Navigating the many roads towards a mAIndshift
As recommended by keynote speaker Stefaan Verhulst, VAIA co-coordinator Ilse De Vos wrapped-up our symposium 'mAIndshift' with questions instead of answers. Each presented case provided a fresh perspective on lifelong learning and AI, so the question remains 'What now?' Ilse compiled a list of the most pertinent questions each institution organising lifelong learning about AI should ask itself:
- Who do we want to reach? Professionals, underprivileged people, students? How can we motivate them?
- What do they really need to learn? Technology? How extensive should that be? And what about ethics, data literacy, trust?
- How do we generate enthusiasm for lifelong learning about AI? What sparks excitement and what holds people back?
On 5 December 2023 VAIA hosted mAIndshift, the Flemish symposium on Lifelong Learning about AI. What an amazing afternoon!
- 1 thought-provoking keynote
- 6 inspiring flash talks
- 2 dynamic and energising panel discussions
I have taken Stefaan Verhulst’s advice to heart, and started my conclusion with two Questions: What brought us here today? What connects the people present?
Well, the people at mAIndshift share a mission: we want more people to learn more about Artificial Intelligence! However, each of us is following our own path to achieve this common goal. Our strategies may differ... But that is only because we interpret certain elements of our mission differently.
Who do we want to reach?
Bursting with energy, Steph Wright told us about the Scottish AI Alliance (Scotland), that wants to empower everyone in Scotland to constructively and critically contribute to discussions about how AI will impact their lives – including children. Knowledge is power, after all. This ties in with Elements of AI, presented by Ville Valtonen from MinnaLearn (Finland). These organisations set out to encourage as many people as possible to learn what AI is, what it can (and cannot) do.
This is quite different from the more targeted approach of Llanos Tobarra from UNED (Spain) and Solenn Daudu from aivancity (France). They target students as the lifelong learners of the future.
The AI4AL project focuses on adult educators to reach as many adult learners as possible on the European continent. As Angeliki Giannakopoulou explained, the European Association for Adult Education (EAEA) pays special attention to vulnerable and disadvantaged groups within society, for whom it is especially important to have access to education. Through their AI Learning Kit, MinnaLearn adopts a similar approach to facilitate workplace learning on a large scale.
At VAIA we opt for another, slightly different track: we target learners directly. We focus on anyone working in Flanders, both professionals and researchers. This is similar to the AI Campus (Germany): Michael Kircher’s project targets trainees in mechatronics and electrical engineering who combine work with studying.
Despite all these differences, we are united in the long run, because we want more people to learn about artificial intelligence.
What do we want people to learn?
Both Ville Valtonen and Michael Kircher referred to Bloom’s taxonomy, which classifies learning objectives into six levels of complexity.
Do we want people to be able to:
- give a definition of AI? However, as Steph Wright stated, not even experts can clearly explain what AI is. Perhaps we want them to be able to list the most commonly used AI tools?
- understand the general principles behind AI? Like garbage in, garbage out. Or should they have a high-level understanding of how specific AI tools work? Elements of AI aims to do exactly this.
- apply AI technology in their day-to-day lives? Our VAIA website lists many courses with this learning goal. A specific example that comes to mind are prompt engineering courses.
- analyse which tools are best suited to do which tasks? (For example: De wildernis van generatieve AI. Welke kiest u?)
- evaluate the outcome of these tools?
- create new AI tools or further develop existing ones?
Just how far up the ladder do we want people to climb? Do we want to push for as many AI experts as possible?
Personally, I think as a society we should expect everyone to at least reach the level of understanding – in general terms – how AI works and what it can and cannot do. I agree with Steph Wright and think it is important to ensure that everyone gets a say in deciding our common future in a world with AI.
To give people a nuanced voice, we need to build trust, but trust does not develop without basic understanding. Just like any other powerful technology, AI has risks. Understanding is the key to equipping people with the means to navigate these risks and to start trusting the technology.
Discover all the different definitions the Scottish give to AI 👇
AI is not one thing
If we focus solely on the tools and how to use them, we will always get caught up by reality. Many speakers highlighted the challenges arising from breakneck speeds at which this technology is evolving.
This is why VAIA is currently investing in knowledge clips that outline the theoretical framework. These clips can be watched by people in their own time, for example, before a hands-on workshop on how to efficiently and effectively use ChatGPT. Flipping the classroom has the additional benefit of reducing class time. This is especially important when catering to professionals who want to accomplish a lot in little time. Additionally, it addresses the problem of simply not having enough teachers. This bottleneck was another issue mentioned by many speakers – even being the raison d’etre of MinnaLearn.
Besides learning about AI technology, VAIA pays attention to the ethical and legal considerations, for example, the need to address stereotyping bias in image generators. In each course that we help organise, we add a section on ethics and law. We are lucky to have a policy partner in Knowledge Centre Data & Society who offers us the necessary – often complex – expertise.
Since AI relies on data, should we not invest in a more data-literate society? How many people really know that correlation does not imply causation? Stefaan Verhulst used this example to enlighten us on the importance of asking (the right) questions. As he pointed out, ChatGPT reminds us why good questions matter: ‘the value of data and AI depends directly on the questions we ask.'
Lifelong learning
How do we get people to learn about AI? How do you ‘teach people’ who have left school? ‘Lifelong learning’, learning after graduating, is more than obtaining extra certificates: it is a continuum.
The learning continuum
Informal learning
Informal learning is something we do all day and every day through conversations, things we hear, things we read. As AI becomes more ubiquitous, the informal learning opportunities about AI increase, for instance through friends or colleagues: ‘Have you ever tried DALL-E? Not yet, please show me!’ We all learn like this – by reading articles, viewing documentaries, chatting with colleagues, watching presentations and demonstrations.
Non-formal learning
Non-formal learning is not primarily focused on obtaining a diploma or a certificate, but it is very goal-oriented. A significant portion of professional lifelong learning in Flanders is non-formal, through various ad hoc courses and trainings, provided either by the employer or by educational institutions.
Formal learning
Formal learning takes place in a structured environment and leads to formal recognition. This is the domain of universities and universities of applied sciences and arts. A well-known form of lifelong learning in this domain is postgraduate programmes. In recent years, micro-credentials have been gaining popularity as a means of recognition in smaller, ideally stackable learning units.
The same question arises as with Bloom’s taxonomy: Do we want to motivate as many people as possible to take up formal learning on AI? Or is this simply not feasible for the typical adult learner, who is juggling work and life? At VAIA, we are not entirely sure about this yet. And we might never be.
For the economy!
Purely economic reasons are not going to cut it. As was mentioned during one of the panel discussions, AI is everywhere. It has been around for quite some time, but the arrival of generative AI unlocked the floodgates. There simply is no escaping it: AI is here to stay. A 2019 OECD Flanders Skills Strategy Report concluded that 29% of Flemish jobs will change drastically and that 14% of Flemish people even risk their current job becoming completely automated. Familiarising yourself with AI clearly gives you a competitive advantage, but this has not yet motivated a lot of Flemish people (nor companies) to take action.
For society!
What about societal reasons? We want a resilient and creative workforce that does not crumble when confronted with technological change. We want people to be flexible so they can adapt to the ever-evolving world around them and keep their skills and knowledge up to date, while maintaining a critical mind that asks the right Questions. As Stefaan Verhulst said, ignorance can easily be weaponised. He pointed out that we missed the window of opportunity with the breakthrough of social media. When it comes to AI, it is not too late yet. This can be a clear government goal, but will this convince many individuals?
For yourself!
Learning is fun! Each goal you achieve, gives you a confidence boost and a sense of accomplishment. In the newspaper article ‘Learning to play the piano benefits your boss’, Ans De Vos, head of the Lifelong Learning partnership in Flanders, explains: ‘By playing the piano, you don’t learn job-specific competencies. But you do develop skills that can be relevant at work. [For example], you learn that persistence pays off. Or that something you can’t do at first, isn’t impossible. Additionally, it can help you stay mentally fit. […] Those who have positive learning experiences – whether it’s learning Spanish or flower arranging – are more open to it in the workplace.’
Once motivated, there are still some obstacles
No matter how motivated you are, there might be barriers that keep you from actually engaging in lifelong learning.
- You might not have enough time.
- You might not have enough money.
- You might not find what you are looking for.
- You might not know what you are looking for.
The OECD studied the customer journey of lifelong learners. There are many obstacles that stop people from taking up lifelong learning, even though most people who have a job are interested in doing so. But it is important to remember that learning is vital to stay future-proof!
So, how do we encourage people to learn more about AI?
It is our duty to identify what encourages people to learn more about AI and to tackle what prevents them from doing so. Here’s what mAIndshift’s speakers did:
- Steph Wright stressed the importance of getting the terminology right. The Scottish AI Alliance decided against naming their ‘Living with AI’ project a MOOC (Massive Open Online Course), because it is not a course. She also stressed the importance of images. Stereotypes like robots and glowing brains – oops! – can be surprisingly unhelpful.
- Michael Kircher (AI Campus) highlighted the value of tailoring your content to your audience. At VAIA, we always try to include relevant use cases in our courses. And we are working on a strategy to better tailor the courses we ‘spark’ to the needs of the targeted professionals.
- Llanos Tobarra emphasised the need for learning paths. You really need to guide the learner from one module to the next – something that is on the VAIA to-do list for 2024.
But first, let’s take a moment to savour the success of a wonderful symposium for just a little while longer… We hope that all our participants were inspired to teach more artificial intelligence to more people. Are you with me?
Ilse De Vos
Ilse De Vos studied Latin & Greek at UGent and obtained her doctorate in Byzantinism at KU Leuven. She lived in the United Kingdom for five years, where she worked as a researcher at the University of Oxford and King's College London. In late 2015, she returned to Ghent as coordinator of an interdisciplinary research project on medieval Greek poetry and as a teacher of Ancient Greek. From 2020, she coordinated the interfaculty honors program Think&Talk and led the Humanities Academy as director. Since 2023, she has been co-coordinator of VAIA, the Flemish AI Academy that strives to ensure that more people (can) learn more about AI.
Lifelong learning is right up Ilse's alley. Every year she takes a course somewhere, from Gothic to Learning Psychology. In 2021, she completed the Specific Teacher Education program and started Business Administration.
