The Four Horsemen of Technological Disruption

Cam Harvey reveals why AI is just one piece of a much larger puzzle. In this episode, he outlines four simultaneous technological disruptions — artificial intelligence, quantum computing, decentralized technologies, and multiomics — arguing that nothing in history compares to having all four converge at once. Harvey explains why AI productivity gains are imminent in 2026, how quantum computing will solve optimization problems that today's supercomputers can't touch, why stablecoins will become the financial backbone for AI agents, and how the combination of all four could revolutionize personalized medicine within years, not decades. The synergies between these technologies, he argues, will dwarf the changes we've seen in recent history.

Transcript

SPEAKER_00 0:00

As always, I want to welcome everyone back to another time with Cam Harvey here with Cam Harvey's Through the Noise. In a previous discussion, we've been doing a lot of time talking about AI and its implications on society, work, and people. And you had mentioned that that this technology is not the only thing we need to be paying attention to. That there's a that there's a list of, I believe you said four different elements that you argue are are really important to understand right now. And I thought it may make sense for us to take some time and just give you a chance to remind me what those are and outline what you mean by this and why we should all be paying attention.

SPEAKER_01 0:47

Yes. So you might refer to these as the four horsemen. They're at the gate. And we've talked fairly extensively about the number one, and that's AI. And that's in our face today. And what I've noticed is that people don't fully appreciate the scale of the AI disruption. People, even with the technology that we've got today, which is incredibly impressive, are not being very ambitious in using it. There's a small number of people that are being creative, but at the average company, they're using it for things that uh are trivial or low-hanging fruit uh applications. So I really believe that AI has the possibility to dramatically increase productivity. We've not seen it in the data yet, but we will see it. And the timeline is short. So we haven't really talked about the timeline, but oh, I get asked, is this like five or ten years? No, it's 2026. So you're saying adoption utilization is far behind capability. Yeah, exactly. So, but it will catch on. So AI is not the only thing we should be talking about. So there are uh three other disruptions that are happening. And indeed, when I look historically, I don't see anything like this. So historically, it's one at a time. And to have two at a time would be historically abnormal, but we have four going at the same time. Each has a different sort of time horizon, but the key thing to realize is that they're all connected and there's a lot of synergy. So let me go through um the other uh three uh disruptions. So number one uh is AI, number two is quantum computing, number three is decentralized technologies, and number four is something called multi-omics. And let me describe uh each one. So quantum computing is in the news, but mainly for the wrong reasons, I think, um, because the quantum computer has the possibility of breaking current encryption sort of um algorithms. And that's that's a long way off. And I've estimated that it's going to be in the 2040s uh before that actually happens. And it's not really a big deal because we've got quantum proof uh encryption already. So it's not like we have to develop it, it's ready to go. So what people miss is the possibility of quantum computing tackling problems that can lead to significant productivity gains. So the quantum computer operates differently than a regular computer. Uh, regular computers just like a zero and a one, uh, the bits, um, whereas the quantum computer has got the possibility of being in between the zero and one. So these are called qubits. Um but I want to focus not on the mechanics of quantum, but what quantum can actually do. What problem will it solve for me? Exactly. So you've got to be solving a problem. So what it can do, what it's really good at, is uh optimization. And optimization problems that are infeasible with uh today's uh computers. So this might be uh a supply chain optimization that could be dramatically uh improved. What does that mean? Reduce costs, increased profit, all of this is good stuff. It's also uh possible to apply this to both institutional and personal portfolios to do a much better job of actually managing our wealth and putting everybody on an even footing uh to be able to take advantage of these algorithms and computers.

SPEAKER_00 5:34

Aaron Powell Is there an easy way to describe to us um why it's why it's feasible to solve those problems with quantum computing and not our current status quo?

SPEAKER_01 5:45

Aaron Powell Yeah, because the the current status quo is just like so limited by the zero and the one. So it's so when you've got this entanglement in between uh the zero and one, you can do things that just are not feasible. And uh and and biomedical research is going to be a big beneficiary of this, where you could look at millions uh or billions of interactions at the same time in doing a simulation. Uh and and this could be done on a regular computer, but it would take decades to pull it off with the fastest supercomputers that we have today. So this is a matter of hours on a quantum uh computer. So this will accelerate uh dramatically uh biomedical uh research. The number three is decentralized technologies, and this is all stuff crypto. Uh, and I'm not talking about Bitcoin or or Ether or anything like that. Um, even though this part of the system, what is kind of low-hanging fruit and kind of connects everything together are things like stable coins. And if you think about the world of AI agents, they need to be able to transact in some way. And the agent cannot go open a bank account. So having this decentralized technology of a stable coin uh empowers the AI agents to actually do transactions. And some of them uh might be small transactions, which is fine. The human would want to oversee larger transactions, but it empowers the AI to act in the financial world. And this is really uh important, and um at very low cost. So think of operating 24-7. Think of transferring anywhere in the world uh in one to ten seconds. Uh think about doing this uh at very low cost. So decentralized technologies allow for this.

SPEAKER_00 8:06

So just to make sure I'm tracking, we've gone through three technologies briefly, and we're on we're on third. First one is scaling intelligence, second one is scaling computational power, and the third one is dramatically reducing or solving transaction costs.

SPEAKER_01 8:23

Aaron Powell Not just transaction costs. So transaction costs are well transaction transactions the ability to transact. Exactly. So uh so this is what I'm talking about, the synergy uh between these technologies. Uh so the decentralized technology will enable uh the AI agents to do transactions. The decentralized technology is also uh important for something else, and that's identity. So identity is a big problem because you want to know if you're talking to a human or a non-human. And right now we don't have the ability to discern. This is the so-called Turing test, named after Alan Turing. Um and what decentralized identity actually does is it allows you to prove that you're a human. And and and this is essential for the future of AI. And you can imagine uh our political system being flooded by AI bots to try to change uh opinion. Well, if the message is not signed in a certain way, then you just assume that this is AI. And if I'm signing it, uh then you know it's me. So so I think that this is also important, and again, another synergy. So so these three are really essential for economic growth, and the fourth is multi-omics. So it's kind of like uh genomics. Um so this one is is completely outside of my this is super exciting, and uh and the rate of progress is so uh dramatic. So we know we've got uh like 20,000 genes, and genetic code has been mapped and stuff like this. But this is called multi because it's three things. It's a DNA, it's RNA. So uh so think of uh the DNA is 20,000, RNA, hundreds of thousands, and the third is the protein structure, and there's millions of elements uh to that. So millions of variations, and it could be due to genetics or environment, things like that. So it's really, really complex uh when you put all this together. And if you think about the way that we do science today, uh we run an experiment, we test a drug, there's people in a trial, and then we look at the average, the average result, and then figure out if that average has had a significant impact. Is it no impact or is it significant? So we're we're focusing on the average. So this technology allows for two different things. So, number one, we're gonna operate simultaneously on DNA, RNA, and proteins. And to do simulations, it's it's infeasible to do that with today's traditional uh computing, but we can do it with the quantum computer. So we can uh if you think about uh how this uh this works, um it could be the AI suggests like a new treatment, okay, and looks at billions of combinations and comes up with the best one that it thinks. And then uh day two, this could be done within a week. Uh you set the quantum computer going to uh to simulate and to see if this AI-driven idea is viable. And then let's let's say it is viable, but then day three, we've got patient number one. And what we do is to look at their DNA, RNA, protein structure, and tailor the dose and the treatment to that particular person. So we're no longer dealing with the average, we're dealing with the person.

SPEAKER_00 12:58

I heard something else. You skipped human trials. Yeah. Was that intentional or was that a lack of understanding of the process?

SPEAKER_01 13:07

It is so uh so costly to do this stage one, stage two, uh, stage three uh trial. It takes years. Uh so we can tailor a treatment almost immediately. So I'm not talking years. I'm talking about uh from idea to actually dosing in in one week. Okay, so that is a dramatic uh shift. And what does this mean? Much more effective uh treatments, uh fewer side effects because the treatment is focused on the individual. It means uh our life um improves, which I think is important. So the first three ideas talk about productivity and economic growth. The fourth idea talks about the quality of life, and in that will shift. Um and and all four of these are happening simultaneously. All four of them are linked, and the synergy is very significant. We focus on AI, it's in the news all the time. You need to look at the whole package, and the package is very promising.

SPEAKER_00 14:26

As we close, I just tell me if I get this right. We have these four technologies are being developed um independently. And each one independently already has uses. Uh they're all being used somewhere in some capacity. What we haven't seen yet, and but we will, is how the combinations of these create a, I mean, I don't have a better word for it than a Lollapalooza effect of benefits to um to society. And those changes, when they come, most likely, as you described with AI, the capabilities are far going to sort outpace the adoption, and we're going to be racing to catch up to these uh these dramatic shifts.

SPEAKER_01 15:10

Yeah. So uh one thing I want to emphasize is that these technologies have different timelines. So AI is playing out right now, and the productivity gains from AI will happen in 2026. Dramatic productivity gains, and that will continue. The quantum computing has got a longer horizon. Um we've made progress, the capabilities are doubling every year and a half, and you can extrapolate forward uh and indeed just within five years, we'll have something pretty good. Um, but it just keeps on getting better, and the capabilities increase. Decentralized uh technologies, those exist today, and we're seeing widespread interest in things like stablecoins amongst banks and companies, fintech. Um so so that's that's happening right now. And then the very last one, uh the bioomics, um, that that depends to some degree uh upon the AI, which is available today. It also crucially depends upon the quantum computer, which is not ready yet. So it's a longer horizon, uh, but still the vector is so steep that we should expect to see uh dramatic changes in the next five years. Uh changes, uh, the sum of these changes will will dwarf what we've seen in decades.

SPEAKER_00 16:49

Well, thank you so much. I mean, this is a wonderful synthesis, and what I hope is is as we continue to do these, we'll be able to revisit some of these ideas in depth. But thank you so much for your time. Definitely.