arthur: can't get
Daniel: With traditional computing technology.
It's so obvious that we struggle
Arthur: to,
Daniel: ultimately choose the right path whether it be for, policy, i- in way, [00:35:00] in every area. And
The- there's such a constraint in
human intelligence
as well as making the right decisions and even having the right, data sets and be able to interpret them.
But even, processing information, for example you hear about the courts in the UK having such a big backlog. Surely quantum could really point us to having, a far better world in terms of working out what policies make sense how value can be attributed
Arthur: to
Daniel: people in a fair way, resource allocation - security across all of these
different areas.
Arthur: Look, again, very good point. I'm generally an optimist. I'm not a pessimist. I wouldn't be building and investing if I were a pessimist. Technology has
Daniel: improved
Arthur: the lives , of most people across the world enormously, and there's nothing to expect that doesn't continue.
But there might be a lag between where we are today and where we are in the future, and that lag can be very difficult. I don't know how quantum can necessarily help with the court systems.
Daniel: AI
Arthur: AI could [00:36:00] possibly help with the court systems, but to try and sell to the public, it's diff- it, the, the Labour government wanted to get rid of jury trials, which I think is a very bad idea.
But trying to sell that as it'll speed it up." Yeah, but what happens if I end up in front of a judge that doesn't like me, and I don't have a opportun... we can see the downside. If you then said wait a minute, we're gonna do all, every court case is gonna be by an AI and not a judge," I think people won't accept that.
And I think quite rightly they won't accept it because how, where's the testing? Where's the this? Have you run it on every court case in the last 50 years and made sure that the, it matched
Daniel: at least
Arthur: we've got today? These are very complicated things.
[00:37:00]
Arthur: Quantum unlocks a new data layer for us. It goes from binary to probabilistic, and because the universe is probabilistic, our biology is probabilistic, our materials are probabilistic it's allowing us to do that. It doesn't necessarily give you better answers for a lawyer 'cause a lawyer wants to look, what's the case law?
What's the contract say? What's this? The probabilistic side of it is not that interesting to them. It's Here's a case, it's between A and B. Can I find a ruling within the last 20 years that proves that B is right 'cause I'm representing B? Yes, I think AI is really strong at that, but throwing in a probabilistic argument is not helpful.
As I say, what, what quantum does is it allows us like a magnifying glass to look deeper into a problem. [00:38:00] At the moment, when we look at a biological problem, we are translating it into ones and zeros so a computer can run it, like a CD versus a vinyl. This is letting us look at the original vinyl and coming up with an answer.
That's more valuable. What I suspect will happen is we will look at the vinyl, and then in 10 years' time, or 20, or 30, or 40, people will realize that the vinyl is just a replication of the orchestra,
and the orchestra was completely different from the vinyl because actually what you enjoyed about the music was sitting in the hall with other people.
So how do you go from the vinyl, which is just a perfect representation of the music level, but not a perfect representation of the environment in which you were doing? And I'm sure you know this. You go to a concert, some of the concerts you go to are insane. The feeling in the audience is incredible.
You don't get that from , a vinyl recording I went to see recently My Chemical Romance. It was really cool. I went to see Oasis last year. There was something really special about those events [00:39:00]because of the way that they were and
Daniel: the people were
Arthur: to them. The recording of that, even a perfect recording, probably only captures, what, five percent of what I felt in that
Daniel: that stadium.
Arthur: And so quantum is like that. We will go from CD to vinyl, and then we'll realize vinyl is that... I don't know when you were young, I'm sure when you were young, you did the same as me, which is read about body language
You know, computers aren't doing body language yet.
They're just doing the talking. Then when we get to body language, there are all the things like pheromones. There's energy. There are millions of data points that we're not even aware of that we're communicating with each other. And so those data points are not so useful in a court of law necessarily.
They might be if it was a jury trial but what you say to each other is the only important thing, 'cause that's what the judgment is made on. So computer's fine. If we're doing, a, a biology experiment, it might be fine. If we're doing something beyond that, we might need something completely different.
And so we [00:40:00] really need to consider that every technology we have been able to produce so far has just allowed us to go layers and layers down into what the final truth is.
Daniel: Investing in technology and
setting up these businesses and teaches, e-endless learning
what has it taught you about
people?
Arthur: Wow. Look, I think . i've had, many roles.
Daniel: The way you in, the way you operate
Arthur: different roles is completely different. When I built derivative trading technology, I was a very junior person. I was-
Daniel: ... Good at
Arthur: it, but I did this when I was saving a bank, I did it differently.
When I ran my first business, I was a very different person and I operated completely differently. When I did my second one, I was very different, and now I take all those things together and
Daniel: I, again,
Arthur: a different person.
Daniel: What
Arthur: what I've really liked in my life is I've
Daniel: been able to
Arthur: be exposed to really fascinating and I tend to surround myself with [00:41:00] those people.
And they tend to think very differently from me. They tend to be completely different. So that I find interesting. And so what I've learned from people is there are certain people that really appeal , and spending time with them and
Daniel: talking
Arthur: them
Daniel: and
Arthur: understanding different perspectives is really important.
In terms of investing, if you're looking at it really specifically in investing,
I think from when I started my first business, which was in '97 what I was looking for back then or what I needed is a completely different person to what you would look for t-today. There are some feature sets that are necessary. Resilience, ability to sell, ability to have teams with you.
Back in those days, you could have a five to 10-year vision and plan of where you wanted to get to and spend the time getting there. I don't think we have that luxury anymore. The speed of acce- acceleration of these technologies and the way that humanity is [00:42:00] reacting means that you have to be building things very quickly.
And so in terms of the question, 'cause it was quite a very broad question which is why it may take, takes me time and I'm only s- answering a tiny bit of it, which is
Daniel: We're looking for people
Arthur: that are much more flexible today, I think, than we are.
Daniel: I really
Arthur: dislike the term tech for good. I have a lot of...
Some friends, I invest in the, the odd tech for good fund, and I really like the people behind them. I think that's a flawed business model. I think a, to decide whether a technology is going to be good when you invest in it is the wrong way around. I believe and the reason Eden Base and Q Base exists is you wanna be working with people who
Daniel: accept
Arthur: that there are different opinions to theirs, and at least consider that as they're building.
And when they s-start to- have actions that have an impact, to assess those actions [00:43:00] about w-how that is operating. I think that's really important. One of the reasons and we discussed this at length earlier, which is, what I think is a good idea is not a good idea for some of the customer bases that I'm working for.
How do I ameliorate those sort of problems? One of the things that I've been doing for the last three years is I draw two images every single day. One is an an- a tiger and one is a dragon, and I do... on AI. And the reason I do it is to see the guardrails of what it lets you do,
Daniel: And it's really interesting
Arthur: to see that.
You can really get into the heads of the people that are building these codes 'cause of what they let you draw. With text it's very difficult 'cause you can't... but drawing. And
Daniel: the reason I wanna do that
Arthur: is I've been very concerned, particularly since writing the book, that these technologies are changing humanity to a much more homogenous set of people.
And I don't think homogeneity is a good idea. The reason I went on that tangent is, it's back to the people I like to spend time with. I [00:44:00] like to spend time with people that are very different from me,
Daniel: generally
Arthur: very bright, and have completely different experiences from me. And I spend much less time with people that think in a similar way to me or a similar background because that does not add to my knowledge.
And that doesn't mean that I don't get upset when I hear things that I don't like. It's just that that is really important. But investing is the same, the best people create the best companies and the best technologies. I don't think we can... I don't, I haven't changed my mindset if you have a really good idea, but you're not a very good operator, and executor, and salesman you're gonna be successful.
It-- That side of it far outweighs , the idea
Daniel: The skills of today to do well, probably have changed,
Compared
Arthur: to 20 years ago.
Daniel: It seems with AI, being able to plug into Claude, there's just so much endless opportunity, and the limit seems to fall partly on people's
Arthur: imagination.
Daniel: Do you have a view on
Arthur: [00:45:00] that? Again, look, , it comes down to the person. One of my concerns with the sort of statements you hear about future of work everybody will just become an entrepreneur, or everyone will start their own business, or everybody will become a creator or an artist or whatever.
Daniel: The way we've
Arthur: been educated and the way that we've been brought up and the way that we've, operated doesn't necessarily mean that we can pivot.
So no I'm not sure that just an access to a
Daniel: tool like that suddenly
brings
Arthur: out the creative element in everybody.
Daniel: I think for people that are creative in a
Arthur: certain way, they're very good. For me, the the creativity is great. I'm a creative person. I can draw okay, but I'm not an artist.
So
Daniel: for me
Arthur: it's brilliant because I can explain what I want and it can show me a visualization of what I want, and I say, "No, that's not right." But for an artist, that is probably the worst thing they can do, 'cause an [00:46:00] artist is all about expression
Daniel: through the medium of what they're using,
Arthur: a paintbrush or a, a pen.
And some artists are great with paintbrushes and
Daniel: not very good with
Arthur: pencils or with sculpture. That's their expression level. So
Daniel: like all these
Arthur: things, it is a phenomenal tool,
Daniel: probably a phenomenal tool for most people, but you
Arthur: have to
Daniel: find the way in which the tool is useful ...to
Arthur: that person. No. I don't think they improve creativity.
The other thing they don't do is, when we wrote our first book we wrote it without AI. It was a book on AI, and we wrote it completely , without AI 'cause it wasn't good enough back The second book I'm now writing with my co-author, we used AI enormously. And
Daniel: - For me,
Arthur: brainstorming with AI was phenomenally useful, but it had this terrible side effect, which is it expanded the ability of my thought so fast that we ended up with [00:47:00] maybe a thousand points to put in the book as opposed to a hundred.
It was overwhelming. When it came to writing it, even though it
Daniel: sounded
Arthur: very effective, the writing,
Daniel: it was
Arthur: actually meaningless when you read it a few times. It didn't actually say anything. And so there is a veneer of-
Daniel: , Value.
Arthur: And so for an author, it's a terrible tool. For someone that
Daniel: is not an
Arthur: author but wants to
Daniel: ideate and express small
Arthur: paragraphs of ideas...
what it was very good for me is my co-author is a very different person from me. We've been friends many years. He found it very difficult to... When I wrote pieces for him to what I was thinking, he couldn't understand it. When I talked to him, he could. What we found did work is when I talked to a device- ...and then had the AI write it in a style that he would understand, works brilliantly, but that's a paragraph.
And so for every area,
Daniel: it is
Arthur: very good for [00:48:00] some things, and it's very bad for other things, and we have to understand that. We also have to understand that where we are today is not where we're gonna be in two years' time or three years or four years. So You know, it used to alarm me massively in 2023 when people said you should become a prompt engineer."
I'm saying, "Why?" Maybe for 2023 and '24. Nowadays when I want a prompt, I ask it what's the best prompt for it.
yeah.
And then I read that. W- programming, end of code is not about
Daniel: no more programming,
Arthur: it's about no more code. These machines will not have rules. They will operate in their own environment.
It's a very important distinction. But if we don't realize that is coming, we are try... And this is why your questions are so good but also so difficult to answer, because we're trying to put a map. A bit like if you said to me,
"How do you operate in Saudi [00:49:00] Arabia?" And I tell you how you operate in the UK and expect you just to walk into Saudi Arabia and it will be fine.
It's not. And so
Daniel: you're asking me, and these
Arthur: are brilliant questions, but you're asking me to say how do we operate today, and how is that gonna fit in a world that is gonna be very
Daniel: different?" It's
Arthur: up to the pe- It's... and the reason I wrote the book and the reason I got it published was
Daniel: it's not about me
Arthur: trying to define what it is for you.
Daniel: It's for me
Arthur: to try and explain to you what I think is coming, and for you to make the the judgment. See- I love the difference. There's a saying, I think it's a Jewish saying actually, what's the difference between money and knowledge? Which is if I have $1 and give it to you, you've got $1 and I don't have it, but there's $1.
If I give you, Arthur, I say to you, "This is a really interesting - fact," and you put it into your brain and go away and use it, then there's double the knowledge or triple or quadruple because you are using it in the way that serves you. And I think that's the same with quantum, it's the same with AI, it's the same with every single technology, it's the same with countries.
Daniel: [00:50:00] They need to be
Arthur: creating an environment where people are not afraid to ask the question,
Daniel: and the
Arthur: more they shut down the ability to ask questions, the
Daniel: more dangerous
Arthur: it is because the, the greater the chance we're gonna end up in the future totally unprepared for what we're going to see.
Daniel: I'd like one take on something that you're very excited about around where things are going.
Look,
Arthur: it's a really complicated one, but this is what my current passion is. So as we discussed earlier, I have very close friends who are very different from me. I've had coincidences that are too difficult to even explain, so what I
Daniel: think
Arthur: these technologies are gonna start to unlock, especially quantum plus AI, is to understand that layer we were talking about, which is humans converse on speech.
Below that is body language. Below that is pheromone. Below that is [00:51:00] energy. Below that is something else. .. The latest theory or a latest theory is the universe is not built on matter or energy, it's built on information.
Daniel: And
Arthur: even Demis Hassabis from DeepMind was talking about it in December and, I found it interesting from him because he's very binary AI.
But assuming that is true that the universe is built on information and the Big Bang was an explosion of un- observed information, i.e. the universe is not filled with nothing, it's filled with an unobserved piece of information, and it only- It's but it's not ... and it only create... It only turns into information when it's observed,
Daniel: and the
Arthur: observer is inside the system, so the observation changes the outcome
Daniel: That
I think
Arthur: will start explaining things like does time exist?
Does distance exist? Does what's a deja vu? What's a multiverse? What are all these things? Why do you walk into a room and m- see [00:52:00] someone on the other side of the room and then go and talk to them and find they're a very close friend of yours eventually, but you don't even know anything about them?
What is that energy connection? We have entangled qubits. All this information about
Daniel: how the
Arthur: universe and humans operate together, I think is completely blind to us. What's consciousness? We talk about intelligence. What's
Daniel: Yeah. and sentience.
Arthur: Sentience? We're building AI systems, or we're marketing AI systems that are gonna be sentient.
We don't even know what sentience is. We don't know what consciousness is. We don't know what binds people together. We don't know what makes people friends. We don't know what the signals are between humans
the thing that excites me most about quantum at the moment, and the reason I'm looking post-quantum, is what is that going to unlock to allow us to get to the next level down?
- Newtonian physics [00:53:00] we thought the world operated like that in the universe. Then Newton invented Newtonian physics, and we could start mapping out some of the
Daniel: things that people had been writing
Arthur: 2,000 years before,
Daniel: but they didn't
Arthur: understand why they were doing it. Ptolemy was writing about, the way the universe was operating without understanding what he was doing, just through observation.
Newton wrote that down. Then people came up with quantum physics and blew Einstein's mind with, which this cannot be possible. But now we're at a stage where we can actually model quantum physics. What's that going to do next? And so the excitement for me is are we moving to a phase really quickly where it's going to break our preconceived ideas about everything?
Daniel: And
Arthur: I actually feel
In our lifetime, we are going to get to that next level, and I suspect what we will find is we're still on the tip of the iceberg. And then we'll get down, and we'll find- we're even more-
there's some water.
Daniel: We'll get to the
Arthur: water level, and we'll realize we're nowhere even
Daniel: get to the water level, and we'll realize- We're [00:54:00] even- ... nowhere even near the start. We're way more ignorant than we realize.
Arthur: Eden Base is clearly about making money. Investing is really great, and we are very good investors in these spaces.
But it's also unlocking access to technologies that are going to increase human knowledge so fast and in such an important way that we might understand that everything we - knew today is wrong.
Daniel: Oh, what a way to finish. Daniel, it's been so kind of you to share your time in what is such a, an exciting
Arthur: space to, learn about, so thank you.
Thank you. It was really cool.
