The end of banking as we know it:
Brett King on AI, agents and the collapse of legacy thinking
4 December 2025
In T8’s exclusive interview with world-leading futurist Brett King, we explore how AI’s emerging agency is dismantling legacy thinking across financial services, from the collapse of product-centric banking, to the rise of autonomous agents, to a future where utility, not institutions, becomes the centre of gravity.
T8
Brett, great to have you with us. I want to start with a broad scene-setter. You’ve described AI as “compressed intelligence,” and I’d like to understand how you feel that reframes the relationship between people and machines (machines we traditionally controlled) and also the relationship with institutions that once owned knowledge. Today those institutions are no longer the exclusive holders of that knowledge; at best, they are curators of it. How is the relationship between humans, machines, and institutions shifting as a result of compressed intelligence and AI?
Brett King
Yuval Noah Harari has a great quote on this. The main difference between artificial intelligence and any technology that has come before is that AI will have agency. It will be able to act independently. Yes, it will act on our behalf, but it will still be going out and performing tasks on its own.
“AI will have agency. It will act independently.”
That is a game changer. It's the first time we've had technology that is independent of direct human control and interaction at each stage.
We’ll still be kicking off the agent and receiving its results, and if we don’t like the output, we can change it. But if you look at this at scale, the way the technology develops in terms of competency and its ability to act independently is significantly different from anything that has come before. Agentic AI is the first real game changer.
The second major shift relates to the value of human knowledge. We’ve always had information asymmetry - it exists in your business, in my business, and in most advisory businesses. Doctors advising on health, lawyers on legal contracts, accountants on tax. The reason we go to a doctor is that the doctor knows more about medicine than we do.
But consider the pace of medical advances. This year alone we’ll probably see half a million medical journals published.
“AI is essentially the compressed intelligence of millions of doctors.”
There’s not a single doctor on the planet who could absorb all of that information to diagnose and treat every condition effectively. That’s where AI has a significant advantage. It can look at your symptoms, blood work, radiology and everything else, and make a diagnosis based on all of that data - essentially the compressed intelligence of millions of doctors being applied to your problem.
So now doctors will have an information asymmetry problem versus AI. AI will be far better at giving advice on health and diagnosing conditions than any individual doctor.
In the short term, if you go to a doctor with a complaint that isn’t straightforward - something chronic or complex - you will insist that the doctor uses AI to help diagnose it. But what does this mean in 10 or 20 years? The best diagnostics will be AI doctors, not human doctors.
For now, humans will be paired with AI, but in the future, what will be the doctor’s role? The AI will say, “Here is the diagnosis and the recommended care,” and the human will implement it. And the same applies to legal, accounting, financial services - all of those areas.
T8
So the idea of seeking a second medical opinion almost disappears, because instead of consulting two doctors, you’re effectively consulting tens of thousands. If we map that logic onto financial services, aren’t we essentially talking about the same shift? As long as an individual - or a professional acting on their behalf - is asking the right questions, AI can increasingly deliver answers that would traditionally come from a CIO or a financial adviser. Or do you see the financial-advice space as more nuanced than that?
Brett King
Yes, we’re very much moving toward that model, and we’re seeing it happen in real time. I just did an event in Dubai about a month, month and a half ago, and when we were wrapping that up, we just generated the financial reports with AI. There were a few mistakes, but only minor ones, things that we could correct. Previously, I would have needed a dedicated financial expert for that, and I think the same is going to be true. Analyzing financial reports, analyzing your financial position - it’s done quickly using AI. I use a tool called Manus to do my personal accounts and tax, which I always hated. It's so fiddly and time consuming. I just dropped all my bank statements into an agentic AI engine. Let it do it now, and it's phenomenal how much progress has been made there, very quickly. But the next step, of course, is financial advice. For example, how to invest your money. It’s obviously coming.
“The next step is financial advice…
it’s obviously coming.”
T8
That naturally raises questions about how far we allow technology to make decisions on our behalf. It also brings up the regulatory dimension: what guardrails might be put in place, and could those guardrails ultimately slow or stifle the progress many of us are excited about, while others are understandably nervous about it?
Brett King
Yeah. In terms of investment in artificial intelligence, I don’t see anything on the horizon that will slow it down. So if anyone is hoping for an AI bubble to burst and for humans to suddenly take back full control at that point, I think that’s futile. I think there’s a pretty simple way to look at that. There are two types of people in the world today: those who are using AI to enhance their business or career and are trying to figure it out, and the second type - those who are going to be replaced by the first group.
T8
You make a strong point about momentum. But what about guardrails? Do you see regulators trying to slow or shape the degree to which AI can act autonomously on our behalf?
Brett King
In terms of regulatory function the core issue of alignment really comes down to ethics and the safe operation of AI in our environment. This requires two things, it requires a broad regional or national regulatory apparatus that licenses the use of AI, that essentially mandates any AI that operates and can introduce risk to society is registered. Secondly, we need to propose ethical guardrails where there a clear guidelines and restrictions to how AI can operate in certain conditions. For example, with autonomous vehicles we five levels of autonomy that are possible and each stage the operating model must demonstrate specific performance characteristics safely - keep in mind Waymo has already reduced the incident of serious injuries compared to human drivers by 85%, along with a 96% reduction in intersection crashes (where 27% of fatal crashes happen).
Ultimately, this will be both an overarching national AI regulatory body and then specific regulations in the vertical sectors where specialization is required, along with regulators building their own compliance AI algorithms that can police commercial AI.
T8
One of the other areas we’ve been looking at is how financial-services organisations are actually using AI. From what you’re seeing across the industry right now, what feels real and genuinely tangible in terms of AI deployment? And conversely, what feels like theatre - brand-led signaling rather than meaningful adoption? You must be seeing very different levels of maturity across the organisations you speak to.
Brett King
Yeah, the one thing I see that is broadly gaining traction - and there was a good report on this that came out yesterday, I think from Morgan Stanley; if you look at Ross Dawson’s feed on LinkedIn, he covered it - is that about 88% of S&P 500 companies say they’re using artificial intelligence.
The one area getting consistent traction and showing results across that field is process automation: finding ways to improve internal processes. That is getting a ton of traction. We’ve talked about RPA (Robot Process Automation) for a while, but the tools are very real now. We can see this even in the tech industry with changes in coding approaches and so forth. In process automation, we’re seeing radical improvement.
Secondly, we’re seeing the implementation of semantic AI for initial customer support. There have been some very successful examples, like WeBank in Shenzhen, which eliminated 98% of their human-level support using artificial intelligence – but also through better design. They use AI to study support problems and then design better built-in solutions, rather than just deploying a ChatGPT-type agent to mimic human support. Their philosophy is: if you need to speak to a human, that is a design failure. So they use AI to model and redesign the process, and that has been successful.
We’re also seeing early implementations of tokenisation with JPMorgan Chase, HSBC and others looking at smart contracting. When we talk about smart-contract implementation and stable-coin implementation at scale, we can’t avoid the reality that this becomes an automated, AI-operated business.
That is the next big shift coming - and I would say it’s the biggest disruption to financial services we’ll see in the next 50 years: the automation of day-to-day banking and finance operations.
“There are two types of people today:
those using AI - and those who will be replaced by them.”
T8
You mentioned JPMorgan and HSBC, but more broadly, who among the incumbents do you see genuinely leading in this space? And are they being outpaced by players outside the traditional banking sector?
Brett King
I think it depends on what you mean by traditional players in the banking space. Fintech has already disrupted banking in a fairly significant way. Let me give you the numbers. The top 50 fastest-growing financial institutions in the world are all technology companies. There’s not a single incumbent among them. That’s because the rules have changed: digital acquisition is now the primary mechanism of growth for financial-services organisations globally.
If you’re still relying on a signature on a piece of paper, then your business is being given away to someone who can onboard and engage customers digitally. That is already very clear.
Here’s how we know. If you take the top 20 retail fintechs in the world - Nubank, Revolut, WeBank, Stripe, PayPal, etc. - and compare them to the top 20 retail banks - ICBC, Agricultural Bank of China, JPMorgan, Wells Fargo, Bank of America - you’ll see the top 20 fintechs have 4.1 billion customers today. The top 20 retail banks have 2.7 billion. It’s not even close. Fintechs have already won the day-to-day banking battle.
People will say Alipay isn't a bank, it’s a mobile wallet, so it doesn't count. But in 2018, they had the largest deposit pool in the world with Yu’e Bao, sitting on top of Alipay’s wallet. It had $300 billion in deposits. Yes, that shrank because of currency and product controls, but the point remains: mobile was proven to be the best form of deposit capture, not branches.
“The top 20 fintechs have 4.1 billion customers today. The top 20 retail banks have 2.7 billion.”
Look at Nubank - 120 million customers. Their nearest competitor, Itaú, has 58 million, and they’re an 80-year-old bank. Revolut in Europe has 50 million customers as of November last year; probably 52 or 53 million by now. HSBC is a 160-year-old bank with 38 million customers globally. You see this happening everywhere: digital players are capturing market share. And that’s before we factor in AI-based banking.
So then the question becomes: who will be best in the world at implementing AI in financial services? Will it be an incumbent that first has to consolidate its legacy tech stack before AI can even run on top of it? Or will it be an incumbent fintech with ten years of development and a cloud-native, AI-capable stack that can plug AI in at the cloud core from day one?
Which of those organisations is better positioned to absorb AI? Which has the cultural agility to implement AI across the company, versus still debating internally whether branches or digital should get the lion’s share of investment? That debate is still happening in places like the US.
I had a DM yesterday from a researcher in the banking industry saying, “This European bank has just doubled down on branches. They don’t like your message in Branch Today, Gone Tomorrow. They think branches are still strong.” And I said, “If they think that, go for it - but they’re not looking at the data.” If that’s their market strategy, they’re not looking at data.
And someone will say, “Yeah, but Chase is deploying tons of new branches.” Sure - they’re right-sizing their branch network. They’re moving out of rural locations and into cities, and they’re creating smaller branches. They’re downsizing total square footage. But at the same time, they’re investing $18 billion a year in technology.
So what do you think they believe the future of the bank is? Do they think it’s branches? Or technology?
That cultural agility is a key element in the AI transformation, for sure.
T8
What I’m hearing is that incumbents - the traditional banks - are essentially plugging AI tools onto ageing platforms, whereas the neobanks have AI embedded throughout the entire customer experience. For them, AI isn’t an add-on, it’s intrinsic. It’s present across the whole ecosystem, which allows them to join up the experience far more quickly and effectively than a bank taking a fragmented approach. Is that a fair way of characterising how things look today?
Brett King
I think that’s certainly a fair observation. Let’s look at Nubank as an example. They have 120 million customers. They acquired Hyperplane in 2023 to build out their own generative-AI capability. They have enough data - over 100 million customers across Latin America - to train their own generative finance model for the region. None of the banks in Brazil or Colombia have the customer scale to do that. Alipay would have the data advantage in the same way.
That’s a key element. And when you look at how their business is built - the way they think about technology, deploy it, and try to figure out what comes next - players like Nubank and Revolut are just far more technically agile. There’s a general acceptance from the exco down to junior staff that technology is embedded in the business. It’s core to what they do. So when AI comes along, it’s just another technology to deploy, and they’ll experiment broadly and simultaneously on multiple fronts.
Whereas in most traditional banks, AI is still treated as a project that sits with the innovation department. It’s not embedded culturally across the organisation as a capability or a mindset.
That said, if you go to some of the larger banks and ask how many staff use AI, the number is actually pretty high. If I went to Emirates NBD here in Dubai, I’d say around 70 percent of staff use AI daily. But if I went to a midwest regional player in the United States and asked the same question, it might be five percent.
So the question becomes: how do you develop that cultural agility? How do you get that level of immersion in AI across the organisation so that people understand where AI fits and how it can reshape the business - if it isn’t something everyone is already experimenting with? “There are two types of people today:
those using AI - and those who will be replaced by them.”
“Technology is embedded in the business. It’s core to what they do.”
T8
Tell me if I’m looking at this the wrong way. Many of the regions you’ve mentioned – Brazil and Latin America, the UAE, parts of Asia - seem to be making some of the most significant leaps because they’re not burdened by legacy systems. Meanwhile, Western markets and established global banks often appear to be moving more cautiously.
What, in your view, is slowing incumbents in the West? Because the need to accelerate must be clear, and institutions like HSBC, JPMorgan and others surely understand the competitive pressure. You referenced some remarkable scale numbers earlier from newer players – numbers I suspect many people still underestimate. So is this really a question of legacy constraints, or something deeper?
Brett King
That’s why I talk about it the way I do - because it has already happened. Ten years ago, when I was talking about the potential of fintech, the arguments were always the same: “They’re only growing fast because they’re small. They’ll never be profitable. They’ll never manage risk as well as banks. And they’ll never have the asset base that we have.”
Well, the first three arguments have already disappeared. Nubank and Revolut are still growing faster than any of their European or Latin American incumbent competitors. They are highly profitable and among the most operationally efficient banks on the planet. And they manage risk far better than traditional players because risk is fundamentally a data-modelling problem.
So what’s left is asset size. And I turn that back on the boards I speak to: tell me again how your asset size helps you become an AI-capable bank? Because the reality is you want to protect that asset size for as long as possible, and anything you do with AI is interpreted as a risk to that base. When those assets start shifting, they’ll shift very quickly to the technology side, and incumbents won’t be able to protect that position if they don’t have the technical or cultural agility.
That’s the short summary. In terms of how incumbents adapt – it’s tough. Chase is doing it because Jamie Dimon understands this, and they can spend $18 billion a year to make that transformation. But how many banks have that level of investment capability? And that’s before we even talk about new AI-based banks coming over the top, and fintechs evolving into fully AI-driven financial institutions.
“Fintechs have already won the day-to-day banking battle.”
T8
That leads to another question: do you see the sheer scale of new entrants pushing the market into even greater fragmentation? In other words, is this going to become an increasingly competitive environment – and therefore even more challenging - for conventional banks?
Brett King
Absolutely, but the reason isn’t AI specifically. It’s changing customer behaviour. And this is something I don’t think many bankers fully appreciate yet, although once you hear it, it’s obvious.
Gen Z and Gen Alpha don’t think about banks the way their parents do. Gen X thinks in terms of a primary bank relationship, with maybe a secondary bank for a specific need. Younger generations don’t think like that at all. They think about banking the same way they think about apps on their phone.
“Oh, I want to buy a new car - what’s the app for that?”
“I want to save more money - what’s the app for that?”
At the moment they have a task, they look for the best utility they can find. They don’t have the concept of “this is my bank, and this debit card is where I go for everything.” That idea is foreign to them.
So fragmentation will naturally accelerate, and it will be based on utility, not brand. These younger customers don’t trust based on brand. Having more branches on street corners doesn’t make HSBC (or any bank) more trustworthy to them.
They trust based on functionality:
“My friends say this works.”
“I can send money instantly and flawlessly.”
“Wise works better for overseas transfers than a bank.”
“Revolut works every time. I can Revolut you the money while I’m in Europe.”
Revolut has even become a verb.
And that utility-driven behaviour is what’s reshaping market share in financial services.
There’s no brand-led trust left - it’s utility-led trust now. “Fintechs have already won the day-to-day banking battle.”
“Younger customers don’t trust brands - they trust utility.”
T8
So we’re really talking about fragmentation driven by behaviour rather than by product or market structure. But it raises a question: how do these younger generations manage their money holistically if everything is so distributed?
For someone like me – older, more used to having everything organised in one place with a few core products from a single bank – that fragmentation feels counter-intuitive. So how does this model actually work for them?
Brett King
You’re going to be different as well in the future – and I’ll tell you why. You’re going to have your own personal AI agent managing your money. It’ll be far more effective than a financial adviser in a bank, and much more personalised. It will be able to generate better returns, and it will have the equivalent of high-frequency trading capabilities internally.
It will manage risk for you. It will automatically find the best portfolio components to invest in based on the algorithm you’ve trained it with. Within the next five to seven years, you’ll be completely comfortable with your personal AI agent managing your finances – and you won’t really care what’s behind it.
So AI agents are going to further fragment the landscape.
“AI agents are going to further fragment the landscape.”
T8
That’s really interesting, and I’m glad we’ve reached the topic of agentic AI, because it’s the biggest theme at the Singapore FinTech Festival this year. What I’m trying to understand is this: if AI agents become the primary interface for managing our financial lives – and potentially much more than that – does the core relationship shift away from the bank?
Who will actually provide these agents? Will they be built by individuals, offered as a service by third-party companies, or embedded by banks and operating-system providers? Ultimately, who owns the customer relationship when these agents start managing our broader lifestyle, not just our money?
Brett King
ChatGPT and Anthropic would obviously like to think their agents will be the ones you use. But I think this will become OS-based. The agents embedded in your personal wallet, in your operating system, on your phone or whatever device you use – that’s probably where this is heading. Then you’ll have cloud-based AI, like the ChatGPTs and Anthropics, that work across different operating systems as well.
Yes, you’ll also see agents built into apps – modern financial-services apps will have your AI financial manager. But over time, your personal AI will have to manage far more than your finances. It’s going to manage your health. It’s going to manage your household expenses and your household budget. It’ll help you with all of those aspects. It’ll certainly help with savings, and it’s going to become much more personalised over the next decade or two.
And then there’s a very real question about how money itself will work in the 2040s and 2050s in highly agentic economies. But that’s a more augmented discussion we can have separately.
T8
Building on that - if we have personal agents and banking agents - are we essentially moving toward a world where agents talk to agents, and we no longer need to engage with the bank directly at all? It’s machine-to-machine?
Brett King
Absolutely We’re effectively months away from that. Within the next two years, I think it will be a reality for many people that you simply won’t speak to a human at a bank again.
And let me remind you of that quote from Henry Ma at WeBank: if you have to speak to a human, that’s a broken design. In a modern bank, speaking to a human is no longer an asset – it represents friction. It slows down what you need to get done. From a financial-services perspective, it’s not an asset anymore.
“Within the next two years, I think it will be a reality for many people that you simply won’t speak to a human at a bank again.”
T8
It really is extraordinary. You once said banking used to be a place you went, then it became something you did. But now, with agents stepping in, it feels like it’s not even something we do anymore. So what does banking actually become in that world?
Brett King
It becomes something your AI does. It becomes pure utility.
As I’ve said before, if you look at what a bank fundamentally does - whether it’s retail or corporate banking - there are really three core functions: it safely stores your money, it safely moves your money, and it gives you access to credit. That’s the core utility you get from a bank today.
Right now, banks wrap products around that utility layer. But the real advantage is simply having that core utility accessible and streamlined, without all the product wrappers. So you’ll say, “Help me move my money. I want to send €50 to Harry.” Your AI will say, “Alright, I know the best way to do that. Let me take care of it.” You won’t have to think about sort codes, transaction fees, or which service to use. It will just work.
That utility is what comes to the forefront in an AI-driven age.
Banks have spent a long time training us to think we need financial-services products. But in an AI- and technology-led world, you don’t need the products – you just need the utility. Fintech has already proven that, and it will be even more true in the AI age.
“(Banking) becomes something your AI does. It becomes pure utility.”
T8
Earlier you mentioned that the real breakthroughs from banks are happening in the “invisible layers” - things like fraud, KYC and internal process optimisation – where AI is already proving genuinely valuable. But it made me wonder: are some banks still investing heavily in areas where they’re unlikely to win, particularly customer-facing agents, given everything we’ve just discussed about personal AI ownership and agent-to-agent interactions?
In other words, do you see banks directing energy toward parts of the value chain where they may ultimately lose out to the very AI agents we’ve been talking about?
Brett King
Well, in my earlier book Bank 4.0 - the one before Branch Today, Gone Tomorrow - I wrote about embedded banking. We’ve all heard that term, but I still don’t think many bankers truly understand it. Embedded banking means you want your bank to be integrated into your customers’ lives, solving their problems when and where those problems occur. You want the utility to be available at the moment of need.
That’s not what we see from most banking apps today. Banking apps still reinforce the traditional product structure. But if you look at a Nubank or a Revolut - and Revolut is a great example – they’re building utility into daily life. They’re enabling actions, not pushing products.
“Banking apps still reinforce the traditional product structure…you want the utility to be available at the moment of need.”
Here in Dubai, Wio is another good example. If you want to invest or save, they help you do it as easily as possible. They embed that core utility into your life and make it effortless to access. They don’t say, “You need a credit card department for that,” or “You need to speak to an investment advisor.” They say: How much do you want to invest? Weekly or monthly? Let’s ask a few quick questions – okay, you’re good to go.
That embedded ability to be in your life and facilitate what you need, when you need it, is the core. Banks haven’t adapted to that because their organisational structures won’t let them. Those structures reinforce product silos. Revolut doesn’t have a credit-card department sitting in a separate part of the building. It doesn’t work like that. From a technology perspective, they’re always looking at new ways to deliver core utility.
T8
You touched earlier on how these agents – or platforms like Revolut – proactively approach us, ask a few simple questions, and then take action on our behalf. In that world, the traditional argument around “financial literacy” almost disappears, because the complexity is abstracted away.
It made me think of something you recently posted about the cognitive impact spectrum of AI, and one of the layers you highlighted was the risk of us becoming passive – where everything is done for us and we stop learning or thinking about how our money works.
So how far should we allow that to go? At what point are we handing over too much responsibility to the agent? And is there still a role for financial literacy when so much of the decision-making is automated?
Brett King
Financial literacy is a lie. And I’ll tell you why: financial literacy is based on the idea that you need to understand the products we’ve created in order to access banking services. That already doesn’t exist.
This is why Nubank and WeBank have captured huge numbers of traditionally unbanked customers. Nubank is a great example. WeBank as well. I think about half their customers are first-time banking customers, and Nubank is at least 40%, if not 50%.
And it’s not because those customers suddenly became financially literate. It’s because these banks designed away the need for literacy. They solved the problem with better design. They’re not saying: “You must understand the rules of the banking system to bank,” or “You have to understand Sharpe ratios, beta, alpha, or meet accredited-investor criteria to invest.” Those concepts are vestiges of the old system.
So if you say financial literacy or financial education is the answer, you don’t understand what’s coming.
So essentially, we swap that out for financial inclusion. We are being empowered to be part of the financial system without needing to understand the old rules. The question becomes: how do I give you access to the utility in a way you understand, and in a way that is financially healthy for you?
“Financial literacy is a lie.”
T8
All of this is fascinating, especially for companies in the design world creating beautiful experiences, screens and journeys. What you’re describing fundamentally changes that. There’s nothing linear or screen-based about these new interactions. They’re not “journeys” in the traditional UX sense; they’re orchestrated moments - micro-moments - rather than structured flows.
So what does that mean for the design industry more broadly? It feels as though emotional engagement, behavioural insight, and consultancy-style thinking become even more important. And design itself becomes disrupted. We don’t know how soon, but pace isn’t gradual - it’s vertical now. Things are changing incredibly fast. How should companies in the design space think about their role in that future?
Brett King
This is tough for people to understand. But here’s the really cool part: there’s an entirely new design industry emerging around these contextual moments.
One area that’s going to be really interesting is the development of mixed-reality smart glasses – interactive content with your AI engaged and content coming into your field of view at the moment you need it.
“If you have to speak to a human, that’s a design failure.”
You’re not going to want the old experience. If you’re doing a complicated banking transaction or a complex engagement where you’d normally need to lay out an extensive process, you’re not going to pop up app screens in a head-up display and use hand gestures to manipulate fields and choose options. That’s not how it’s going to work.
You’ll need a completely new design language – one that surfaces exactly the right content in the right moment and interacts with the user as needed.
T8
So that raises a design question, but also a human one: what happens to us in this shift? I caught a moment in one of your recent videos where you mentioned that JPMorgan’s employee numbers could fall from around 130,000 to 8,000 over time - which is an extraordinary thought.
What did you mean by that, and how should we think about the human and workforce implications of this acceleration?
Brett King
Yeah. I think technology and employment is the big issue, and we’re only just starting to grapple with it. If you look at computer-science job placement in San Francisco, that’s already an early warning sign. With the level of investment going into AI, the contraction in the job market is going to be severe over the next five to ten years.
We’re going to need something like UBI (Universal Basic Income) as a safety net for people who are displaced. And we’re nowhere close to ready for that - particularly in markets like the United States, where UBI is labelled socialism or communism. But the alternative is far worse. Look at the panic over 30 million people being on SNAP (Supplemental Nutrition Assistance Program) benefits. Imagine that number rising to 150 million. Imagine half the population on unemployment benefits. The social strife would be immense, and there would be huge resistance to solving it.
This is the biggest challenge we have with AI. The problem isn’t making money. The problem isn’t producing incredible experiences or fixing hallucinations. The problem is that AI will structurally change the way we work as a society, and we’re not ready - economically or from a policy standpoint.
When people ask me if I’m scared of AI, I say: I’m not worried about AI. I’m worried about human adaptability to AI.
When people ask me, Are you scared about AI? Are you worried about AI? I say I'm worried about human adaptability to AI. I'm not worried about AI itself.
T8
One thing we’ve been discussing internally is how design careers evolve in an AI-native environment. Many organisations are already using AI to generate basic journeys or first-pass concepts, which shifts designers toward more senior roles - guiding, validating or shaping the AI’s output.
That raises an interesting question: how do junior designers grow in this new world? How do they build foundational experience if AI handles so much of the early-stage work? And what does the path to becoming a senior designer look like when the traditional apprenticeship model is disrupted?
Brett King
I think you’re going to have juniors working with AI agents side by side — learning from AI and learning how to interact with AI. That will essentially make them AI-native seniors over time.
Ultimately, there is still a role for humans to shepherd AI through these processes — AI whisperers, if you like. We need people embedded in organisations who can do that. But you’ll simply need far fewer people to get things done than before. That’s the challenging shift.
As we talked about with my illustration of JPMorgan Chase going from around 140,000 employees to five or eight thousand – that’s the scale of disruption possible, particularly in heavily process-driven industries. If a company is built on process, all of that can be automated, and the people managing those processes today can be largely eliminated. That’s where the hardest challenges arise.
But when we start talking about creativity and other elements like that, there will still be a need for human interaction and working side by side with AI. That’s true collaboration. The shift is from using technology as a tool to embracing collaborative intelligence.
T8
If I can zoom right back out to a much higher, helicopter-level view before we close – there’s something you’ve said elsewhere that really struck me. You wrote that capitalism runs on money, while AI runs on energy. I found that idea mind-boggling, and I’ve been thinking a lot about what it implies.
So at the broadest possible level: what happens to capitalism in an AI-driven world?
Brett King
No, capitalism won’t survive AI. It’s incompatible with AI. I don’t know whether we’ll end up talking about energetic economics or something like that, but remember: we measure data centres today in gigawatts of energy - one-gigawatt, five-gigawatt data centres.
In markets like the US and UK, we’re going to need a 40–50% increase in energy just to power automation at the scale we’re discussing. China is the only nation right now investing in energy transformation at that scale. That’s why China will become the world’s first major smart economy, even though the US has players like ChatGPT, OpenAI and so on.
Here’s the core issue: the more automation you put into the system, the less utility money has. As the system tokenises, money becomes just ones, zeros and digital tokens. Over time, money loses its utility – which is one problem for capitalism.
But the bigger problem is wealth distribution. If you have large-scale technology unemployment and you’re paying everyone Universal Basic Income because AI has taken the work, then go back to the very first page of The Wealth of Nations by Adam Smith. The opening paragraph is about wealth distribution – capitalism as the most equitable way to distribute wealth through labour.
Capitalism is built on the idea that human labour is the core of economic value. But if AI eliminates human labour from the workforce, then capitalism simply becomes incompatible with the future.
“Capitalism won’t survive AI. It’s incompatible with it.”
T8
Well, that is quite the note to end on. Brett, thank you – as always – for the clarity, candour and the provocation. It’s been an incredible conversation.
What’s next
If your organisation is still exploring how AI reshapes your customer experience, or how agentic intelligence should lead it, talk to T8.
We’re helping banks move from digital-by-design to AI-native, building experiences that think, adapt, and act.
Brett King
(Top-10 Global Futurist, Author, and Relentless Challenger of Legacy Thinking)
Brett King is one of the world’s most cited financial futurists, a bestselling author, founder, and the host of the award-winning Breaking Banks and The Futurists podcasts. His influence stretches across continents, from advising global tech firms to serving as a policy voice for governments including the Obama administration, Xi Jinping’s leadership group, and key GCC ministries.
With over two decades at the forefront of financial innovation, Brett has become the industry’s leading provocateur on the future of banking, AI-driven economies, and the collapse of legacy systems. His work predicted the rise of digital-first banks, mobile-wallet dominance, and now the age of agentic AI.
Brett’s books, including Augmented and the landmark Bank 4.0, have reshaped how executives and policymakers think about technology, society, and the evolution of value. His latest work, Branch Tomorrow, explores the rapid disappearance of physical banking and the emergence of AI-native financial ecosystems.
Known for his clarity, courage, and humour, Brett continues to challenge institutions to evolve, or be overtaken by those who already have.
