PassW0rd – 12 June 2024 (Will AI Make You Redundant?)

This program is brought to you by Resonance FM. If you like what you hear, please support our work by making a donation at resonancefm.com/donate.

Hello and welcome to Password on Resonance FM with me, Peter Warren. In this month’s program, Will AI produce your P-45? We investigate the fears that the revolutionary new technology may take our jobs. In a series of interviews with leading technology figures and industry experts, we find that the future is quite literally, at the moment, in our hands. According to the tech titan Risto Silasma, The man credited with turning around the fortunes of the former phone giant Nokia, who has made his own fortune with a number of high-tech startups, including the leading cybersecurity company WithSecure. AI will be a boon to humanity. Nearly a decade ago, at his own expense, Salazar took time out to learn how to program AI systems, and then, armed with that knowledge, went around the world on a mission to educate world leaders and decision makers about the potential of the technology and how it works. Here’s Salazar on the AI epiphany that has definitely given him a job.

Somebody said that, you know, if we had taken the, the airplane that the Wright brothers built and decided on the future of aviation based on the, the safety of that design, we would never have airplanes, passenger planes like today, which are very, very safe. And that’s what engineers do. We take a preliminary model and we continue improving on it until it fulfills the criteria that was set for it. And that’s what we need to do with AI. And the only way to get that done is to start using it. And of course, we need to be smart about how we use it. We shouldn’t have built a passenger plane back in the Wright brothers’ days, and luckily no one did. But we continued using the technology where it worked, and that’s what we need to do with AI. And there are many things that it does very, very well. Many things that it does better than human beings can do at the moment, especially many things that it can do better than the average human professional in that field of work. And it can do many repetitive things very well, things that we find boring. And when— what we find boring, we typically don’t do very well. And there’s so many opportunities when you get an organization into the right frame of mind, looking at what the technology can be used for and experimenting and learning. And lots of good things can come out of that, especially in the Western world where we have a huge demographic issue and the population pyramid looks really nasty, and we need to automate a lot of things. We need to reduce unnecessary work in many areas, and we need to focus the human hands and brains on things that only humans can do.

That’s the point, isn’t it? What AI promises to do is change our economy. It means that the economy that we have previously been used to, this 20th century economy where there were a lot of boring repetitive tasks, they’re going to go because the AI can do them.

Yeah, exactly. You know, I did a small desktop study using online information to try to understand what’s the value of technology, and there’s no perfect way to calculate that, but as a proxy I took the cumulative or the total value of all publicly listed technology companies in the world. And in 2014, according to the information I found, and I cannot verify that it’s correct, that number was about $5 trillion in 2014. 10 years later, 2024, it was $24 trillion. So a growth of over 400%. At the same time, Again, according to the best information I found, all the other business domains grew by about 40+%. So technology created value 10 times faster than all other domains. And if we think about the next 10 years, now that so many new technologies have just come online, AI included, and perhaps the most important one,, but lots of others like CRISPR for biotechnology. Quantum computing is still in the future, but during the next 10 years it will be usable. It will be sort of online and so forth and so forth. If the same absolute value generation happens, that would be $19 trillion, and that’s not that much. If the same sort of relative growth in value happens, that would be over $60 trillion of new value from technology, and a large part of that from AI. If the same progressions, the same speed of growth or the acceleration continues, the number would be so astronomical compared to the size of the world economy today that the world economy would be completely unrecognizable in 10 years. So being conservative, let’s assume that it’s at the same speed will continue, not the same acceleration, but the speed. And that would mean $60 trillion of new value. That’s a very meaningful number. And all countries, all politicians should be interested in how do we grab more than our share of that pie.

Because that’s one of the things that we’re going to have to do, isn’t it? Because there are going to be structural changes to the economy. We’ve got to think about whether there is going to be a universal basic income, or how we’re going to share out that information, not just in those states and those companies, but to individuals because of these transitions.

Exactly. If we implement solutions such as universal income, or we commit ourselves to something like that, it reduces fear. And fear, fear is the mind killer. It’s the killer of courage. It’s the killer of experimentation. And if we are all afraid of the future, we just move slower. And we stagnate. And then some other economic area that has less fear experiments more, learns faster, evolves faster, and will take a bigger share of that pie that we, we just talked about. And that’s why it is important to create the kind of an environment where innovation floor issues. And one part of that is to reduce fear of the future.

Many of the people that we’ve been talking to have said that this means that we have to grasp some significant challenges in areas like education, for example, that we have to change the way that we think, the way that we’ve learned hitherto, the models that have essentially been that we build specialist knowledge, which we’ve done since the Renaissance, and we do that, we have lectures and the way that we teach, we’ve actually got to teach people to think to be able to interrogate the AI systems.

I think we should be teaching people to think even without AI. And I think we should be teaching people how to learn because it’s kind of odd that in the primary school or primary education, there’s typically never a subject focused around learning tools, personal tools. How do I— how does my memory work? How do I train my memory? How do I train my ability to solve logical problems? How do I train my capacity We just focus on learning math and we focus on spelling and we focus on reading geography and biology and history instead of learning how to do that better. And learning how to think is a great reason to study programming, because when you do programming, you learn the fundamental element of thinking, i.e., You learn to break a big problem into smaller pieces that you can solve one by one. And suddenly something that felt completely insurmountable feels quite solvable. Maybe it’s still challenging, but solvable. And that’s something that is not limited to just programming. A lot of other professionals need that, like doctors. All sorts of symptoms with the patient, and you need to sort of break it down to components that you then figure out one by one, and you connect the dots and you understand what ails the patient. Or lawyers, basically a contract is like a program. It describes exactly what should happen when A is true and B is not. Which is exactly what a program does. So a well-trained lawyer and a well-trained programmer actually learn to think the same way in the end.

The tech billionaire Risto Salasma, who embarked on a one-man world tour to enlighten world leaders. Salasma is not alone in seeing good in the technology. The futurist David Wood, who thinks that breakthroughs in medicine aided by AI could soon see the banishment of death, a topic Wood expands upon in his book The Death of Death, says the potential of AI technology and the jobs it can create are huge.

It’s changing so fast that nobody really fully understands what it’s capable of. What it isn’t capable of. There’s a lot of hype because people are excited. There’s a lot of substance because frankly, there is remarkable capabilities in there. And the most frightening thing is that people can use AI to improve the development system. So lots of people are being put out of work, and AI engineers will be increasingly put out of work as a lot of what they presently take a long time to do will fairly soon be done much more quickly by AI itself. And what will we end up with? Nobody can be sure, and that’s why we need to stop and think very seriously.

That thing about stop and think, that’s going to involve a lot of politicians, and unfortunately there are not a lot of politicians who really understand very much about this, are there?

Well, it’s always the case that politicians aren’t experts in every field, And what good politicians do is they find suitable advisers. And frankly, there are some good advisers around most of the political leadership around the world. Not all of them, by any means. There are people who are meeting quite regularly to say, you know, we shouldn’t really be racing about this. We know that by default there is an intense race on because the company that who succeeds with new AI is likely to become the most valuable in the world. Let’s just look at Nvidia. As we are recording this, they have overtaken Apple, I think, to become the second most valuable company behind only Microsoft. And as AI becomes more successful, the companies who are behind it will thrive. It’s not just a commercial race, of course. It’s also a geopolitical race. Do we want the world to be dominated by Chinese-based AI systems, or Russian-based AI systems, or American-based AI systems, or whatever? So there is this incredible race. But people around— most politicians realize, actually, this is counterproductive. It would be better if somehow we could collectively slow down. And what we’re going to see in the next 12, 18 months is whether that agreement amongst advisors is going to succeed, or whether instead it’ll be a race, a sort of a suicide race in the worst case, to who gets in charge of this very explosive technology first.

I mean, that’s going to be the issue though, isn’t it? Because some companies are going to say you can’t hinder our development because you’re stopping us from being able to achieve these wonderful prizes. Also, governments are going to say, we’re not interested in becoming involved in this mitigation because we see great advantages to us. I mean, there’s an interesting paradox in this as well though, which is that the Chinese don’t really want to see the development of an AGI because it’s a very control-based economy that the Chinese are engaged in, and the Chinese Communist Party don’t really want to see themselves replaced by AI.

Exactly so. So the Chinese, like I think many other sensible people in a way, they want to get the benefits of, we might call that a narrow AI, without letting the whole genie out of the bottle, an AI which will leap beyond our control. So can we get the best of what I call advanced AGI minus, advanced AI minus, which has many of the capabilities of humans, but doesn’t have free will, doesn’t have volition, doesn’t have a sense of drive and identity and purpose. If it remains as a tool, then there’s more of a chance that it can be used for genuinely valuable things like finding better cures for cancer, finding better ways to manage green energy, nuclear fusion perhaps. So we all want that without having the system going out of control. But that’s what politics is about. Politics is something sometimes by intervening with the economy, sometimes at the bequests of the corporation leaders themselves, strangely. If I go back in history to Britain in the 19th century, 18th century, there were factory owners who weren’t very happy to be employing young children. They saw it was pretty terrible. But there was such a commercial pressure to use these young nimble fingers. They basically said to the politicians, please regulate us because we need a level playing field here. And by the way, please have a factory inspectorate to make sure that we’ve all not just said we’re going to stop employing children, but we actually do stop employing children. So sometimes the leaders of the corporations that are in this race realize it would be better if there was a level playing field. And to an extent, that’s what’s happening with occasionally the leaders of people like Microsoft, OpenAI, and Google, saying to politicians, say, you ought to be regulating us.

Some people are actually quite cynical about that, and they’re saying that the reason that those large companies are doing that is because they don’t want to see some post-AI-introduction litigation occurring, and that they are then sued in some sort of class actions by people who say that they’ve been exploited in some way.

I think there’s a complex mixture of motivations, no doubt. These people are— people who are running these companies, their advisers— are complex creatures. They are wonderfully intelligent sometimes, and wonderfully dense at other times, you know, without naming names. And so let’s put pressure on them. This is what civil society is meant to do, not just to lecture from a hectoring point of view— you bad capitalists— that’s not going to work, but to to understand what the technical possibilities are, the upsides and the downsides, and to engage in an honest, thorough, deep conversation about how these things can be used for the common good for everybody. And if we get it right, it will be wonderful, frankly. But I think it’s like electricity. Electricity wasn’t just useful for one tool. Initially, people thought it’ll do one or two things, and it turns out you can have enormous number of different electrical goods, even steam. Steam power was invented to do one thing: pump water out of mines. And then people realized, hey, you could use it in lots of other industries as well. Hey, you could use it to power trains. So I think AI is the same, in that it is a general-purpose technology. Here’s this GPT again, but it’s a different GPT from the ChatGPT. GPT is a GPT, as it turns out, and it will be able to invade more and more areas of life that people hadn’t expected it will come into.

The futurist, author, and advisor to the AI consortium Singularity.net, David Wood. According to Wood, the issue is that we are suffering from the same syndrome as that faced by the handloom weavers, a cottage industry wiped out by the northern textile mills of the Industrial Revolution.. But it’s not jobs that will go, it’s roles and industry sectors that are now redundant. And because of that, new ones will develop for people to become engaged in. As Wood pointed out, the weaving machines put some weavers out of work, but also meant more woven materials, so people ended up having woven carpets and woven curtains and multiple suits of clothes. They were richer. Though it’s not a process that is quite following the pattern of the Industrial Revolution, according to Dr. Claire Walsh, an advisor to the UK government on AI regulation and the director of education for the not-for-profit think tank the Institute of Analytics, which promotes the value of analytical technology and business transformation.

So at the moment, I think we’re in a very experimental phase. We can’t say with any confidence. We know that, for example, 10 years ago there were predictions that AI would by this point have replaced over 40% of all roles, not jobs, but roles. And clearly that hasn’t happened because we are not a deterministic society where the technology is invented and we immediately adopt it. There are human resistance to that. And so, you know, it’s very difficult to say where we are going to go. I think when we think back to job losses, we go to the past immediately. So we remember what it was like when industrialization deindustrialized. We had offshoring and whole factories closed down, whole towns, whole cities were destroyed. And in our minds, that’s what’s going to happen now. And I don’t think we’re going to see that repetition. And so I think this doomsday picture is probably not going to happen. But we do need to start adapting because machines will take on parts of our job, but perhaps not the whole thing.

That’s the interesting point, isn’t it? Because what appears to be happening is that parts of the 20th century that were obsolete and out of date are being replaced by AI. And it is replacing process, and that seems to be the area, and that’s the most visible part of this process, isn’t it?

So I think where people are embracing it and enjoying it, or where it’s able to do things like automate meeting notes and perhaps do some of the scheduling, helping us find and source information is, you know, these little tiny wins are definitely great. We had these predictions that, for example, that radiologists would be unemployed by this point because machine could detect cancer as well as a radiologist, but at 3 o’clock in the morning, at any time, and it could scale, and it had all these other advantages. But what we found was that at the moment, what we’re working with is very narrow AI, very specialized technologies. They can only perform a very specialized task. Even something super, super intelligent looking like ChatGPT looks like a person talking, but not a lot else. And so until we get what we call interactive AI, where these machines start talking to each other. So for example, in the case of radiology, the machine that can detect cancer in one part of the body can talk to another algorithm that can detect cancer in another part of the body. And so on. And so they start talking to each other and we start to get a more holistic diagnosis. We still need that radiologist to say, I think this is cancer of, and then wherever, and I think I need an image to help me. And then of course, you know, that radiologist then needs to link in with whole teams of care, support, nursing, treatment. And again, you know, this narrow AI just does not have that sophistication at the moment. To take over this complex relationship.

I mean, that again is the problem, isn’t it? And the media is actually at fault for a lot of this because the media keeps pushing this myth of the sort of Terminator-style robot that is going to be taking over our lives. And what isn’t really realized is that AI is probabilistic and it is not deterministic. And that is the issue, isn’t it?

And I think we all have a really poor understanding of probability, especially machine probability. We think if it’s 80% correct, oh, you know, probably correct, when in actual fact it could be wildly inaccurate at 80% accuracy, or even 99.99% accuracy could still be useless. And so we do need somebody who understands the probability behind these machines to be driving these decisions and deciding how to act with them.

That’s a very good point, isn’t it? Because one, this interview is being transcribed at the moment, and every now and then the transcription turns to gibberish, and we have to listen back to that particular bit to actually find out what it is that, in this case, you are saying.

And imagine if that gibberish happens to be in some other kind of machine where, you know, it’s deciding what mortgage rate you pay.

Really problematic.

Exactly.

So at the moment then, the role of the human in all of this is to actually go through the conclusions that the machine has taken and say, are those acceptable?

Absolutely. And then what action do we take next? Or perhaps why do we think this pattern has happened? Machines can tell us what’s happening, they can’t tell us very well why or identify cause and effect, and That’s problematic because most business decisions rely on identifying cause and effect so we can take decisions and impact change in the future.

Okay, so if we’re saying that machine is picking up apples and it’s putting apples from there into a crate, we know that that is doing what it’s meant to be doing and that that’s acceptable.. But if it’s doing something else, then we have to be able to double-check that.

Yeah, we need people who can work in human-machine teams where some of the decisions will be taken by the machine and probably will be taken better, more accurately, or quicker. Or maybe the machine can do the really boring decisions that actually none of us want to do. Maybe the machine will do some jobs we enjoy and value and we will miss them. But what we’re looking at is human-machine teaming, that it’s not all of the decisions, it’s some of them. It’s augmenting humans. And we were working with an organization earlier this year, just this year, which surveyed 16,000 employers and asked them what skills they were looking for. And all 16,000 said they were looking for digital and data skills, which I’ve never seen results like that. And when I speak to employers, they assume that these skills are coming, that someone out there is magically training people up with these skills, and in my experience that’s not the case. Universities haven’t quite shifted yet unless they are very specifically focused on analytics training. We’re not seeing insurance graduates coming out with strong analytics skills when companies perhaps are expecting their employees to upskill themselves, which is the way it works in our field. There is this strong expectation that we will self-learn and update as we go, but I’m not sure that that translates very well to a wider population outside the field of data analytics. I think she has a really good point because our best opportunity to instill new skills in the population is while they’re in school, and the education system has been terrifically slow to adapt. Those critical thinking skills she’s talking about are not firmly embedded in a curriculum that is basically aimed at performing well in international comparison tables. And that’s a problem. One of the things I wonder is how are we going to manage that transition from school to respective professional? If the machine is doing all of the entry-level jobs where we would normally allow young people to cut their teeth, to slowly make their way into the profession and up the profession, I do wonder how that transition is going to be managed in the future. And I have no immediate solutions for that.

Dr. Claire Walsh, the Director of Education at the Institute of Analytics and a government advisor on AI regulation, who is calling for a profound rethink by politicians and educators about how we engage with AI. And she’s not alone. Dominic Holmes is the Principal Consultant with the workplace education group Cornerstone. He works for its thought leadership and advisory services which provides ongoing education for workers in 180 countries worldwide.

I think if we look at all the industrial revolutions that have been in history so far, many of them have predicted the wipeout of jobs, and it depends what you mean by the wipeout of jobs? Are you talking about a change in the overall level of employment, or are you talking about some individual roles disappearing to be replaced by new roles? So I think it’s important to sort of explore that. I think that there’s automation And AI is going to drive further and more sophisticated heavy lifting of work in society. And I think, for me, that’s, that’s a glass that’s half full, but also comes with a challenge that we need to step up to. So, first of all, I think the world of work is going to become much more challenging going forwards. I would describe it as risky and an opportunity in equal measures. I think the key thing is to recognize that that change is coming. That change is probably coming quicker than industrial revolutions that we’ve seen in the past. And the really important thing is to prepare the workforce for that future so they can meet the future ready.

But that’s going to be the very difficult thing. You talk about the Industrial Revolution. The Industrial Revolution caused incredible social dislocation. You had large amounts of people moving from the country into towns and cities. You had a lot of jobs and roles that used to exist completely and utterly replaced. And if Marx and Engels were to be believed, quite a lot of misery. This, if it’s going to happen quicker, means that we’re going to have a lot of jobs and a lot of roles completely and utterly wiped out that people were quite happy doing.

Actually, they wanted an income from those. Yes, and I think this is where Certainly, I’m starting to see some green roots in business and ultimately academia in terms of what are the steps we need to start to take now so that almost we don’t repeat the distress that’s been caused by industrial revolutions in the past. And like I say, we greet the future ready. But ultimately, what I would say is that’s going to take system thinking across ultimately a partnership between government, public sector, academia, and business. And ultimately, only through acting in a coordinated way Can we support society as a whole to make that transition?

Well, we interviewed Professor Rose Luckin. She’s the UK’s leading authority on AI and education. Uh, Professor Rose Luckin was saying that we need to radically change schools. It’s not academia. We have to start right at the bottom and that we have to start to train children in analytical thinking, that we have to actually completely and utterly change the way that things have been done. You will not need to remember huge amounts of information because those huge amounts of information will be in the databases the AI systems are using.

You’re going to have to teach people to think. I agree, and I think it does need to start in schools, and critically That role needs to continue seamlessly in the workplace. So fortunately, we are all living longer. There are 4 generations in the workplace. Interestingly, people are beginning to question, do we want to retire, or certainly, do we want to retire at traditional ages? And I think this is where it requires this system of thinking where ultimately we make it so that people want to continue to work and critically want to continue to learn to keep up with the curve. But I think for me, it almost starts re-engineering our belief system of what the future looks like. We have to embrace change, whereas I think often we push back against change. And I think that’s maybe where, if you like, building the skills and capabilities of the future needs to start in school with maybe even— one of the things I’ve talked about is if you look at education, we’re taught to compete. Everything is about competing. And in a period of such change, we don’t need competitors. We need collaborators, and that’s collaborators with other humans, but collaboration with new technology. And that’s, for me, the fundamental reshift. For me, it needs to start with a purpose. They need to have a purpose for their business that is welcomed by society, and then ultimately build out, if you like, from that social purpose, such as learning, to ultimately drill down into putting more control in terms of the directors and the subject matter experts and give them the time to research and then scale their findings back to the workforce as a whole. One of the things I talk about is building the agile workforce. It’s the future of talent, but it starts by being a whole-of-workforce thing and a way-of-life thing. It’s no longer about working with the concept of top talent. It’s about upskilling and reskilling the workforce as a whole to really tap into that collective energy.

Dominic Holmes of the workplace training company Cornerstone. But is he right? Right. As Claire Walsh says, people are adopting the technology at work to help them do their jobs, but that would appear to mean that some specialist areas have been attacked. A point made by one of the world’s leading ghostwriters, Joshua Lysak, who says that because of AI, many of his spectral colleagues are being consigned to the graveyard of history. AI, he says, is coming for some of those in creative industries.

What worries me is that not enough professional writers are worried. I once posted a few weeks ago that in my experience already since ChatGPT dropped in December of 2022, that roughly 95%. This is a guesstimate. Roughly 95% of professional writers’ output is inferior to chatbots. Now, that’s not worrying for the 5%, but it ought to be. Why? There is a saying that no professional services provider can be good, fast, and cheap. The saying is, good, fast, or cheap, pick two. Well, with a $19.99 a month chatbot, you do get all three of those, particularly if you’re using the latest large language models. Already I have noticed aspiring authors, thought leaders, CEOs, so on and so forth, who pre-2023 would have and even were asking recommendations for ghostwriters, for copywriters. Well, we have a full-time copywriter. Now they’re using ChatGPT, they’re using Claude, They’re using PickYourChatbot. Why? Because it’s good, fast, and cheap. Now, the key is it’s good enough, especially for branding content, marketing communications, web pages, blog posts, etc. It’s good enough, whereas a professional writer, human, is 2 to, in some cases, 100x or more, more expensive. And so those professional writers, particularly ghostwriters, who want to have something like a career path still before them, ought to realize how good these things really are relative to the best or good enough. If the best is thousands in terms of cost, But good enough is 20 or 1999. That’s a great way to look at it for most individuals.

The leading ghostwriter Joshua Lysak, who came out of the shadows to warn that AI is becoming attractive because it can, like the textile mills, produce machine-made materials cheaply and in bulk. Beware. The handloom wordsmiths. To get a premium for your work, your designs must be very compelling and unique. A situation some are saying that spells doom to us. AI-driven machines will wipe out the most basic of jobs like bricklaying and plastering. Shop assistants will go to be replaced by automated tills, just as drones are taking over the battlefield people will disappear from the world of work. Not quite so, said Henry Chen, the founder of the data labeling company Sapient, which employs people for the essential task of labeling data for consumption by AI programs and is gamifying the process to make it more interesting.

The industry relies on us to power its human-in-the-loop, sometimes called RHF. For AI training. So when there’s data that need to be structured in order to train the AI model properly, there’s a concept that’s popularized by OpenAI as a human— RLHF, reinforcement learning from human feedback, which involves a very large volume of human to provide the data annotation work in order to train the AI model properly. So Our current business is essentially providing a large network of humans, the equipment with our custom-built platform, data training platform, and train the data. Now, the big problem of status quo of this industry is it’s a very fast-growing segment even within the AI industry, is the job is really boring. I think not many people well, when they want to just sit at a computer and just do repetitive data labeling work, it could be like drawing bounding boxes around cars and label it. It’s a red car, it’s Mercedes, it’s whatever. So the job is really boring. There’s very high turnover rates. A lot of the work is done in very concentrated, it’s more like a factory job to a lot of people around the world than actually, they don’t feel like they’re advanced doing advanced work to advance AI. So what we’re working on right now, so the next iteration of our platform is going to be, would be more decentralized and gamified approach to data labeling. We have quite a bit of projects on the go right now, and we’re slowly rolling out the alpha version of our gamified platform that was the goal of reduce churn rate, increase engagement rate, increase, uh, so All the bells and whistles will install into this platform, gamified platform, to improve accuracy, to improve efficiency, to improve retention rate of data labelers. We know for a fact that the biggest bottleneck for the data labeling industry right now is the high turnover rates. People don’t like the job, people don’t find it engaging enough, or finding that their work is meaningful. And you start seeing a little bit of articles popping up here and there, even from Washington Post, about the treatment of data labelers in a very confined space, more of like— we kind of call it labeling farms, so in almost sweatshop-like conditions. And so there’s a lot of problems deriving from the current methodologies to Gathering large network of humans to label data, it also leads to all kinds of biases, which leads to hallucination by AI models. So there’s obviously the, I think, the labor issues, but there’s also some quality issues with the data that is trained by this current status quo. So for us, it’s important for— because every year there’s more and more data that comes out. There’s a huge increase in budget, huge increase in demand by leading LLM models, especially hungry for more and more data. So there is going to be huge increase in volume of data that needs to be annotated. There’s going to be huge increase in complexity of the data that’s going to be annotated. So we see the future to be a more decentralized model where instead of work— showing up to work, working a office-like space. We see a world where instead of playing Candy Crush on your phone while you’re waiting for the bus, you’d be labeling data for a few minutes while the subway or the bus comes, and a lot more people. So it’s a new wave of kind of like a gig economy.

Henry Chen, the founder of the data labeling company Sapient, on a process that echoes the images of Charles Dickens’s devastating critique of the Industrial Revolution, Hard Times. Where workers known as hands minister to the machines. Though, as with everyone interviewed for this programme, he pointed out that this is an evolving world and one where it is very difficult to say with any certainty what is happening. Take passionate beekeeper Simon Hayton-Williams, for example. As well as being concerned about worker bees in his day job as CEO of the Adaptivist Group, a leading UK-based digital transformation company, Hayton Williams has also been monitoring what workers have been doing, and he’s found that like bees, people are capable of evolving— findings they released in a report called Digital Etiquette.

Well, one of the things that our research focused on is the different approaches across ages and generations of people with their approach to technology. And we’ve found that, you know, certainly across all generations, there’s a wide adoption of tools like ChatGPT and GenAI. And there was kind of before this, there was a sort of presumption that perhaps the older folks were not grappling with these technologies. And our research shows that in fact they are. And I think 30% of people over 50 are using GenAI in their work already, which is not as many as the younger folks, but it’s very much coming in there. I think one of the things that is really important is to recognize that people, this kind of technology is embedded in so many things that we use every day that people are using it without necessarily realizing it. And I think like most other technologies, given enough time, this is just going to vanish into the background and be part of daily life. And right now it’s interesting because it’s new and shiny and people are taking the, I guess, the dividend of this new technology. I’m strongly of the belief that it will just become part of everyday life. And I kind of liken the technology today to the early kind of pocket calculators. And so I’m of the era where my dad made me learn to use a slide rule because he said, you can’t do anything without a slide rule. And of course, you know, now we all have a pocket calculator everywhere. Has it, has it reduced some jobs? Possibly. Has it created many more? Of course it has. And I think it is, it’s really about technology augmenting people rather than replacing people.

But that’s the interesting point, isn’t it? One of the things that it will do, and there does seem to be a general consensus on this, is that in the short term there will be jobs that will be removed, and those are jobs that are involved in processes. There’s also certain jobs in accountancy for the first time that will be going. People are saying that paralegals will be going. Anything that involves a process can be automated.

I mean, I think it’s true that anything, you know, if you think about software full stop, software is just business process automated and talked to a computer. Gen AI and the tools that we’re using today extends that. And yes, there is some truth to that and there will be some pressure on those jobs. But whilst those jobs may be lost in the short term, it’s perhaps a reordering of those jobs and it creates doors for more opportunity to do more things and more sophisticated things. Again, AI is just a tool. It’s a tool that allows people to be more effective and more efficient. Does that mean we’ll need less of those tools, less of those people? Potentially, but there’s other things for those people to do as a consequence of that.

The apurist and digital transformation expert Simon Hayton-Williams, head of the Adaptivist Group. It’s a process of digital evolution that has also been noticed by podcaster and recruitment expert Matt Alder, author of Digital Talent. Though according to Aalder, it’s a process that’s not quite following the script. Indeed, as Lysäk told us earlier, often it is writing the script.

I think it’s, yeah, none of this is following the script. I think that’s the thing. You know, there’s been a script about we’re gonna automate all of these physical tasks. We’ll have robots doing things for us. And that, you know, very well may still be the case. But as you say, it’s kind of come the other way around. It’s taking a lot of the thinking jobs and has the potential to do that even more. So, but actually it’s inventing new ways of doing things. And I think that’s the critical part of this. And it’s thinking through what the implications of that are. And I know that in my own industry, there are debates about this going on in terms of, well, will there be any recruiters in the future? And they’re difficult conversations to have because people automatically get very defensive and protective about what they do. And they think of all the reasons why. That couldn’t happen rather than the reasons why it could and how we might move to something better and what those implications are. So I think it’s a kind of a very difficult mindset for people to work through. You know, I’m old enough to remember when the internet turned up and we had some similar conversations and you kind of look back and yes, it did fundamentally change a lot of things, but a lot of things also stayed the same and just became quicker and more digitized. And I think that there was just a, just a potential here that lots of things change completely and we are not really thinking about it enough.

I mean, that’s again, an interesting point, isn’t it? Because all of the justifications that people are coming up with, all of the anecdotes that they’re coming up with, they’re saying, right, the industrial revolution it caused dislocation, but then the jobs flowed afterwards. The internet revolution, that caused dislocation, but then the jobs flowed. It’s not necessarily the case that you can always use experience as the model for understanding, is it?

That’s a really good point. I think you can always learn a lot of things from history, but things never, you know, are never quite the— never quite the same the next way around. And there’s also— there’s always the opportunity of kind of massive unexpected disruption happening, which is kind of where I think we might well be right now. And it’s not to say that this won’t create new jobs and new opportunities and interesting things. I’m absolutely sure it will. But what speed does it do that at? And how far are these new opportunities away from what people do at the moment? So even if it created the same amount of opportunities, one of the things about the The dot-com boom was the intense, just intense shortages of people with tech skills and all the things that were needed because things had changed and it took a really long time to get that balance back. And in a lot of ways it still isn’t. So it’s kind of like, what shock does that have in terms of a skills gap? Yeah, I mean, specialist skills are still 100% needed. And there is that kind of arms race around the AI specialist skills at the moment, and that’s obviously going to continue for quite some time. But I think what the most forward-thinking employers are doing is they’re looking at what they need in terms of skills in their business. So it’s like, we need people who can do this, we need people who can adapt to do this, and that’s kind of informing the way they think about talent.

Recruitment expert, podcaster, and author of Digital Talent, Matt Alder. So what’s the answer? Will it take jobs? Will it give us a life of ease? As a futurist David Wood told us at the beginning of the programme, AI will be like electricity and we will find it used in surprising places. One of the points made to us in previous programmes by the University College of London Professor Rose Luckin, a leading expert on AI and education, has been that the advent of the technology means that we have to develop critical thinking skills, a point underlined by many interviewees. As billionaire Risto Silasmaa said, learning how to think in the intelligence age is key, and Aude pointed out being able to learn quickly is key. And it’s something that Dr. Robert Harrison, Director of Education and Integrated technology at the ACS International School says is something that we must learn.

I really don’t think that teachers are likely to be made redundant by AI. I do think that AI will be changing the nature of teaching and assessment to some extent, and I think it offers real opportunities for schools to rethink curriculum as well as to prepare students for a very new sort of world with this kind of alien intelligence that AI offers. So that we need to prepare students both to be better citizens, to cope with the, the world that they’ll be living in, and also to prepare them for ways to work with machines in new and creative ways, and also to build better AIs in the future.