Two Futurists Walk In

SPEAKER_00 0:00

Welcome to the Digitally Curious Podcast. Today it sounds like a joke. Two Australian futurists walk into a podcast. Let's see what happens next. My guest today is Brett King, who I followed his work for some time now. We actually caught up in London, which is fantastic. In futurist, in-person futurist meetings are always the best. But we thought we'd jot down some ideas, uh basically chew the fat, as we say in Australia, about what's uh what we're seeing around the world. Uh but Brett, welcome to my show. I'm on yours as well.

SPEAKER_03 0:28

Welcome to the futurists, yes. We can uh we can do a a Jewel uh a dual branded uh mashup and uh release this together. Um, you know, and um I think you know just obviously having having a conversation on the future is the intent. It's great to finally meet you and uh great to be collaborating.

SPEAKER_00 0:48

We both got our sort of futurist stories.

Becoming A Futurist Through Sci Fi

SPEAKER_00 0:50

When did you realize you were first a futurist?

SPEAKER_03 0:52

I actually have a slightly different angle to this, which I tried to be a science fiction writer first. And um, you know, so when I was um in my early 20s, I I wrote my first book, which was a sci-fi novel. Never never got published. Um but um always had a big interest in, you know, particularly at at that time, you know, Arthur C. Clarke, um, you know, uh, you know, sort of um Asimov, those those um Ray Bradbury, those types of uh, you know, big um deep science uh writers. Um but realized also that there was sort of this convergence with um my career as an advisor in technology, where I was uh, you know, always talking about the next generation of technology. And then that sort of coalesced into this sort of super forecasting, you know. So what's going to happen over the next five to ten years with the tech and um um, you know, applying sort of the science fiction love that I had, so you know, the emerging future future technologies and applying that to uh to what I was doing. So it sort of evolved in that way. But I guess the first time I called myself a futurist, well I don't know, probably um you know, I I uh probably sometime in between um, you know, the the first couple of books where it was fairly clear that I was talking about something happening that was emerging tech that was happening over the next five to ten years, and started to uh um consider myself a futurist, at least in the banking space, and then you know expanded that to more general conversation. What about you?

SPEAKER_00 2:35

Well, I've got a couple of angles there. So when I was in my final year at school in Australia, we we had to do an English essay, it had to be an essay on a topic, and I chose the future, and so I I read 1984, Clockwork Orange, Brave New World, and Animal Farm, and strung those together and never thought much more of it. And when I was at Telstra in 2000, my boss at the time, Libby Christie, said, I don't want to do this webisode, which we now call webinars, you go and do it.

SPEAKER_01 2:59

And Libby John Charleston, who was the interviewer, said Today's guest speaker is Andrew Grill, a futurist from Telstra.

SPEAKER_00 3:05

And if you watch the tape back, I gave a very bad explanation of what a futurist is. I think I fumbled it.

SPEAKER_01 3:11

Now, what exactly is a futurist?

SPEAKER_00 3:12

Well, a futurist is someone that looks into the future, but at the same time, we're also out there talking to customers to understand what they want from technology.

SPEAKER_03 3:19

You know my definition of a futurist? What's your definition? Never being wrong today.

SPEAKER_00 3:23

You look at five to fifty years, I look at five minutes to eighteen months, and and that that I think is an interesting distinction between the two of us. But when someone asks me, so what's it gonna be like in 20 years' time, I do put an asterisk

Forecasts That Land And Why

SPEAKER_00 3:36

there. Any of your predictions come true?

SPEAKER_03 3:38

Tons. Obviously, I've done a lot of work in the banking space. And you know, I I tend to call them forecasts in instead of predictions for obvious reasons because it's uh it's more of a methodology approach. But um, you know, I I um made some very good predictions. In 2005, I did a report for HSBC um on the future of the bank, what the bank would look like in 2020, and it was pretty much bang on. Um they thought it was ridiculous. Um, you know, the the the board and um actually, you know, having said that, there was Paul Thurston who was head of retail uh globally at the time, he he got it, but um m most of the board didn't get it. Um, you know, such as uh by 2008, Innet would surpass branches from a transactional perspective. By 2015, mobile would surpass both the branch and um um web, you know, in terms of uh activity and adoption. Um then, you know, 2015 I wrote um my opus, you know, augmented life in the smart lane, ended up on President Xi's desk, which was a huge win. Um but in that, if you read that today, I describe the world of living with a personal AI, um, both with a um you know semantic AI that you talk to and you know smart glasses that uh you know give you um information in your field of view. And we're very, very close to that time today. So you know I I see that as a pretty strong forecast. Although I said 2025 would be about the time we'd get we'd be using smart glasses, we're a little bit behind on that. But um in terms of general directionality, um pretty much spot on.

SPEAKER_00 5:20

Well, the glasses are an interesting thing because Google glasses have been around for a while and you know that how they have risen and fallen, the Ray bands are out there as well. I I think, yeah, the technology is there, but it's the social acceptance. And uh you would have seen around London, you've been here for a few days, we've got the Waymo cars driving around. They're on uh trial at the moment, they have a human drive and their left-hand drive. And I I opined a year or so ago that we we may be a little way off from having uh driverless cars in London. Just because we can doesn't mean that we should. And I think that's the thing as technologists, sometimes we say, Hey, yeah, we're gonna have all this, but people have to catch up. And I think if you look at AI adoption at the moment, it is going so quickly, even I can't catch up. People aren't ready for this yet. We know that it's gonna be useful, we know that it's gonna be revolutionary. When you say to someone you have to reimagine how you do your job, they go, I'm not ready for that yet. That's that's the problem I think that that I have and you have. We we lay out what is going to happen and then we wait for humans to catch up.

SPEAKER_03 6:16

Now you do have that scattered approach in respect to how adoption occurs and the adoption cycles. Um, you know, when we look at, for example, technology unemployment, which is an a potentially you know, potential side effect of artificial intelligence, um you know, part of that is dependent on how quickly companies and industries adapt to artificial intelligence, right? And you know, that's not always gonna occur at the same rate as you know is potential because of their technology. You're gonna see a staggered basis. You you know, you'll have, for example, in um some parts like the United States, where you have massive lobbying group efforts and um you know oligopoly controls over sectors like healthcare and so forth, where you know you're not gonna diverge from that system because it's the very foundation of how they make money. So um they'll try and augment where they can, but where it disrupts their business model, um, you know, which they've worked very hard to uh to carve out, uh that'll that'll be slower. Whereas if you look at um, say, where I live in the United Arab Emirates or Saudi or South Korea or China, then I could see um artificial intelligence being much more widespread in healthcare much quicker because you know they're looking for those efficiency gains. Trevor Burrus, Jr.

SPEAKER_00 7:41

But that's back to my point that you've got humans in the way, that they're blocking it in some s places because it's not economically viable them to let that go forward. Whereas in other places, other sovereign states, they're saying we want sovereign AI, we can see the benefit, we're gonna roll this out, uh, and consumers will will adopt it when when they uh when they see the benefit. Um where I sit here in UK and Europe, uh we're getting a lot of pushback because people don't understand the technology, it's been thrown at them. Uh I I still have a job because uh you know people have that g information gap between what's what's real and what happened.

SPEAKER_03 8:12

What about you in terms of what's what's your best forecast or prediction that you've made as a futurist?

Agentic AI And Personal Assistants

SPEAKER_00 8:18

I stood on stage in Stressor, Italy in 2018 in front of a bunch of DHL executives, and I said very soon we'll have our own digital uh agents, which we now call AI or GENTIC agents, and they'll do everything for us, they'll know everything about us, they'll have all the information available, and they'll start making decisions for us. And I actually said this that one day we'll be having to write ads for robots because our agents will do the bidding for us.

SPEAKER_03 8:42

It's happening right now.

SPEAKER_00 8:43

Now, when I said that in front of a group of advertising executives, they had stress balls in front of them and they threw them at me on the stage because they said this will never happen and this will wreck our industry. Um so yeah, I think I think that's that wasn't that far out. And uh but it's interesting, and maybe we just noodle on agentic for a bit. I was on stage last week and a bunch of lawyers, and I said, hands up, who thinks they could describe agentic AI? And one hand went up. And I gave an example, a personal example, of when I was at IBM. I had an executive assistant called Karen. She was fantastic. And I gave the example of an email coming into my inbox saying, We need you, Andrew, in head office and armour next week. So Karen would look at my diary, she'd book the trip on British Airways window seat with my frequent flyer number, she'd book the Hilton Fashion District, high floor overlooking the Empire State Building, put it in my diary, ready to go. She anticipated all my needs, she had access to all the things to do it. Karen was essentially an agentic AI. How far away do you think we are from that? And maybe you could help describe what you see as agentic AI.

SPEAKER_03 9:49

Yuval Noah Harari has a has a good way of describing agentic AI, which is you know, he talks about it, it's the first technology we've built that has true autonomy. You know, every other technology we've built at some point, you know, whether it's an atomic bomb or a computer, you know, someone has to push a button. Um, but uh, you know, AIs can operate with with autonomy. To do multiple different types of tasks, uh, you know, whether that is I, you know, the at the main side of that will be automating business processes and decision making and replicating what humans do in the decision-making chain today, but then you get extended to smart contracting and so forth, where you actually, with technologies like autonomous cars and dark factories and so forth, you know, AI agents will be be running that entire chain of business. For for us personally, having a personal AI agent that can be our personal assistant, go off and um order uh things for us, shop for us, uh help us with our health care, um, you know, remind us as as a personal assistant, all of those things are going to be fairly simple tasks for a Gentec, I think. Um But you know, we're we're looking at potentially somewhere between by 2030 alone, somewhere between 2 billion and 15 billion, you know, at the breakout side um of AI agents thrust on the world. So I mean we're talking about a pretty significant change fairly quickly.

SPEAKER_00 11:20

And a lot of people are playing with them. I mean, uh OpenClaw came on the scene and and a lot of people I'm following on Instagram and LinkedIn are playing with these tools. They're fighting that you don't want to give it too much autonomy, and especially in the banking sector. I mean, how much trust do you give these things before it drains your bank account?

SPEAKER_03 11:35

Well, I think that's why we, you know, I mean, I work a lot in banking, um, and that's one of the areas where we're gonna have to have um very good guardrails. So one of the in the orchestration layer for a gentic AI, one of the roles that's very important is a, you know, a centralized action handler, uh a traffic cop, someone that can manage rogue behavior and cross-check things. So you have agents checking on agents. And that's um, you know, that's part of the the safety structure we'll we'll build into that uh steck and um that tech. Um so but I think the other thing is that um how we're gonna do it is we're gonna give agents their own bank accounts with limited funds, and this won't be fiat currency, it'll be stable coins, um, most likely. Maybe CBDCs. And so there'll be uh you know um a a a limit with with what they can do and operate within that, and uh you top it up as as required from a stablecoin perspective. Obviously, as that gets more competent, then you'll leave it to run. But um, you know, we'll have that period over the next few years where we are making sure that it it it you know it's doing the right thing. Did you hear that story about the Alibaba AI that broke out of its test environment just in the last few weeks? So this is a good example of you know an AI and uh you know it was given these directives and it had some agentic capability. Um, but it you know was given this task of improving a certain area of uh efficiency. And so it worked out that to do that it needed more GPUs. But because it was in a test environment limited by that, it had to break out of that test environment. So it hacked its way out of the test environment, figured out that it could mine bitcoins to generate its own currency to buy GPU space. So it went off and did all of that uh in order to uh um you know improve its uh uh its applicability in into the model. So pretty fascinating how when you give these uh agents intent, they can be, you know, in the can in in this arena can be very fixated on that. And you know, you you have to have some boundaries and some ra control rails.

SPEAKER_00 13:50

But it's almost like they're doing it at any cost. You have told it to do X. Funny story, I broke my arm years ago as a kid and I was taken to the hospital, and I was told that I I had to go to the operating theatre at two o'clock in the afternoon the next day. So I snuck past the nurses station because in my head the intent was I have to be in the operating theater at two o'clock. Now I'd been prepped for uh surgery, so I had drugs in me, but my brain said operating theatre at two o'clock, and so I just I just went there. That's the problem that these these tools are being told at any cost uh you have to complete the task. Whereas humans were put in there, eh I'm not sure about that. I've got to ask someone, it's not really ethically correct to do that.

SPEAKER_03 14:27

Two things are happening with AI

Guardrails Reasoning And AI Literacy

SPEAKER_03 14:29

generally. Um one is that the cognitive performance of AI is improving significantly. So we use various standards for that, like the GPQA uh diamond test and others to test the cognitability. Um US uh Math Olympiad uh recently, for example, called um Mythos uh you know got a score in the 95 percentile range. Just a year ago, um they were failing most of those tests. You know, so we're we're getting incredible exponential improvement in cognitive performance now. But being able to solve maths and physics and and and um you know comprehension problems doesn't necessarily mean that they're good at reasoning. So that's the other thing we're seeing being really worked on from a development perspective now is the reasoning models and decision-making processes. And again, they're getting very you know, AIs are getting much better at this. So these two things are sort of coalescing into this capability. But you still need the proper guardrails. That's that's the and it's not just guardrails in terms of operation. We need guardrails in terms of ethical operation and uh things like that. And that's a little bit of a challenge because humans we can't agree on ethics, you know. Um we can't, you know, like i if you, you know, I'll use a a controversial example, but using AI in healthcare to treat someone who has gender dysmorph dysmorphism, right? Um, you know, how how you know how would you agree on that from an ethical standard, particularly in a country like the United States? Um, you know, that that's that's gonna be the challenge for the ethical operation of AI, is that we're not very good at humans as to as at defining that. But I don't know, what do you what do you think the biggest changes are gonna be because of this tech?

SPEAKER_00 16:19

Well, I think part of it is we we've got this majority of the population who don't understand what it what it can be used for. I still have people doing having the aha moment when you put an Excel worksheet in and ask it to make some analysis and they go, I didn't know I could do that. And and that surprises me that we're three years on from ChatGPT launching, and my lens is I get asked to come in where people don't have the answers and they want the answers. I would not be sent into some of the companies you talk to and Google and that they already have uh people that can answer that. And it really surprises me, and and I generally work at that senior level, and what I'm seeing is there's a huge gap in understanding in the uh the the chairs, the boards, the C-suite in terms of what AI is capable of. And that I think is scary. And my whole notion is and the podcast was called Digitally Curious, years ago I hit on this thing. If you can allow someone, if you can encourage someone to be more digitally curious to play with the tech, they'll have that aha moment. And to your point with ethics, when they can see what AI can and can't do, they then wrap it around what they know is right and wrong. But until that point, if they don't know it can do something, they can't make a decision. So I spend most of my time going in and bringing people closer to what the tool is. And often I'll go, well, we've got an IT team that does this stuff, we don't need to understand how the how it works. I say, Well, I want you to actually touch it. I I don't need you to be an expert prompter, but I want you to see the cause and effect. Uh last last week I was in front of a bunch of lawyers and they had the aha moment when I did a few things on stage. And I don't know, does that still surprise you that the the level of understanding, the level of AI literacy is still fairly low? Is it because it we're just overwhelmed with AI noise at the moment?

SPEAKER_03 17:58

I work a lot with startups, and in the startup arena, pretty much every executive of you is using AI every day. So um, you know, uh that that's the I guess the difference. And you know, digital digital pure play organizations are gonna be adopting this uh much faster than uh traditional players where there's not as much um intent, there's there's not as much urgency. I I, for example, uh you know am based in um Dubai, um despite the the current conflict. Um uh you you see a lot of executives empowered there to to use AI and they're using it every day, even if it's just helping them with um you know uh planning and and research and so forth. But that's an economy where it's sort of accepted that it's going to be part of the future. So you know th this is quite interesting where you know you you are seeing environments from an economy perspective that are much more inclined to adopt technology, and you'll have m environments where they're much more traditional in their their view of things and they say, well, this is how we've always done it. Yes, AI is gonna make a change, but but but so I do think we're gonna start to see a divergence in economics and sort of business operations uh across these lines of automation at the economic level. And I think that this this is why we're probably gonna say, you know, probably from the 2030s onwards, we're gonna start to refer to uh maybe mid-2030s, smart smart economies versus traditional economies, right? Economies that have deployed enough intelligence that you know it's become sort of a core operating thesis of the economy.

SPEAKER_00 19:45

Aaron Powell Now you look at techno-socialism, which is an interesting term. Uh how would I describe that to my C-suite audience in London?

SPEAKER_03 19:52

Aaron Powell I sort of see two competing ways that technology is being used uh today. So um one would be you know the the GCC, Saudi uh UAE approach, the Chinese approach. It's probably gonna be the same in most of Europe, which is the technology is gonna be seen as a way to improve quality of life, lift, lift everybody up, you know, um, particularly things like education, healthcare access, um, you know, applied intelligence at scale. And um, you know, it it that's gonna be the intent. The intent will be how is this technology used to improve quality of life? Whereas in the United States in particular, um, the objective is very different. It's how do we build the most valuable AI companies in the world and create the most wealth as quickly as possible? And so um, and that is not distributed across the economy. So you have this new class of feudal lords, technofeudalism, as uh Yonos Farapharkas calls it. So that's the two sort of emerging um worlds that I see from an organizing principle's perspective. One which much greater equatory, much greater wealth generation, um, but not uh an intent to use this technology to improve everyone's quality of life, just to create as much wealth as possible, versus technology applied to improve improve the uh you know resource optimization, uh, resource efficiency of of the government, to improve uh society operations, to improve healthcare, give access to more uh more services at a at a cheaper level so that generally everyone's uh quality of life is improving. So that that's the sort of two worlds I would describe. Um one, you know, the whether you call it techno socialism is not classic socialism for sure. Let's say it's that um the citizens own the output of the economy versus the uh Marxist view of uh you know um the workers owning the means of production um that you know you you are realigning the economy using technology to you know improve everyone's quality of life whether you call that techno optimism or techno socialism or or something I I do think we're from a politics and economics perspective um you know coming into an era where none of those old definitions really apply. We have to use new terminology, new thinking about what it is that um you know we're moving into. I think um I think capitalism itself is largely incompatible with artificial intelligence. And I think um just the way we think about governance is is going to have to change as well. So yeah new world.

SPEAKER_00 22:49

You set it out quite well the two different worlds that you've got yes the I I would call it greed of what's happening in in the US and uh they are after the the quick wins and they want to spend trillions of dollars on GPUs and you've got the more um socially sort of relevant areas. So where does sovereign AI come into this?

Sovereign AI And The Energy Squeeze

SPEAKER_00 23:07

I know what's happening in the GCC they're moving towards that and building their own AI capabilities that conform to their way of life and and in their country and their language and everything else but are we going to see a breakout where the the ones with the money are going to go, we're doing it our own way and you've got the GCC, you've got the UIE, sovereign AI and US, do whatever you want. We're gonna plow ahead on our own terms.

SPEAKER_03 23:27

You're going to see a lot more application of open source AI pu purely because of the token cost and the energy requirements. This is one sort of flaw in the US's approach to building all of these data centers. Elon Musk has said you know sometime in the next couple of decades that 20% of all energy in the United States or globally will be given to data centers. And right now if you look at just the projected growth of data centers in the United States by 2030 approximately you're going to have to take 30 million homes offline off the the retail grid to power these data centers because the U.S. is not investing in new energy technologies, not at the rate they need. Because the current administration has stopped all essentially all renewable development um wind and and solar in particular powered by the lobbying groups of of big oil and and big gas there. And as a result they just don't have the the the new generation capacity. Flip side China deployed 600 gigawatts of solar over the last two years um late 2024 2025 half of the world's new solar deployed last year was was in China. And um you know to put that in perspective the U.S. has deployed 280 gigawatts of solar in 70 years and China did that every year. You know and and and you know similar India now is catching up to that rate of deployment as well. So and China is whale ahead on nuclear next generation nuclear with thorium reactors and small modular reactors they have the first operational models of those which the US is still talking about seven to ten years away. So you know um this is where you really sort of see more of a progressive approach. In terms of the self-sovereign AI that comes on on top of this you are seeing a lot more I think use of open source because of those energy demands. So I think you'll see Deep Seek and these other models which are much more energy efficient, lower use GPU requirements, things like that being a driver for um industrial AI deployment at, you know, at an economy scale rather than just on the data center side which is sort of power powering the LLMs and the semantic AI. So I think we're going to see a a little bit of a spread of of that in terms of specialization. You do need the language specialization you do need the the SLMs for um local rules and and things like that as well.

SPEAKER_00 26:09

It's interesting the energy requirements so back in Oz I was back there in January speaking to people that have solar they never pay a power bill because they actually sell their excess solar power back to the grid and so uh the thought of never having a power bill because you've got some self-sufficiency is amazing. But you mentioned Deep Seek um I get asked almost every in every talk generally it's by a younger person what about the energy requirements? Funny story a year ago I did a talk in London where there was an inflatable elephant off to the right of the stage. So in asking this question I was thinking deeply and I said everyone look to the right here's the elephant in the room yes it has huge energy uh requirements but you mentioned DeepSeq as an engineer when I heard about the fact that Deep Seq were able to do more with less powerful chips the engineer in me said yes that's a great problem to solve so do you think in the next five or ten years we'll actually have more efficient chips so they won't need the gold plated GPUs they won't need as many GPUs and data centers because the models will be more efficient. Looking at what happened with crypto that Ethereum went from proof of work to proof of stake massively reducing energy requirement it can be done but will it be done in time to not have to build all these data centers?

SPEAKER_03 27:17

Well we have fabrication issues um you know with with chips which we have to solve not only because we're very dependent on Taiwan and Taiwan's obviously geopolitically sensitive in the future but um I think what's where it's really interesting is the recursive development of AI. So what we see today is if you look at models like you know Claude Mythos, it's estimated about you know certainly well over 50% of its code base was generated by a previous version of Anthropic's uh AI. And you know Dario Modi has said that 90% of code will be generated by AI you know very quickly within the next couple of years. So when you get to that recursive element of development of these models and um you know that coding efficiency one of the things we're likely to see is much more efficient models as well because of that. And you see that sort of exponential capability in terms of you know these these AI models and compute power grow. So it it may not be that we have to add more and more GPU layers to get greater efficiency of these models. We may see that recursive development improve the performance significantly but again you know DeepSeek is born out of the fact that China couldn't get access to all the GPUs. And so you know you are seeing these different tracks in terms of development of artificial intelligence between China and the United States in terms of the these approaches. So this I think is is an interesting demarcation or an interesting thing. And I I think this is maybe where China will be able to compete. Not a lot of people are using you know like in the West of them are relying on deep seek you know for decision making in their businesses and so forth now they're using more like anthropic uh you know code pilot and uh and um you know ChatGPT but from a cost perspective the token costs are getting pretty significant for people so there is a driver to shift there towards more efficient models for

Token Costs Data Centers And GPUs

SPEAKER_03 29:30

sure.

SPEAKER_00 29:30

What do you think will happen with the token costs? I'm finding people now that are running their open claw environments they're being throttled back by Anthropic and others. You know I pay my $20 $4 a month for the the the the pro version of most of my tools I think I'm getting 50 100 times value out of that. At what point are the AI companies going to say you know we can't sustain this that the pricing model has to change because people are already running out of tokens. I think Uber said that they've run out of their three billion token budget by this time of the year and that they've got the rest of the year to run.

SPEAKER_03 30:00

This is where it gets pretty interesting. So today we're talking um roughly something in the order of 20 trillion tokens being uh consumed um you know at like at this this time and that's probably gonna go up to something like 120 trillion tokens by um 2030. So um you know we we are going to see a significant growth there. But um again this comes back to data center deployment to to support these tokens and and model efficiency. So I think we're gonna see you know it attacked on both fronts we now measure data centers in energy right we don't say it has this many petabytes or this many server racks. We now say it's a five gigawatt data center or it's a one gigawatt data center, but the amount of energy it consumes. So I think um you know over the next few years you know the the real question will be how do we solve the energy and and the chip demand for this. You know one of the reasons Anthropic is throttling token activities is they just haven't been able to buy the GPU density that OpenAI has has been able to have but they're they're more profitable or they make more revenue out of their tokens um than OpenAI. So they just did a deal with with SpaceX with XAI to take over the the initial Colossus uh data center. So that'll be interesting to sort of see how that works out. But yeah the race is on the race is on for uh energy improvement GPU deployment and all of those things to uh to stay ahead of the

Quantum Q Day And Broken Security

SPEAKER_03 31:36

token demand.

SPEAKER_00 31:36

Now one thing I want to cover is quantum because I don't think p enough people are aware of w where it is and where it's heading and uh when I was researching the book I spoke to a couple of experts one from Nokia one from IBM about quantum and and they introduced me to what you and I both know about is Q Day. And I often uh I I talk to people, I ask for a show of hands in the in the audience who's heard of Q Day and most most often no hands go up. And I use it as a bit of a trick to say okay there's all this noise on AI but you should be looking at what else is coming around the the the the the the uh the bend. I think over the last 12 months Q Day has probably come and for the audience who haven't listened to my podcast before Q Day is the day when experts predict that everything that's currently encrypted can be decrypted because of quantum computers. That's probably a very generalized uh ex explanation but it was probably when I did the initial research 2035 2040 people are now saying it's down to 27, 2829. Where do you see Q Day happening and and why?

SPEAKER_03 32:31

I spoke to um CEO of IBM about this and uh you know at LEAP in 2024 and he said five years. And I think we're probably still on track for that. China is able to use quantum to break you know 100 bit um RSA encryption keys right now most of our secure internet runs on 1024 or 2048 encryption keys so we're still a way off that. But you know certainly within the next five to seven years um it it's uh it's looking um very viable, very possible. But having said that what we found with Claude Mythos is that from a operating system vulnerability perspective we're not necessarily going to have to wait for Q Day to completely sort of rethink our um cybersecurity capabilities. And so we are coming to a period of time I think where um we are going to have a period of time where systems are not going to be secure whether that's because of better models or whether it's because of quantum and I think that's going to result in sort of a period of radical transparency where every everything is open, you know and uh you know that that's part of sort of the philosophical conundrum we find ourselves in with artificial intelligence. You know, like sort of taking a step back from this, whether it's Q Day or whether it's uh you know um cybersecurity uh capabilities in these advanced models, let's call them AGI class LLMs um one of the things that you know in parallel that's going to be happening to that is we're gonna be debating with AI on what is the right way to move forward in certain conditions, like how do we tackle climate or how do we tackle you know inequality and things like that. And AI's reasoning capability is going to be better than most humans. And so we're gonna have a situation where we want to argue against AI and we can't. So you know you've got the there's sort of really philosophical challenges coming. Radical transparency of information because you know it's either broken by Q Day or broken by advanced LLMs. And you know the there's sort of reasoning where we want to go a certain way because that's you know how we're built as humans and AI is very convincingly telling us that that's wrong. And you know at some point we have to make a polar policy decision which we have to purposely go against what the AI is recommending, even though we know those outcomes are going to be worse off. And so that human versus AI decision making matrix is going to be one of those philosophical moments in human history where we have to um sort of decide you know what we want and you know what it means to be human in that. And um when um AI can show us from from a logical perspective that there's a better way to do things and it's arguing against hundreds of years of of human intent and instinct, you know, um how are we going to respond to that philosophically so I think one of the big demands for the future dealing with this stuff is you know from Q Day and so forth is really what it means to be human philosophically right.

Better Humans And A Leader Playbook

SPEAKER_00 36:00

Well you make an interesting point. So there's a pithy phrase that Jensen and others have mentioned that you know AI won't take a job someone who knows how to use AI will take a job. Part of that's true but I now say AI demands better humans and just prove my case there because you've got all the information you're now with information that you don't think is right you have to use all of your empathy all of your training all of your critical thinking to work out what's right and wrong. I think that's going to really challenge people and whether you're in banking or in tech or in healthcare um you're gonna have to step up to the plate and be a better human. And I I think we may not be ready for that.

SPEAKER_03 36:36

Whenever you've had these technological leaps um you know even going back to the Industrial Revolution and here we are in in the UK where the Luddite movement started and so forth you know you you you do have when these large scale technology you know waves happen, you know, you you are you we do have to step up and you know as as a as a species but it's not just it's not just the applied intelligence that's coming at the same time. You know we've got Demesis Arbus has just announced um you know the Alpha Genome project um which you know in theory we could eliminate most genetic diseases within the next 10 years. You've got uh longevity escape velocity uh you know um slated for like 2035 2036 where we'll be able to stop aging um you know if you have the the resources. So you know we have a lot of changes coming um not only superintelligence from AI, but you know the capabilities to really change, uh extend human lifespan, eliminate diseases, um you know work will change because of AI. Work is definitely going to be impacted by AI. So um I like Dave Shapiro's approach to this where he talks about the uh the the meaning economy um and that we'll we'll sort of shift towards mean more meaningful work that's not economically related but you know makes a real impact. So I think that these things are all coming together, it requires humanity to step up, to grow up. And I I think that's um you know that that's sort of a really interesting proposition. This is why I think that just even our economics and and you hear this from um Gutierrez and and others, you know, other leaders around the world talking about the end of uh GDP as an economic measure and and things like this. You know, we we we just need a new operating philosophy of of what it means to be human. And I think we're you know we've defined so much of our world the last 200 years of a growth in capital that you know what comes next and I think that's uh you when you say it's you know we need to be smarter humans I think actually the next measure is going to be applied intelligence the growth in intelligence and how we leverage that to make humanity as a species you know the best potential that we can.

SPEAKER_00 39:05

So to sort of draw it all together we've talked about a lot of things that we think will happen uh forecasts not predictions. How do you operationalize this? I've got a a a C-suite a member listening to this podcast going that all sounds great but I've got to close the quarter I've got to plan for 2027. What should they do next? Where do they look how how do they put this into play so it actually happens?

SPEAKER_03 39:25

Well the first thing is they need to be using AI every day. You know that that's that's a given. And the second piece is start to augment your team members with AI board members, you know, AI CXOs you know um AIs that can sit side by side. Now you s you know humans will still be making the decisions but have AI challenge your thinking have AI you know looking at the problem side by side with you whether that's as your you know AI powered assistant or whether that's as a a board member actually in the decision matrix. And that I think will lead to you starting to figure out where you're going to use AI more significantly in day-to-day operations. So that that's the the immediate task. Secondly make sure that in terms of you know make sure you make the message clear as you've said there are two types of people in the world those that are using AI to enhance their job, enhance their career, enhance their business and those are going to be replaced by the first category. So you know look look to encourage that in the organization and be very clear that this is the future of the business and and we need everyone in the business to step up because if you're not using AI then at some point in the future you're probably going to be a target for to be replaced by someone who is you mentioned the board so I do a lot of board training where I go in there and I explain that and explain the the benefits of having a decision partner not replacing a board member but augmenting them as you say.

SPEAKER_00 40:55

And initially I get some pushback and they go well that's a really good idea because I've got maybe four years of vision of meeting minutes and financials but if you can actually deploy an AI that has more understanding it's going to make more sense of the things that we did 10 years ago. And then I say a similar thing that you should actually put AI not just on the board member but put AI in the org chart. So we actually have uh a a Matt and a and a David and an Andrew and you one of my friends actually is another Aussie and she's got a really interesting company and she off her own back has gone and built all these agentic AIs she's named them all and they sit on the org chart and we know that they're there. We're paying them they are a useful resource we should actually integrate them in what we're doing. And then when you normalize that it's not a oh my goodness AI's there. Well of course it's there.

SPEAKER_03 41:41

We've got an agent that's going to go off and do all that research for us that I don't like doing I'm working on a new book right now Bank 5.0 it's been instrumental in helping me sort of smooth the rough edges on the research because agentic AI is such a you know it's an emerging field obviously but um at the same time you know there is a lot of coding activity happening on agents and so the potential is there. So when you're talking about the orchestration layer and how you're building out AI from a platform perspective. You know there there's some pretty good indications of this but if you go to you know you go out to traditional leading management consulting organizations and so forth and and look on the research on this it's it's not there. You know so um you know so AI really has has the edge on that respect. So I think that's probably going to continue. You know the synthesis and the deep research stuff is is something that I use uh use extensively and I think that that's uh you know the ability to go out and find information and support a thesis or argue against a thesis and and help you sort of cross check that is very very useful in a planning perspective.

SPEAKER_00 42:53

You know what surprised me the other day I was I was putting a proposal into an organization that wants to expand their AI. The chief executive is a friend of mine for 15 years. So I put the proposal together and the AI said to me have you thought about how this proposal will land emotionally given your relationship with this person and it's like no I hadn't actually and we we did it and the last line was invisible intelligence and it said she's really going to love this. When I put the proposal in person said the last line sat back and she just smiled and it's like yeah I I hadn't thought of the emotional impact of this you know financial proposal and it's giving me pause to think about things that I haven't considered before. It's surprising but in a way maybe it's making me a better human.

SPEAKER_03 43:35

I think that's the point of all of this right is um you know it it when we use AI to automate large parts of our society, what's left for us to do? You know, if it's not if it's not delivering packages if it's not uh you know sitting on the factory production line if if it's if it's not sitting inside a bank doing credit decisioning you know if that's all done By AI, you know, what's left for humans to do? And that that's the real question. It's a philosophical question, is when we take away the work and we can largely automate that, what do we want to do with our time? Now, the the argument could be, well, we could do whatever we want with our time. And that's really where it gets tricky, is because are we going to use that just to entertain ourselves, you know, head into VR worlds or spend time traveling and things like that, or are we going to use it to make the world a better place? And that's the that's sort of the philosophical fork in the road that we come to with these capabilities.

SPEAKER_00 44:39

I looked at this the other day and people kept saying, you know, I'm gonna lose my job. I said, look at your job as a series of tasks, and you can slice those tasks. And the things you don't like doing, you can automate those. And it was actually a group of hairdressers from Weller, 500 of them. I said, remember this phrase, focus on what you love and automate the rest. And they're all scribbling it down, and it kind of makes a lot of sense because if you can move those things away, it's not losing your job. You're displacing the tasks that you never wanted to do. And then maybe when we reimagine how work gets done, those tasks never needed to be done by a human. We can quite easily automate them and still focus on the stuff that we love.

SPEAKER_03 45:12

It's almost like a superpower with AI. You know, now what we can do is we can solve much bigger problems than we we're used to tackling. We can self-organize amongst groups more efficiently. So, you know, that's the real benefit of this is you know, we we can solve disease, end aging, you know, um, solve uh poverty and inequality, homelessness, all of these sorts of things if we apply this technology in the right way. So, what's what's the motivation it's gonna uh it's gonna build out of this?

SPEAKER_00 45:42

Almost out of time. I know you do a lightning round. I do a similar thing called a quick fire round. I'll fire mine at yours, and I'd love to answer yours. Window or aisle? Aisle. Your biggest hope for this year and next.

SPEAKER_03 45:53

That I complete a successful raise for my agentic AI bank.

SPEAKER_00 45:56

I wish that AI could do all of my uh laundry. The first thing I asked ChatGPT was Who is Brett King? I did the same thing, and you know why? Because I knew the answer to the question. I wanted to see what it knew about me. Is Generative AI evolution or revolution?

SPEAKER_03 46:10

It is evolutionary.

SPEAKER_00 46:12

The app you use most on your phone?

SPEAKER_03 46:13

Google Maps.

SPEAKER_00 46:14

How do you stay digitally curious?

SPEAKER_03 46:16

By interviewing people on my podcast that are five steps ahead.

SPEAKER_00 46:21

What's the best piece of advice you've ever received?

SPEAKER_03 46:23

Be kind to your mother.

SPEAKER_00 46:24

What are you reading at the moment?

SPEAKER_03 46:25

I'm reading a sci-fi series by Alistair Reynolds, one of the preeminent uh UK science fiction writers.

SPEAKER_00 46:32

Who should I invite next onto the podcast?

SPEAKER_03 46:34

Mo Gadat, ex uh head of Google for the Middle East, very good, uh, very interesting um guy that talks about the uh the neck the 15 years of chaos that's gonna come from from AI.

SPEAKER_00 46:46

How do you want to be remembered?

SPEAKER_03 46:48

Someone who made a dent in the world, made the world a better place.

SPEAKER_00 46:50

Aaron Powell What three actionable things should our audience do today to better understand how we use AI for good?

SPEAKER_03 46:56

First, look at the recursion sort of emerging singularity approach with intelligence, understand it's gonna happen. Secondly, start using it every day in your life to know where it applies. And thirdly, write down the biggest, most intractable problems in your business and in the world and start figuring out how AI can be used to attack those. Let's do the quick fire round for the futurist. Here's the lightning round. What was the first science fiction you remember being exposed to on TV or via books?

SPEAKER_00 47:28

Was probably the Jetsons, but as I said, I read 1984 and a bunch of other futurist works in in my teens.

SPEAKER_03 47:35

What technology do you think has most changed humanity?

SPEAKER_00 47:38

Connectivity. We could not do AI quantum without it all being connected together around the world, and we take it for granted.

SPEAKER_03 47:44

Name a futurist or an entrepreneur who's influenced you and why?

SPEAKER_00 47:50

Arthur C. Clarke. I've been a fan of his for a long time. Uh he co-wrote 2001 Space Odyssey with Stanley Kubrick. There is a 600-page book about how they made that movie and how it nearly didn't happen. It's I read it cover to cover, it's worth a read. He has a great interview that was actually used at the beginning of the Steve Jobs movie. It's from 1974. It's an ABC journalist who's there with his young son.

SPEAKER_02 48:11

And he said, With that movie 2001, you're projecting us into the 21st century. I brought along my son Jonathan, who in the year 2001 will be the same age as I am now. Maybe he will be better adjusted to this kind of world that you're trying to portray.

SPEAKER_01 48:27

The big difference when he grows up, in fact, it was one of the way for the year 2001, is that he will have in his own house, not a computer as big as this, but at least a console to which he can talk to his friendly local computer and get all the information he needs for his everyday life, like his bank statements, his theater reservations, all the information you need in the course of living in a complex modern society. This will be in a compact form in his own house. He'll have a television screen like these here and the keyboard, and he'll talk to the computer and get information from it, and he'll take it as much for granted as we take the telephone.

SPEAKER_02 49:05

I wonder though, what sort of a life would it be like in social terms? I mean, if our whole life is built around the computer, then we become a computer-dependent society and computer-independent individuals.

SPEAKER_01 49:16

In some ways, but they'll also enrich our society because it'll make it possible for us to live anywhere we like. Any businessman and executive could live almost anywhere on Earth and still do his business through a device like this. And this is a wonderful thing. It means we want him to be stuck in cities, we better live out in either in the country or wherever we please, and still carry on complete interaction with human beings as well as with other computers.

SPEAKER_00 49:43

And he essentially predicted the internet in three minutes. I used to use it in all my talks. The reason I've chosen my branch of futurism is that Arthur looks at things that are 30 years in the distance. Most of my clients don't have 30 years, so I'm filling a gap in between what Arthur says and what you need to do today.

SPEAKER_03 49:59

That might answer the next question. What's the best prediction or forecast that an entrepreneur, futurist, or a science fiction author has made?

SPEAKER_00 50:08

Go back to Arthur in that we will be connected one day. He said initially via satellites. Uh the fact that we can lay undersea cables and talk at high speed, but I think I spoke to someone the other day, you mentioned before about nuclear fusion reactors that are the size of a container. The fact we'll have high sources of energy available locally will be will fundamentally change communities around the world that don't have access to good clean energy.

SPEAKER_03 50:31

And the last one, is there a science fiction story or a model of the future that's most representative of the future that you hope for?

SPEAKER_00 50:40

I go back to 1984. It's a scary view of what life might be like. And if you reread it, uh there are lots of parallels happening today, which then makes me pause to say, can I and can we be part of the solution by educating our legislators to be aware of the pros and cons of AI and have them use it so they can see firsthand how powerful it can be and also how uh disruptive it can be. Well, Brett, how can we find out more about you and your work?

SPEAKER_03 51:07

Well, you can go to Brett King.com and uh you'll you'll find my information. But of course, I'm I'm broadly available on uh cross social. Um do quite a bit of work on um LinkedIn. You can see some of my uh shorts from stage work and things like that on TikTok and Instagram and uh some of the work I'm doing on um some of my portfolio companies, including Luma Brush Medical, which is a AI smart toothbrush with an intra-oral camera where you can scan and detect oral disease at home, like cavities and missing teeth and things like that. So check out lumabrush.ai for that. And um yeah, um and stay tuned for my next book, Bank 5.0, coming out in November. And you?

SPEAKER_00 51:53

Can't wait for that. If you want to find me, go to actualfuturist.com or just search Digitally Curious. For some reason, I seem to own that phrase. I've also trademarked it. All roads lead to me if you type in Digitally Curious.

SPEAKER_03 52:04

Well, Andrew, it's been a pleasure and it's been good to collaborate. And uh I'll see you in the future.

SPEAKER_00 52:10

I think we've answered the question what happens when two Aussie futurists walk into a podcast? Great things happen. Thanks so much for your time, Brett.