Since the launch of ChatGPT and the mass popular adoption of chatbots, Large Language Models, and generative tools in the cloud, much social coverage about Artificial Intelligence has focused on faddy or frivolous uses, and on the controversy around copyright theft.
It is not hard to see why: to persuade people to buy into the technology at scale, vendors saw an easy, if cynical and exploitative route: offer instant creativity at low or zero cost, and thus create a dependency culture. In the process, the work of skilled human creatives got trampled on – after feeding the AI beast without credit or recompense.
Put another way, AI vendors have given users the illusion of effort-free, professional-grade talent for a few dollars a month. What lazy primate could resist that?
In this way, AI promises to be the great leveler and equalizer, claim vendors, transforming every sector of the economy, while curing diseases, reversing climate change, and creating unparalleled productivity growth.
Yet there is little sign of that transformation happening in the real world, let alone any measurable payback. Meanwhile, a great deal of bad AI art, shoddy AI text, ersatz AI music, and lame AI video stands in place of real gains for society – plus some useful summaries of complex topics.
That growing mismatch between vendor claims and reality may be why AI-focused investment funds have been underperforming against other tech sectors over the past 12 months, and in some cases falling against their performance last year –for more on this, see my March reports.
If another AI winter is beginning, therefore, vendors are partly to blame for trying to play both sides: promising Artificial Superintelligence while pushing fake cartoons and instant pop songs – just 2025's equivalent of ringtones. Lost somewhere between those extremes is the average enterprise.
FinTech and AI - promised transformation or empty hype?
So, what of an industry that promises to be transformed by AI: Financial Services? One of the UK's technology hotspots – alongside robotics, quantum innovations, and AI itself – financial technology (FinTech) stands to make measurable gains from AI. At least in theory.
That was the subject of a recent Westminster policy conference – a Business Forum rather than the usual eForums on the topic, suggesting that the focus would be on enterprise gains and adoption rather than on recycling vendor marcomms.
So, was that the case? Hmm.
Stanford alumni Dr Pinar Ozcan is Professor of Entrepreneurship and Innovation at the Saïd Business School at the University of Oxford. She is also Academic Director of both the Oxford Saïd Entrepreneurship Centre and the Oxford Future of Finance and Technology Initiative, and has advisory roles at the Bank of England, the UK's Competition and Markets Authority (CMA), the Financial Conduct Authority (FCA), the European Commission, and the OECD.
An industry heavyweight by any measure, therefore. Giving the keynote, she said:
So, what does AI do for us across industries? That's a question I've been asking myself for a while.
One of the examples is that Amazon trusted its AI so much that it filed a patent in the US to do something quite unusual. Rather than us shopping for what we want, and Amazon shipping it to us, the company trusted its ability to predict our preferences. So much so that it wanted to start a service whereby it ships a bunch of things to us, and whatever we don't like, we return.
The only reason that's not in effect today is because Amazon still cannot use drones legally, because if they drop boxes on our heads, that's a big problem. But other than that, the AI is ripe enough to predict everything about us and sell us things that we didn't think we would want, but when we see them, we will want them!
Really? My own experience of Amazon's algorithm – and Meta's too – is largely, 'So, you've bought some shoes and a kettle. Here are some similar shoes and kettles.'
That aside, the fact that a Stanford-educated Professor at Oxford University – speaking at a Westminster policy conference, no less – is pushing such concepts is baffling. At a time when most families are struggling to pay their bills, the idea that the mass delivery of unordered non-essential items by swarms of miniature helicopters would be a win for society suggests one thing: that our policymakers have lost their marbles.
Sadly, she is not alone. Some politicians have become the unquestioning marketers of trillion-dollar corporations too. Take Feryal Clarke, Parliamentary Under-Secretary of State for AI and Digital Government, who last month stood onstage at another Westminster conference and claimed that AI would usher in an era of "infinite productivity" and mass equality, in which all parts of society stand on a level playing field.
This, I suggest, is a delusion; thirty years of tech innovation have passed since the popular uptake of the Worldwide Web, and UK productivity growth has been flatlining for nearly 20 of them. Factor in Bill Gates' recent assertion that nearly every job on Earth will be replaced by AI, and it is clear we need grown-up policymaking and strategizing for the future – urgently – and not this cultish claptrap.
But alas, the signs are that a lack of critical thinking is now pervasive in Westminster. Palo Alto-based composer Ed Newton-Rex, who is fronting the creative communities' fight against AI copyright theft, recently shared a letter from the UK's Department for Culture, Media and Sport (DCMS).
It was a reply to his protest about Downing Street's plans to tear up UK copyright conventions. Rather than engage with the issues he raised, the letter read more like a Microsoft press release on why copyright rules must be changed. Indeed, perhaps it originally was one, given that the vendor has placed AI policy advisors at the heart of British government. (Clever Microsoft.)
Recent academic studies – including, ironically, one from Microsoft Research – have warned of a collapse in critical thinking as our AI dependency grows. Clearly, the collapse among policymakers has already happened.
(Newton-Rex, incidentally, is no Luddite. He has worked at Stability AI, Snap, and Jukedeck, and helped develop TikTok's music algorithm. As founder of new organization Fairly Trained, he simply wants creators to be treated equitably.)
Financial inclusion or corporate opportunism?
But back to Financial Services. Did the eminent Professor have any insights about AI's impact there, given that she was giving the keynote?
She said:
An Oxford graduate went to work for a bank in Dubai and realized that the service people around her – people who were washing cars or waiters in restaurants – were living in poor conditions and were not financially literate. They were losing half the money they earned while trying to send it to their families in Bangladesh, the Philippines, India, Turkey, Iran, and other countries.
So, what these founders did was train chatbots in the languages that those service people spoke. They created an app that talks to them, many of whom may have difficulty reading and writing. Through chatting in their own dialects, they were able to educate them about their finances and create ways for them to send money securely, or to save money by putting it in an electronic account, rather than under the pillow.
So, some of the improvements that come from AI are incredible.
She continued:
Another at Oxford is the founder of GoHenry, Louise Hill, who can use AI to train teenagers in the value of money and make them understand that it doesn't just fall from the sky; we need to learn how to manage it.
Another one here is a company that tries to make sure that everyone can use AI even in Financial Services – that you don't need to be a data scientist. The examples are many, but what I want to bring forward is that, while a reduction in costs is the mechanism, what comes from that is something wonderful, which is financial inclusion.
Perhaps – though the average teenager, swimming in debt and facing a future of stratospheric rents and unachievable mortgages – understands that the only money that might fall from the sky one day is their parents' equity. Designing financial products for them, therefore, needs to acknowledge their reality, not patronize them.
Therefore, a cynic might observe that the kinds of application Ozcan talks about are also means for banks to upsell services to vulnerable people. Bear in mind the recent PPI scandal, not to mention banks' industry-wide market rigging – Libor, Forex, and more – for which they became notorious earlier this century.
Ozcan continued:
It means that we're able to reach millions of people with financial services, whereas in the old world, it was way too expensive for a bank or any other financial institution to do so.
So, we're able to use AI to not just talk to them with chatbots and understand their needs, but more than that, we're able to create financial products with Gen-AI that are exactly tailored to their circumstances.
Data is at the heart of all of this and, as an economist said a few years ago, data has become the world's most valuable resource. But in my opinion, though it's depicted as the new oil, it is much more valuable than that because it's something that becomes more valuable and gives more value, the more we have of it, because it gives us new insights.
Then she added:
Amazon throws digital confetti whenever you put something in your basket. So, next time you're on Amazon, I would love for you to notice that, subconsciously, this is also motivating us and making us feel good about having bought something else!
So, we also need to understand that the arrival of AI in Financial Services means even greater Big Tech power, as well as, perhaps, the amplification of biases.
For example, for every five men that get a business loan, only one woman gets one. So, if I train my app with that information, then I'm going to be multiplying that bias, then giving business loans to men only.
My take
An odd and dispiriting keynote: lots of PR flag waving, of the kind that helps no one make sensible decisions. Then, right at the end, a serious point that was left unresolved. None of this is helpful, people. We urgently need to raise the standard of our public debates on these subjects.
