Future of Work Commission: Our Second Public Evidence Session


Our second public evidence session was themed around encouraging and harnessing innovation. We heard from Professor Rose Luckin from the UCL Knowledge Lab; Paul Wilmott, Global Head of Digital, McKinseys; and Gi Fernando, social entrepreneur and founder of Freeformers Ltd.

Their testimony focused on how we could use new technologies to support in-depth learning and revolutionise exams, to address skills gaps in the workforce and to work collaboratively in order to advance our goals.

Innovation in education

The use of technology in education is a fantastic example of how we can direct innovation to socially useful purposes. Professor Rose Luckin works in the field of Artificial Intelligence in Education (AIEd). She gave us the benefit of two decades spent developing and evaluating technology for learning. I took two main points away from her evidence – the first being that we need to update our ideas about the purpose of education, and the second being the amazing possibilities AI opens up in the classroom, especially when it comes to deep learning and assessment.

Rose was clear that we need to stop thinking of education as a process that has a finite end, and whose purpose is to solely to instil knowledge of facts. She made the point that in the future education needs to become something that we go back to again and again, and that – alongside teaching us facts – teaches us skills. She said that one key skill we should aim to foster in everyone is resilience – we don’t know what skills we’ll need, so we need to make sure everyone learns to cope with challenge and change. We need individuals who can accurately assess what they don’t know, and feel confident going about getting help to fill that gap. Essentially, we need a creative nation of life-long learners.

Automation is creating problems that our education system needs to adapt to. But it is also providing some of the solutions. Rose told us that education AI systems have the potential to revolutionise assessment, to make it much more accurate and efficient. She called the current assessment system the ‘bung in the bottle of innovation’, pointing out that it tests a small part of people’s knowledge in a narrow way over a short period.

An AI alternative could map and track how individuals are doing over the whole of their careers, starting from primary school - it could assess areas where they need development, and provide that help. And not just individuals. Such a system could analyse data at the level of class, cohort, institution, or country. The applications are huge - this could give us the ability to map and respond to skills gaps in the workforce with the people we are training on a moment to moment basis.

According to Rose, the biggest obstacle to this as not money (the exam system costs £845 million a year... which according to her would buy a pretty good AI system) but worries over data. Parents, teachers, learners – they all have concerns about the way their private data might be used. So the big conversation we need to have is about ethical collection and permissible access to prevent this becoming an unnecessary barrier to innovation.

A nation of innovation

Paul Wilmott talked to us about his new report - 'Harnessing Automation for a Future that Works'. The report looked at 800 occupations across 46 countries, looking at underlying skills required, and the current capabilities of tech to do those. It found that very few jobs are candidates for full automation – less that 5% - but that almost every occupation has partial automation potential. Paul reminded us that a couple of hundred years ago, the UK led the way in technological development and deployment.

We are a country with a proud history of innovation – for example, the first commercially viable computer was invented in Manchester. He then looked ahead to a future where demographics indicate the UK workforce is predicated to stay relatively static. That means any increase in GDP will have to come entirely from increased productivity – and we will need every erg of human labour working alongside robots. He addressed the UK’s perennial problem of productivity and pointed to how possible solutions could and should come from the world of tech. The UK needs policies to encourage more investment, innovation, and adaption by institutions and workers -as well as safety nets for the transition.

Diversity gives you a better answer, irrespective of the question

Our final speaker was Gi Fernando. Gi started off by talking about a 16 year old he employed in the early days of Facebook. ‘Shane’ had just been kicked out of school, and Gi hired him by accident. Yet he proved to be brilliant – leading his employer to really question his recruitment practices. Gi identified what he called 'systems of conformity' – which work against people like Shane. He was inspirational about using innovation – like AI and connective systems - to stop people employing or promoting people just like them; to exclude basic bias and instead find the exact right person for the job. The problem with systems of conformity, he told us, whether of recruitment or progression, is that they don't generate diversity. And Gi gave a simple reason why this wasn’t a good idea, particularly in the tech world. If all your employees are basically the same person, he said, when you ask a question they will all come out with the same answer. If you have diversity of thought, you will probably come up with a better answer, irrespective of the question.

He also talked to us through his amazing one_for1 programme which trains a young person for free, every time a business person is trained. This lead to a discussion about obtaining private investment in programmes with socially beneficial outcomes. Gi recognised that venture capitalists need 'a different story' – he told us the best way to attract investment for social outcomes is to lockstep them with profit and return. Gi is obviously an enthusiastic innovator - he ended by talking about extrapolating out from video games to find new organisational structures. I might volunteer to help him with that research...

Data is the problem

A large part of the discussion centred around data ethics. Just as Rose pointed to data issues providing a sticking point in education, Gi identified them as a key issue in his field. Information is his business, but he recognised employers will want to use the data he collects in a different way to the employees whose working lives it details. If we’re going to maximise the potential of these new technologies, we have to make some choices about data – how it’s gathered, how it’s interpreted, and who owns it. There's very little work in this space right now - we should be doing more.

If this is a topic you're interested in, the commission is looking for submissions from interested parties. Have your say by responding to our call for evidence here