Adrift in a sea of AI

I'm currently on my way to Learning Technologies 2024 - I always love attending this conference even if I do need a lay down in a dark room for quite a while afterwards! There's so much to look forward to and I can see so many great talks, but one topic is once again dominating the agenda - AI.

I'm curious and dubious as to how the next few days are going to pan out with regards AI. I know that Learning Technologies is a huge sales opportunity for lots of vendors and I expect to be told how wonderful each different provider's AI solutions are and how much they can help me and my team. But I'm also a little nervous about this mad rush to embrace AI and see it as answer to all of our problems. So I want to get my thoughts about it and the potential risks I see out there before I get confronted with flashy sales people trying to sell me things I don't really need.

What questions/issues is AI answering for you?

For me, this is a biggy. It feels like AI is being billed as THE answer, but people aren't necessarily sure what question it's actually answering. But its the new shiny kid on the block and everyone wants to get in on it. I was on a really interesting webinar with Ross Stevenson the other month where he made a series of great points about AI. One of the most important one's was around having a clear use case for it. Really understanding what you need AI to do for you. Rather than just buying into it and then hoping it'll solve problems, but then it not really solving any problem any better than any of your existing solutions. This really should be the case for all learning technology purchases - know what problem you are needing it to solve - don't just buy in to the hype of it.

Say hello to more content dumping

Over the last few years, I've been heartened by a move away from the practice of dumping 100s/1000s of pieces of content on to learners. I'm not hearing as many ambitions to be the "Netflix of Learning" (Sigh) any more - which is great. My worry is the advances in AI are bringing this back full circle. I fully expect to be told how much faster AI can produce content, and how my teams could produce so much more content using it. But should they? Surely we've reached a place now where we're of the understanding that quality and relevance trumps quantity?

Quality is an interesting point too. There's an awareness of issues of accuracy and hallucinations of AI which risk the production of poor quality or inaccurate content. Part of the issue is how your generative AI is trained and what it is trained on. The GPT in Chat GPT, stands for Generative Pre-Trained (GPT), but unless you're buying your own GPTs, how do you know what materials the AI has been trained on? if its just been trained via the wider internet and Google, ask yourself this, do you believe everything you read online? Of course not, because you've got the ability to critically analyze your sources and seek second opinions - does your AI have that?

My overriding worry with this technology is that its being billed as a panacea, its the answer to all of our problems. I also worry that senior members of organisations may buy in to the idea that content dumping is the way forward again and that they don't need as many L&D people any more because AI can do it instead.

Over the last few years, I've seen a rise in authoring tools billed as being so easy to use that even subject matter experts can use it to create elearning, and you can turn a text book into digital learning really quickly! But no one is stopping to ask whether you should! Instructional design is a skilled profession which at times feels like its diluted by the impression that anyone can do it. No they can't, not well at least. I worry that AI will exacerbate this trend further.

AI still needs people - it won't replace team members

But it's not all negative. AI, when used in the right way can be a great assistant to an instructional designer or learning professional, but the key here is "assistant to". AI still needs a human pilot. It can introduce some great efficiencies into your work, can speed up processes but it still needs a human being to drive it.

A partnership between human pilot and AI assistant can result in great achievements. One of my team had a play with Descript the other week. They were impressed, but also pointed out the need for a clear use case for it. They said that it felt like a great tool for a YouTube content creator who needs to create a large volume of content regularly, but for our needs there isn't as much of a use case for it. They did say though that the audio clean up tools were amazing on it and it saved him a lot of time cleaning up audio for a podcast. Again, a clear use case.

A safe space to practice

I did get to pilot some incredible AI a few weeks ago which had been produced by the Greater Manchester Cancer Academy and a company called Re:course. This was an AI solution which created practice environments for having significant conversations with cancer patients. It made use of AI characters who were able to hold a conversation with you and had their own unique personalities - one was confused about his condition, the other was frustrated and agree about the treatment her partner (character one) had received. It was an amazing and lifelike tool to use to practice these conversations and provided a safe space for professionals to practice in. This is the kind of innovation AI could and should be being used for.

This work is innovative because traditionally advanced communication skills training in healthcare is provided using actors at great cost to all the participants. It also has limits on the number of people who can attend which limits the reach of the training. Introducing an AI based practice environment with life like AI characters expands the reach of this work massively whilst also making it more inclusive to those who aren't able to travel or afford the more traditional training.

Enjoy LT2024 and all the AI talks

So hopefully like me, you're going to go to Learning Technologies and engage with loads of great AI talks, be open to some amazing possibilities, but also maintain a healthy skepticism about some of the claims made about what it can do for you.