A note(book) on AI

Note books have played a key role throughout history. It isn't time to stop now.

Many years ago, our first Engineering Manager gave us two invaluable pieces of advice. One was to never switch on a calculator before we had at least a rough idea of the answer (advice all the more relevant in today’s chatbot-frenzied world); and the other was to keep a notebook. We still do - and chances are, talk to any scientist or engineer today and out will come a notebook at some stage in the proceedings.

One of the most interesting books we read last year was on that very subject. In Note Book: A History of Thinking on Paper by Roland Allen, we were taken on journey through time, finding out about how ideas were developed, captured and shared by such exemplars of human expression and learning as Leonardo, Newton, Curie and Chaucer. The humble notebook was – and still is, we would argue – key to the process of learning, in particular self-learning. It captures problems and the process towards solutions, often learned from experienced practitioners showing how they tackled each challenge. It documents the journey from initial engagement to mastery of each subject, perhaps even discovering how things can be done in better ways. It is a working portfolio of what its owner knows and has done. If written up properly, it demonstrates competence: a living curriculum vitae.

Over time the notebook has been formalised into student logs and apprentice portfolios, to show how learners have understood a subject and – more importantly – shown how they have applied knowledge and skills, and demonstrated competence. To use it to provide evidence of that competence, we now have VACSR - valid, authentic, current, sufficient, and reliable – as a recognised way of showing what you did; how you used the specified skills, behaviours, or knowledge; what you learned as a result; and how you might apply that learning to similar situations in the future.

Plus, of course, there’s been the inevitable move towards digitisation. There are lots of examples, like Pebble Pad which has been around since 2004. These platforms are designed to capture progress through a range of media and formats, together aimed at describing each student’s the learning journey.

Why is this of interest to The Green Edge? Simple: the scale and speed of the (re-)skilling of millions of people across the UK for the green economy means we need a highly efficient and – ideally – a real-time way of qualifying competence.

For example, many electricians, plumbers, and other trades will be involved in the installation, commissioning and maintenance of new forms of low carbon heating and their associated control systems. This work might be intermittent, perhaps every other week or month, and each job might combine commonalities with some degree of uniqueness.  We need a way for all this to be captured, recorded and evaluated, to demonstrate the growing competence of the individuals involved in the work. A portfolio of greening competence, if you will, alongside other future skills around things like design, circular economy, digitisation and so on.

And this is not simply a case of do-it-once-and-you’re-done. As we see continuously across the world of sustainability, technologies are evolving, and competences will have to evolve alongside. We read, for example, in MIT Technology Review and elsewhere that new approaches are enabling heat pumps to reach higher temperatures, which could bring them in to applications like low-temperature manufacturing and steam for food processing. That’s new heat pump competences in sectors where heat pumps might not now exist. All in addition, of course, to new competences for next-generation heat pumps (and, presumably, their control systems) busy at work in heating (or cooling) the world’s buildings.

So, we’ve made the case for notebooks. But who’s going to do all this reading through – on paper or screen – of these notebooks to figure out who’s competent or not? Well, this has to be a candidate application for AI, doesn’t it? After all, we’re already seeing AI in education in the UK and in the USA, with new areas being explored all the time around curriculum development, administration, and assessment. From our viewpoint, having read a range of AI reports (with more coming every week) AI has real potential to around capturing and recognising competence, whilst recording emerging skills to inform emerging green and digital curricula.

Here’s where we see AI making a major contribution. First, around automated grading and giving on-going and instant feedback to students. Second, in skills and competence mapping, pulling out skills and competences embedded within a portfolio and make then explicit. Third, personalising learning paths: by capturing the skills and competences captured in a portfolio, additional actions and projects can be proposed to close off any gaps. Fourth, enhancing learning depth: using and assessing a living portfolio can identify those challenges that a student might be finding difficulty in mastering, flagging early and pointing to additional work to be put in. And last but by no means least, in driving continuous improvement, by looking at trends within a portfolio – and possibly within a business – to identify those areas where improvement can be achieved and the best ways for learners to do the needful.

We suggest there’s an important piece of work to be done here, to see how AI can be used like we described, to help close the green skills gap and support the delivery of the Green Jobs Delivery Group’s UK upcoming Net Zero and Nature Workforce Plan. We might envisage this being sponsored by the DfE, perhaps supported by the various progressive skills charities and drawing upon the expertise of bodies like the Ada Lovelace Institute, The Turing Institute and IfATE. Looking at the private sector, there’s a host of recruitment and staffing companies that are already using AI and could also help here.

The Net Zero and Nature Workforce Plan is due in mid-2024, so it would make sense to us to tackle this quickly in 2024. Takers, anyone?