Oxford Sparks

[Music]

Emily Elias: Most workplaces are dealing with AI and how exactly they should use it. But did you know your knowledge of using AI could actually translate into a bigger pay packet?

On this episode of the Oxford Sparks Big Questions podcast, we’re asking how can I use AI to raise my salary?

[Music]

Hello, I’m Emily Elias and this is the show where we seek out the brightest minds at the University of Oxford, and we ask them the big questions. And for this one, we have found a researcher who wants to help you make more money.

Dr Fabian Stephany: So my name is Dr Fabian Stephany, and I’m a departmental research lecturer at the University of Oxford. And at our department, the Oxford Internet Institute, I run a research group that is called SkillScale. And here we investigate how artificial intelligence and other digital technologies are changing labour markets and specifically the demand for skills.

Emily: And so you’ve been working on a study that looks at using AI to sort of help raise salary instead of being a fear monger, AI is going to take my job.

Fabian: That’s true. I have to say, though, maybe that’s a big disclaimer that I should put up front is I have this secret agenda, this hidden agenda for this, why I started this, the SkillScale project, that came from a personal annoyance. And I’ve been increasingly, I am still increasingly annoyed or am annoyed by, yeah, you call them fear mongers, the dystopians, people that see an imminent threat, big threat, in artificial intelligence when it comes to the labour market. So they essentially think we’re all going to be unemployed because of this technology, AI is going to steal our jobs.

And while there is something about this in the sense that AI, like any other major technology, will certainly bring tremendous changes to the labour market, it is already bringing tremendous changes to the labour market. I think the fears of massive unemployment, that is 50% you or me losing our jobs within the next decade, are completely overblown. AI brings potentially better work qualities. Also, as you mentioned, higher wages to people. And the question is like, how do we make people, what are the skills that you need? How do we manage as many people as possible to access these skills? And yes, and so we’ve done a couple of studies, including one where we investigated how profitable is it to have skills centred around AI. And we see a pretty clear picture here.

Emily: Do you mind just sort of like explaining, like, what is an AI skill?

Fabian: AI skills is a very broad term. It’s an umbrella term. But very quickly said, it contains two types of skills essentially – creator skills and user skills. So the skills that you need to develop AI algorithms like deep learning, like the chatbots that we work with, and the user skills. So people that use this, that do prompt engineering, they effectively learn how to communicate with AI agents. They know how to truncate data. They have high levels of digital literacy, of understanding how this technology works.

And the important thing is both are valuable. It’s not only that the creator skills are the most valuable. It’s where the money and the profits are. We see in our study it’s equally important and valuable to know also how to work with this technology, to integrate the AI skills, to integrate the technology into your work.

Emily: Okay, and so what kind of jobs are we talking about? Let’s get on the same page here. Are we talking about like I’m a hardcore coder that’s using AI to like speed up my website making abilities? Or are we talking about like something else?

Fabian: We try to picture as much as we can. We try to picture the entire labour market. So we conducted several studies – so the answer is a bit more nuanced in the sense that it depends on the study that we’re talking about.

One of the studies where we investigated the wage premium, that means how much more money can you earn compared to people that have the same education and experience and industry background than you, how much more can you earn with AI skills. And that’s a study that’s been leveraged on job vacancies, so job postings, roughly speaking ten million job postings from the UK. And here we have a pretty broad representation of the labour market. Of course, there’s some jobs that are not as represented as they should be. So much of manual work, for example, is not often shown in online job vacancies. But we are definitely not only talking about the hardcore machine learning or AI developers, right? We’re talking about people that integrate AI into their workflow.

Those could be people in marketing. Those could be office assistants. Those could be people like you and me, right, that do research, that do interviews, work with digital technologies, record videos, pretty much of the so-called knowledge workers, we try to cover them all.

Emily: And so what did you find out?

Fabian: So in that specific study, we actually wanted to know, okay, what are the synergies that people have when they integrate AI into their workflow. And if we talk about quality of work, of course it’s not the only feature, but the feature that comes to our minds probably most immediately is like, okay, how much is this worth, right? How much more money can I make with AI?

What’s fascinating about this is that even when we control, in a statistical sense, control for education and for experience and the type of occupation, type of industry, there is on average, a premium of 23% that can be attributed to artificial intelligence.

So to put this in simple words, if you kind of find your twin on the labour market, so in a sense a person that has the same level of education, same level of experience, works maybe in the same office, in the same industry, same occupation, but this person does not have AI skills and you have AI skills, on average it’s likely that the wage that you’ve been offered is 23% higher than the offered wage for your twin.

Emily: 23% is really, would add up in my life if I was making 23% more.

Fabian: It’s more than just a holiday. It’s more than just a fancy vacation, you know? Yeah.

Emily: So how do I get these skills then, if I want? You know, I’m a working woman. I’ve got a life. I like family. I like to see people. How do I add in learning how to use AI to my day to day to get that extra 20%?

Fabian: That’s the key question, right? And that’s also actually the starting point for us for our study in the sense, because we realised that the conventional way that we acquire skills in schools or in universities is, I hesitate to say, is no longer functioning, it does function of course, but is currently, right now, slightly too slow for the development of the market.

AI is penetrating industries at a pace that is hard for conventional education providers to keep up. But where do these people acquire skills? We don’t know this from this specific study, but we know this from other investigations. It’s very, very often a different place, and this place is the firm. So the firm, the company, your workplace, in a broader sense, is with regard to many aspects a more efficient and more sustainable place to acquire these specific AI skills compared to schools and universities. And there are many reasons for this.

The main reason for this is because your workplace is much closer to what is being demanded quite naturally, right? So if in your workplace the demand, for example from customers or from the industries or from what technology has been demanding from us, becomes much faster, more apparent, clearer to us than, for example, in universities, in schools, because there are so many filters to this, right, because then it kind of trickles down. So first of all, your workplace is the place where to acquire these skills.

Emily: But what if my workplace isn’t up to date and I am trying to, you know, find a new job outside of my workplace? Like where are some places that I can go to skill up in this really quick moving AI world?

Fabian: So what we’ve seen is that there are many, in light of the scarcity of conventional education avenues, there are many people who have acquired these skills autodidactically by themselves online.

So there’s a great volume of courses that are available on various online platforms. I think actually the problem is not so much the availability of these classes that are out there online, but the problem is rather to say like, okay, what is my AI skill, right? Because there are so many things that you could learn that are centred around artificial intelligence. It’s should I do prompt engineering? Should I learn, in that sense, how to communicate effectively with AI agents? Should I be learning data analysis? Should I know how to use AI to truncate data and to understand, you know, what kind of value certain market data has? And that is something that, there’s no one size fits all. That’s the problem. But there’s also the beauty of it, because it really depends on your existing skillset.

Emily: How can you know what you’re getting is what you’re getting and that it’s reputable and that it’s not like whoopsie gypsy, I’ve just given somebody a bunch of money and I know nothing about AI?

Fabian: Absolutely. So you’re putting the finger where it hurts. Um, whenever this kind of uncharted territory in terms of technology and learning is being discovered, we see these, you know, snake oil merchants, as people call them, the people promising tapping into these new resources, into this promised land. And in the end, the quality of what’s being taught is not really high. And what is also important, the signal is not that strong.

You can still go with established credentials. So for example, we did a study with recruiters just recently, which is going to be out as a working paper soon and hopefully as well as a published article, where we looked at the value of AI skills in the eyes of recruiters, of professionals, HR professionals, so the people that make hiring decisions for companies. And Oxford, for example, places at Oxford, still has a strong market value, has a strong signal, if you have it as a brand name with you. Very often people think about this in terms of a master’s or bachelor’s degree, but you could think of this as in a micro credential way as well. What we see is that there was an additional effect of this credential, particularly for people that had disadvantages in this competition, so for people that were slightly older in our experiment, for people that had slightly lower levels of formal education. So long story short, look for credible providers of education.

The fact that universities currently struggle to have large scale structured courses, maybe for teaching artificial intelligence, does not mean that they’ve lost their value, right. There are micro credits that universities offer, often in collaboration with online courses like Coursera, where you not only have good content, that’s important of course, that’s the most important thing, but what also matters is that you have a strong signal, that you have a strong brand that’s attached to this high quality class. Because that’s, let’s face it, that’s something that’s also of value for recruiters, for firms, just because time is limited, right? They don’t have the time to check every single detail of every class, but they often see like, oh okay, this looks like high quality. Oh wait, and it even has a great brand name with established university to it. That must be a high-quality course.

Emily: Does the University of Oxford offer this sort of micro coursing thing that I could take and like get some sort of AI skills to add to my CV?

Fabian: Absolutely. So now I’m in a difficult situation. We can praise the University of Oxford endlessly, of course, but we get into a bit of a difficult situation when I start advertising for certain departments, particularly those departments that are not my department. But what I can say is that across various departments, including our department, um, there are courses already available. Some of them are standalone courses, so some of them are courses that have been exclusively and only been developed by the University of Oxford and provided via our platforms, via the University of Oxford. Some of them are already provided or will be provided in collaboration with platforms that have been doing this already for many years, online learning platforms.

Emily: So the future, I should not be afraid of AI. I should instead embrace it to help my pay packet look a bit healthier.

Fabian: AI, if you do it in a smart way, is not an enemy at all. It’s probably your biggest ally for improving your work quality, but also maybe finding a new and sustainable career.

[Music]

Emily: This podcast was brought to you by Oxford Sparks from the University of Oxford, with music by John Lyons, and a special thanks to Dr Fabian Stephany. And if you want to find some resources on how to learn AI, just go and check our show notes. There are some links in there that are wonderfully curated for your knowledge.

Tell us what you think about this podcast. We are on the internet at Oxford Sparks or go to our website at oxfordsparks.ox.ac.uk.

I’m Emily Elias. Bye for now.

[Music]