How can we avoid wasting money on AI? Let’s start with a story…
Once upon a time there was a restaurant called ‘Chalk & Fork’ whose menu consisted entirely of large slabs of chalk. Despite the friendly faces of restaurant staff, it was not terribly popular.
Those plucky patrons who could be persuaded to take a seat, rarely made it through a single course. To boost popularity, someone suggested breaking the chalk slabs into smaller ‘bite-sized’ chunks, but it seemed these remained unappetizing. The restaurant staff employed all manner of schemes: points cards, mobile delivery, marketing – all to no avail. The chalk remained unpopular. The restaurant owner eventually placed his hope in a new technology – artificial intelligence – which promised to do two things. One, to make it much cheaper and quicker to produce the chalk, and two, to render the pieces into interesting shapes – for example sculpting them into the customers’ favourite celebrity.
Sound familiar?
Learning in essence
Learning at work isn’t as complicated as some would have you believe. In essence, learning is driven by the challenges people face: they hit a problem and draw on the resources around them to solve it (or just figure it out for themselves). This means that for a person working in learning there are really only two things you can do:
- You can challenge people in new ways
- Or you can find out what challenges people already have, and create resources that help them. A resource might even be a colleague.
Despite this, many L&D people find themselves mired in activities relating to neither aspect of learning: they find themselves distributing content.
To cut a long story short, this doesn’t do anyone much good: people tend to avoid it, and there isn’t any return on investment for doing this. You might wonder why anyone does it at all, to which the answer is that this is what education does: come up with some facts & try to get people to memorise them. It’s an odd little ritual that doesn’t have anything to do with learning.
One of the great things about L&D is our shared aspiration to make a difference to people. What generally stops us from doing this are the conventions around content distribution.
Where AI comes into all of this
The latest technology to be enlisted in this fruitless pursuit is AI. And so you will face a choice: whether to apply AI to learning or education.
If you choose to apply it to learning there are two worthwhile applications:
Use AI to create challenges
AI can create challenging situations and give you great feedback as you practice tackling them. For example, leaders practising difficult conversations and conflict management, sales people practising pitching, procurement people practising negotiations. In future these experiences will extend to more complex procedures of the kind that flight simulators exemplify.
Use AI to guide performance
You already use AI, in your car, to get from one place to the next. Your GPS tells you what to do, improving your performance and making your life easier.
Now imagine GPS for everything. This is especially important in a world where people don’t stay very long in a job and training is costly and wasteful. Remove the need for training by building performance guidance systems.
The most obvious application would be an AI buddy (chatbot) that helps your new starters get up to speed quickly. You could also capture scarce skills (e.g. HR expertise, engineering expertise) in the form of an AI Zoom character that can join calls as required and help you realise a self-service operation.
Avoid wasting money on AI…
There is also a completely pointless application of AI – what I call the ‘more crap, quicker’ option.
You could use AI to churn out more content. AI will merrily spit out a mountain of modules in a flurry of new formats – AI lecturers, podcasts and summaries. But since these are neither useful performance support for the challenges people face, nor new challenges, you will continue to grapple with the same old L&D dilemmas: poor engagement and questionable ROI.
Today, most of the applications of AI to learning that I come across fit this ‘more crap, quicker’ model. Avoid wasting money on AI of this type.
Step away from convention
One of the great things about L&D is our shared aspiration to make a difference to people. What generally stops us from doing this are the conventions around content distribution.
We’re at an interesting juncture with AI: IT folk avoid getting involved in the ‘people stuff’, and L&D folk in the ‘IT stuff’. But if we don’t step in soon and lead areas like performance guidance and expertise capture, then systems like Microsoft CoPilot will lead the way. And senior stakeholders will be left wondering if they need an L&D department at all.
Your next read: Generative AI: The implications for L&D
