A contrarian take on AI in L&D
What if 2024 isn’t the year you need to embrace artificial intelligence?
Before joining the hallowed ranks of L&D, I spent two years teaching English at Sciences Po Bordeaux — a university specializing in political science.
Because most of my students were pursuing careers in public service and international relations, much of our time in class was dedicated to debating the big issues of the day.
Often, this would require me to play devil’s advocate.
If it looked like consensus was about to emerge in the classroom, I would deliberately stake out the opposing position to keep the conversation flowing. I’ve continued this trend whenever in conversation with my L&D Dispatch colleague, Ross G.
Over time, this impulse has become a deeply ingrained habit. Now, whenever I encounter a widely held belief (even one that I share), my first instinct is usually to ask: ‘What makes you so sure?’
A recent example of such a belief in L&D (or, at least, L&D on LinkedIn) goes something like this: ‘You need to start exploring AI this year. If you don’t, you risk being left behind’.
Before I go any further, I’d like to issue the disclaimer that I kind of buy this argument.
But here’s why I think it might also be absolute hokum:
🤔 There are very few examples of AI being deployed effectively in L&D
As Donald Taylor and Egle Vinauskaite’s recent report on AI in L&D demonstrates, excitement about artificial intelligence vastly outstrips meaningful application. This isn’t to say that there aren’t examples of good practice out there (*cough* our collaboration with Learning Pool *cough*), or that the technology doesn’t have the potential to revolutionize the industry. What it does show, in my view, is that there remain significant barriers to implementation of AI. Until these are overcome, most L&D teams will simply use AI to do what they’re already doing. They’ll just do it faster, and at lower cost.
💩 The easiest thing to do with AI is probably the wrong thing
Faster? Lower cost? That sounds compelling, right? I’m not convinced. As Ross G wrote last year, the problem with generative AI tools like ChatGPT, Bard and Claude is that they are ‘yes men’. Ask ChatGPT to write a health-and-safety course for you, and it will answer ‘Certainly!’, before spewing out the perfect pastiche of a boring e-learning module. This might seem like a strange thing to say as someone who works on a content team, but I don’t think content creation has ever been L&D’s problem. Creating the content is the easy bit. Knowing what to create (and, importantly, what not to create) is where things get tricky. If you think your LMS is bloated now, just imagine what it will look like when the cost of content creation drops to zero.
⏳ Things will change this year, but not as quickly as we might imagine
Last week, everyone from Mind Tools got together for our first ‘Connections’ event of the year. On the second day of the event, our friend Myles Runham from Fosway Group delivered a presentation on AI. The presentation included the following quote, attributed to Bill Gates:
‘We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten. Don't let yourself be lulled into inaction.’
I love this quote, and it feels like an apt description of the current state of play for AI in L&D. While I don’t necessarily believe that you’ll be left behind if you don’t jump onboard the AI train in 2024, that doesn’t mean you should be ‘lulled into inaction’, or bury your head in the sand.
🤖 AI is coming for you, whether you like it or not
Finally, even if you’re not paying much attention to AI yet, there are a few reasons it will start becoming difficult to avoid in 2024. With every tech company rushing to incorporate generative AI into their existing product suites, many of the tools and applications we use every day (Microsoft Word, Google Sheets, Slack, LinkedIn, etc.) are set to include an expanding array of AI features. Over time, AI will likely become a foundational technology that’s built into every tech product. You didn’t care when Word corrected your spelling — so why should you care if it starts making copy suggestions?
What are your thoughts on AI in L&D? How are you planning to incorporate it into your practice? Get in touch by emailing firstname.lastname@example.org or reply to this newsletter from your inbox.
🎧 On the podcast
Jackie Kennedy is Learning and Development Lead at London Borough of Camden, where work contexts range from libraries and schools to waste management and social care. How do you develop managers in these diverse contexts, with a public sector budget?
In this week’s episode of podcast, Jackie joins Ross G and Owen to discuss the unique challenges faced by local government, and how to create a management-development programme without providing any ‘teaching’.
Check out the episode below. 👇
📖 Deep dive
If AI is over-hyped, automation is under-hyped.
As I wrote a couple of weeks ago, one of my goals for 2024 is to get on top of admin — not by spending more time on it, but by exploring opportunities to automate repetitive, manual tasks.
In a recent edition of his newsletter, Steal These Thoughts, our pal Ross Stevenson breaks down how Zapier’s L&D team did exactly that, leveraging their own product to save 1,000 hours annually and minimize the risk of human error. They did this by:
Analyzing tasks: Identifying and mapping out all the tasks that were potential candidates for automation.
Selecting and integrating tools: Creating workflows that connected different applications, handling everything from workshop registrations to feedback collection.
Improving iteratively: Continually refining this process, with feedback loops set up to ensure that automation met the dynamic needs of the team and the business.
Of course, Zapier’s L&D team had an automation tool close at hand, but I still think this case study is an inspiring call-to-action.
Stevenson, R. (2024) ‘How This L&D Team Reclaimed 1000 Hours To Unleash Real Business Value’. Steal These Thoughts.
👹 Missing links
Since 2012, the text comprehension scores of 13-year-olds in the United States have dropped by seven points on average. This has caused alarm on both sides of the political divide. But what has driven this collapse, if it can’t purely be explained by the impact of Covid lockdowns? According to a recent, soon-to-be-published (though not yet peer-reviewed) study by researchers at Columbia University, the answer might be down to the effects of reading on screens vs paper.
As we’ve noted in previous editions of this newsletter, a common pitfall in L&D is failing to think about how you will encourage people to engage with your interventions. In this article for Training Journal, regular Mind Tools collaborator Houra Amin explores strategies for maximizing engagement, including social proof, storytelling and nudges.
The ‘compliment sandwich’ (more accurately known as the ‘s*** sandwich’) is a common method of giving difficult feedback. The only problem is that it doesn’t work. As Adam Grants points out in this newsletter, research shows that primacy and recency effects are powerful. Even if the person you’re speaking to thinks your compliments are insincere (which they probably will), they’re more likely to remember them because of their positioning at the start and end of the conversation. It’s time to take this sandwich off the menu.
👋 And finally…
This article from the New York Times caught my eye. In a special version of a ‘Skinner Box’, photographer Augustin Lignier set up a button which released a dose of sugar and took a photo whenever it was pressed by one of the rats inside. Interestingly, the rats seemed to enjoy taking selfies, sometimes ignoring the sugar but continuing to press the button.
Thanks for reading The L&D Dispatch from Mind Tools! If you’d like to speak to us, work with us, or make a suggestion, you can email email@example.com.
Or just hit reply to this email!
Hey here’s a thing! If you’ve reached all the way to the end of this newsletter, then you must really love it!
Why not share that love by hitting the button below, or just forward it to a friend?