There is a human condition, there is a human spirit and I think we are a long, long way from automating that, and actually why would we?”
Paul Feldman is talking about artificial intelligence (AI). As chief executive of Jisc, it’s not the line you might expect him to take. Yet it’s refreshing to find such a pivotal advocate for technology in higher education – or, indeed, elsewhere – who takes such a pragmatic line about perceived future digital capabilities, and who so shamelessly celebrates the abilities of people over the potential of computing.
It’s not as though Feldman hasn’t seen the leaps technology has made. When he graduated from St Andrews with a computing degree in 1981 “there was no concept of personal computing”.
He went into commerce, building a system for the Sellafield nuclear site for dosimetry, the measuring of internal radiation. “I did a few jobs all of which made me feel very unsatisfied with the result. It didn’t seem to match what the real needs were,” he recalls.
He read for a PhD for a couple of years before moving back into consultancy, moving through companies and ranks before settling for 25 years in retail financial services, mostly with Nationwide Building Society. Despite starting in an IT role, he eventually became its products director. It was, Feldman admits, “quite a commercial role”. During the ascent he had moved away from computing almost entirely. “It’s something I really love and have enjoyed and I love using technology to make a difference,” he says. He moved back to the IT side, becoming chief information officer globally for Barclaycard. Then 2008 came and, with it, the credit crunch.
“I didn’t want to be in a job where I was the person everybody hated, so I went to do some work for Cancer Research UK. It’s great to give back after being deep in the financial mess. It was doing a bit of soul cleansing, I suppose.”
There followed other work – among them a digital organisation that was bought by Thomson Reuters
(“I did not want to be in a big organisation”) and advisory work for Gartner – before the call came from Jisc.
“I thought, once I’d worked out what Jisc was, actually it sounded perfect,” Feldman says.
“It gives an awful lot back doing this and it’s a really exciting organisation going through a lot of change, so it was just a perfect thing.
“And somehow they made a mistake and appointed me.”
What is Jisc?
For anyone who has tried to define Jisc, there will be a suspicion that he got the job simply for working out what it was. Formed from 11 different organisations (“now 12, and we’re about to have the 13th”), it is, in Feldman’s words, “the digital body for UK research and higher and further education”.
To many, Jisc is Janet. “That’s at the heart of everything we do. It’s the biggest thing we do and the most expensive. It’s the biggest in the world, it’s the best in the world,” says Feldman, going on to extol some of the services built on top of the ubiquitous IT network for higher education. “We have things like eduroam that is probably one of the most valuable things to academics – their ability to travel the world and get wifi anywhere in the world. It’s an aside, but some universities left the New Zealand national research and education network, but soon returned when their academics complained about not having eduroam. It’s these little things that actually make some of the biggest difference. It’s that range of key IT services that makes researchers’ and academics’ lives so much better than they would be.”
Then there is library support, including a project with the British Museum to digitise 13th- to 19th-century English texts. “That’s a resource that will be there forever. At the launch I was told about a Shakespeare researcher who said for the first time they could do textual analysis of the complete Shakespeare works. Doing little things like that, that have such a big impact, is fantastic.”
And Jisc produces digital courses, and guidance, and IT procurement. As Feldman reels off the many reasons everyone finds his organisation so hard to pin down, phrases recur: “It’s about digital and research and higher and further education”; “it’s about that technology piece”; “it’s digital and technology stuff”. It’s hard, he admits, to explain coherently.
The LEO data is really interesting but it’s a survey at the end of the day, and surveys come with a set of health warnings
Jisc is in the process of incorporating some of the Higher Education Statistics Agency (HESA).
The first phase will bring across the commercial arm, leaving the designated body which handles the collection of data untouched. But will it follow?
“I think everybody agrees it makes sense for that to be a part of Jisc – it completes the picture of digital for the sector – but then there’s a timing piece. We need to make sure that when that merger happens it doesn’t disturb the focus on Data Futures [a strategic data programme at HESA]. The desire is still to bring the two organisations together. But that decision is still to be taken.
“The expectation is that if it happens, it will be a standalone and ring-fenced subsidiary within Jisc. Partly that needs to happen for them to retain their designated body status. They’re also quite a different thing to the rest of Jisc. There are definitely similarities in terms of delivering key infrastructure for the sector but the way they do it and the regulatory nature, certainly in England, around it is quite different so it makes sense for it to be ring-fenced with a very clear governance structure that means it can retain independence from the rest of Jisc. That’s critical, at least in the first few years. Who knows where it goes after that?”
CV: Paul Feldman
Name: Dr Paul Feldman
Job title: chief executive, Jisc
Born: London, 1959
Education: University of Warwick – Warwick Business School (PhD,
industrial and business studies; University of St Andrews (BSc (Hons), computational science)
Awards: The Sunday Times 2016 list of Britain’s 500 most influential in engineering and technology
Career highlight: Apart from becoming CEO at Jisc, winning Computerworld Smithsonian Award for innovation in finance
What would you have done if not education? Education was a third or fourth career, and if I hadn’t moved to Jisc I would have sought jobs in executive advice across all sectors. Other sectors I’ve worked in have been finance, charity, digital tech, consultancy.
Best leadership advice: Always trust your gut/instinct, and always remember this advice.
We live in the age of data. Big data. Good data and bad data. Data quality. Is there too much data?
“I don’t think there is such a concept as too much data. I think our ability to manage that data historically has restricted the need to collect data. There’s no point in collecting data that you actually can’t use, but I think that technology now has got to the point where you can’t have too much data, in my view. Yes, the data needs to be collected well, it needs to have integrity and it needs to be well protected, but our ability to really delve into and mine that data and get the messages the data is giving us have never been greater and machine learning will allow us to do that in better and better ways.
That can only work through having rich data sources. So you can’t collect too much data.”
Not only does Feldman not think you can’t have too much data, he also argues that HE doesn’t have enough.
“I think the sector is behind other sectors,” he says. “The more commercial sectors, because they’ve been able to see the payback, have invested much, much more in data than higher education has historically. Universities up and down the country are doing digital transformation. They have appointed people to do this and I think the leading universities – and there are quite a number of them – are really taking digital transformation seriously. To do that they need to be treating their data properly. We’re seeing that trend happening. It’s happening in individual institutions. As a sector, that data is essential. OfS [the Office for Students] talks about wanting to be data-driven in their regulation so it will be demanding data and institutions need to understand their data to match the regulatory push that OfS will put on them. I think policy decisions have been taken historically based on less good data than we’re able to do today. But again it’s understanding the data and what the messages are.
“I think there are some quite dangerous assumptions that can get made from the data and there are some very bold statements that are made. So the LEO data is a really interesting source and it’s indicative but it is a survey at the end of the day and surveys come with a set of health warnings. There’s a real opportunity to take that data into places that aren’t credible and I think it’s about really understanding the quality of the data collection and how far you can actually rely on that to take decisions. I think it’s indicative and useful for taking policy decisions in DfE [the Department for Education], for OfS to help understand the quality of institutions to an extent, but it is to an extent and it’s important not to use it as a gold source of data because every survey has flaws, and understanding those flaws can help take the best decisions.”
Competition between institutions should be on the best ways to support students and lots of other things, but not how to teach the theory of relativity.
Last year, Jisc launched Education 4.0, the HE silo of Industry 4.0, the catchphrase for the fourth industrial revolution, driven by technological change. A conference was speculatively planned for this autumn. “We’ve been talking about an Education 4.0 conference, but I’m not sure we’ve pressed the button on it,” says Feldman. He mentions self-driving vehicles in logistics or robotics taking on paralegal work in law firms.
“Just about every sector is getting changed by these new technologies and our belief is that higher and further education and research is not going to be immune to that. There are lots of students getting a great education but it’s quite an impersonal education, whereas we believe the technologies can give every student a personal learning journey that’s tailored into their needs. It’s not an unknown concept that the way we teach at universities hasn’t fundamentally shifted almost since the 12th century.
“We are talking about something for 15 or 20 years’ time. Some of the technologies we need just aren’t there yet, but they’re coming, so it’s directional at the moment. You think back 20 years: 20 years ago it was the millennium. You just think how different the world is now to what it was then. But actually it’s taken us 20 years to put some of those things in place. We had the internet at that point, we had the world wide web. We are only now starting to see the high street transformed because of things that happened 20 years ago. So you need to start thinking now about the technologies that are going to transform the world in 20 years’ time.
“It takes time to actually train your lecturers and to understand what the lecturing job is. I think the lecturers coming out now, the postdocs that are just starting lecturing will have a very different experience in 20 years’ time and we need to help them understand what that looks like so they can be ready for teaching in 20 years.
“I believe the theory of relativity will be taught very differently in 20 years. It will be an experiential thing. Whatever virtual reality is then, it will be very different to putting on the sort of headsets that we all see. But where is that experience going to come from? Are we going to develop it a thousand times, in every university around the world or is the world going to get together and do it once? How do we get ourselves, as a community, to develop these critical learning resources so that it’s done once and the best it can be. Competition shouldn’t be on those things. Competition between institutions should be on the best ways to support students and lots of other things, but not how to teach the theory of relativity.”
Alexa on steroids
“Our vision is one where the students gain the knowledge themselves though self-directed learning supported by machines, so it’s Alexa on steroids, if you like. But there’s an intelligent mentor who works with the student individually, helps them get the knowledge they need through the set of resources that will be around. But it’s not a solitary existence. We believe there will be a lot of active and project-based learning so the students can be working collaboratively with other students on campuses and wherever else they need to be and really exciting, very immersive activities that are building a lot of soft skills as well, building the skills they will need for whatever their work is because we know that’s a critical need for the future.”
But, Feldman argues, returning to his “human condition”, that doesn’t spell the end for lecturers. “We do believe that there’s going to be lots of need for lecturers through this process, whether they’re lecturers or mentors or coaches or something different. It doesn’t really matter what they’re called. And we talk about lecturers giving their wisdom, not their knowledge.
“I was talking about this at a conference recently and there was a question from the audience asking does this mean we no longer need contrarian lecturers? But actually we need more contrarian academics who are going to really challenge the students’ thought process. It really expands their thinking and that’s what we, as people, do best.
“It’s spotting the real spirit and creativity in students and building on that, which is something machines won’t be able to do within our lifetime, is my belief.”
“It’s using people for what they are best at, which is the human condition, not the repetitive task.
“In lecturing, standing up and giving the same lecture every year is a repetitive task, so we can free lecturers up from all of the mundane stuff and help them spend much more time on the the much more exciting business of really growing their students and embedding the knowledge and challenging and using their wisdom and insight in really exciting ways.
“That has to happen gradually because that vision is a vision for the future. We can’t do that today, but institutions should be starting on that journey and starting to use the technology. Suddenly a lecturer will turn around in 20 years and think, ‘God, whatever happened to me standing up and doing a lecture?’”
But it’s also the nature of learning Feldman would like to see changed. “Fifty per cent of the country have been sheep-dipped through a course. We need to teach them to learn because they will need to reinvent themselves. Some of the skills they’re learning now, many of them will not be needed in the future. Their profession might not be needed in the same way. If you’ve got a 60-year working life as opposed to a 40-year working life you probably do want to have a number of different careers.
“If you’re 40 years old and want to change career and you have a family, you have a mortgage – a bloody expensive mortgage – you can’t afford to go away for three years. You need a different way to get that higher-level learning to become a different type of professional.”
Investing more in our colleges, not at the expense of our universities, is what we should be doing. We need both.
Rites of passage
But Feldman is not about to throw his weight behind the idea of a future where virtual learning takes place through a screen. Again, he subverts the technological ‘ideal’ with a human perspective. “There’s a rite of passage – an 18-year-old going to university – and there’s a growing up that happens there for those people that want that. That’s the piece that really gives you the foundations to be able to do that reinvention over the years. It also is that wrenching out from the family home for many, and that ability to explore who you really are. I still think there’s a need for that campus university for those students who want that. Having been through it myself, and having seen my children go through it, it’s really valuable.”
The publication of the Augar report has cast a beam of light on underfunding in further education, and it has not escaped Feldman’s eye. “I think there are still that 50% of kids that don’t want to go to university for whatever reason and they need that same learning, and that ability then to go on to university. I hope that whether it’s from the stimulus from the Augar review or just because it’s needed, that it’s not just university-educated kids who need that lifelong learning partnership, but those kids that go to colleges also need to think about how they can use universities in the future.
“University students, if they want to retrain to be a plumber, a college environment is the best place to do that, so I think much more symbiosis between higher and further education and helping our students feel comfortable in either environment over their lifetime is a critical thing. We do need to think about what that lifelong journey looks like and how it works, and the university’s place in that, and how you can make it practical.
“Kids can learn photography at college and start to practise it in an apprenticeship with a photographer.
They may then want to go and get much more capability in that space. So they might start at a college and end up at university and that seems to me a great path. But I think our colleges have been under-invested. I really support that view. You go into any college and you compare the resources between the college and a university, and investing more in our colleges, not at the expense of our universities, is what we should be doing. We need both.”
Again, there’s a marked humanity about Feldman.
It’s a far cry from the idea of computer boffins crowding around a screen, plotting how to do away with everyone’s jobs. Is artificial intelligence overhyped?
“Artificial intelligence is a misnomer,” Feldman says. “It’s not intelligence today. It is machine learning. It’s a better mousetrap. It’s way better software and way better ability to analyse data and get inferences from that data, but it’s not intelligence in that sense.
“There are some things that it’s going to be better than people at. If you look at medical diagnosis, there are lots of examples where machine learning is better at spotting things than people because it is looking fresh every time, in a repeatable way and is looking at a mass of data. It is going to fundamentally change some professions but that doesn’t mean we’re not going to have doctors. We still have people who need to be treated. And you still need that human interaction. You still need that ability to say, I know that’s what the machine is saying but actually I think it’s wrong and I think we need to question this.
“You can get artificial intelligence that can dream up recipes. It doesn’t make those recipes something that you want to eat. It’s a bit like a million monkeys given an infinite amount of time will come up with the works of Shakespeare. You put in a file to create thousands of recipes and you’ll get one or two that are great. One or two pieces of music that are great. You ignore all the poisoning that’s happened along the way.
“There is a human condition, there is a human spirit and I think we are a long, long way from automating that, and actually, why would we?”
Paul Feldman on…
… edtech vs the world
The problem edtech has compared to the fintechs and the medtechs is that the money’s not there. So the UK’s got about 160 mainstream, state-funded universities, a few hundred small ones, and that’s a small market so they need to look globally. The universities can’t afford to spend millions and millions of pounds on a trial of a piece of technology. A lot of the fintech in the investment houses doesn’t take much payback because they can charge an arm and a leg, which is why there’s so much fintech out there. The payback is really there in fintech; it’s much harder to see payback in edtech and it’s much harder to grow a business when you’ve got that sort of investment challenge globally. We need edtech.
From statistics I’ve seen, I don’t think higher education has been any more attacked than other sectors. I think there were some statistics saying globally around 65% of organisations have been hacked, or have admitted to being hacked, which is a number that’s not that dissimilar to higher education. I don’t think that higher education is any more exposed. I do believe that higher education is taking it as seriously as any organisation. It’s a never-ending task, and governing bodies need to take this seriously. Make sure you keep cybersecurity high on your risk register. We do believe every university is taking it seriously and we don’t believe you can take something like that too seriously. That’s our message: universities are doing well, we think they constantly need to do well, and those that aren’t need to catch up.
… being hacked
I myself have fallen victim to a phishing attack. I received an email that I thought was genuine from a member of my staff. It’s really, really easy and it’s almost impossible to put technology defence into that. I like to think I’m quite educated in this, I quite understand what I’m doing.
I checked the email address and it was from the member of staff, who’d had their email hacked at some point. I looked at it, I thought ‘this is pretty genuine’ and it was only after I put my credentials into the website I thought ‘that’s not right’ and immediately got my password changed, so we weren’t exposed, but I fell for it and no education was going to stop me. I was in the middle of a meeting, it looked urgent, it was from a member of staff, it was a bloody good one, but criminals are really good at this. Every organisation we’ve tested – and it’s only a proportion of universities – has had someone who’s fallen foul of that sort of attack.
… student mental health
Students leave digital footprints all the time in lots and lots of different ways. And the more of that data that we can collect in a sensitive way and protect, but also analyse, gives us clues about what the student’s doing. Today we’re focused on using that to help the student succeed, really understanding if they’re on track to achieve their potential.
Today we don’t think the data we’re collecting can particularly tell us whether the student is not understanding the subject or suffering from a wellbeing issue so we are working to really understand what’s needed to understand wellbeing. We believe you can collect different types of data to help give an indication that a student needs support, not because of academic, but because of wellbeing reasons. We do believe that there are solutions in that space to finding those students who need help. If we could have a piece of technology that we know how to use, that could identify some of those students and save some of those lives, we absolutely should do that.
You might also like: HESA and Jisc eye merger to create data-digi superpower