eLN Connect: AI, Behavioural Change, and Elephant vs Rider!

November 20, 2018

It’s a slightly strange title for a blog, granted, but the talks from last Thursday’s eLearning Network ‘eLN Connect’ event gave us a lot to think about – and it’s clear that the twinned topics of tech and human psychology are as alive and kicking in the learning space as ever.

Toby Harris – Vice Chair of the eLearning Network – kicked off the day with some context around the shift in technology across our day-to-day lives; “as we move into 2019, we stand on the threshold of huge change”, he asserted. “Against the backdrop of mass, open communications, powerful computing in our pockets and ‘always-on’ networks, we could go as far as to say we’ve moved from a society run on oil, to a society run on data!”

“The good news about the incredible depth and range of technologies now available to us is that we’re seeing that L&D is finally dragging itself into the 21st century”, he continued. “Go back thirty years, and I think maybe we were designing learning interventions that were unrestricted by the ‘traditional’ formats that we now conform to – and excitingly, we’re getting back to that. The tools available to us have benefited from wider advancements in technology, and what we’re moving towards is a model of quality, live experiences, complemented by data driven resources. Personalisation is the key – and truly achieving that relies upon a need to reskill; a need to tackle resistant organisational cultures head on; and the need for marketing practices to become commonplace in learning and development.”

So what did keynote speakers Donald Clark and Julie Dirksen make of the current L&D landscape?

Our top takeaways:

1. The popular AI rhetoric makes it sound far scarier than it really is

“Ask people how they learn, and the number one answer will be Google, followed by YouTube”, says Donald Clark, Tech Entrepreneur and CEO of AI company Wildfire. In Donald’s view, we need to wake up to the reality that AI is not some far-distant concept; in fact, it’s completely embedded in what we do already. Smart phones? Running AI. Social media? AI. Popular voice-activated home devices (Alexa)? You guessed it…

“It’s not just our day to day lives,” he says, “it’s going to change the workforce, and we are duty bound to pay attention to those two small letters.”

Yet, if we rewind to the mid 2000s, and there were some incredibly bold stats flying about. Here are the top two that Clark ridicules as being ‘absurd:

63% of jobs that will be performed by children starting primary school today don’t even exist yet
74% of jobs will be automated in the next 2 decades

“Why do we feel the need to scaremonger like this?”, he asks. “Yes AI is highly smart at doing very specific things”, (think the algorithms running your phone, or social platform), “but it’s also extremely stupid at doing other things.” He follows this statement with an entertaining anecdote about his smart-hoover; essentially fantastic at running a specific task and learning as it goes, but rendered completely useless when faced with an unfortunate mess left in the middle of his carpet by his beloved pet dog (!)

Clearly, the point is to be a little discerning about what we employ to tackle a certain job – a machine, or a human. Where learning is concerned, the mistake is not recognising where AI has the potential to add huge value to build the same personalised, frictionless experiences that we’ve come to expect in our consumer lives.

As an amusing counterpoint to Donald’s AI-hoover versus dog mess, here’s his little list of ‘things that make brains bad learners’:

Take 20+ years to educate
Inattentive
Conative overload
Forget
Demotivated
Social and cognitive biases (ineducable)
Sleep 8 hours a day
Can’t download
Can’t network
Die

…Something to think about the next time you’re designing a learning intervention!

2. Our eagerness to draw conclusions from data is admirable, but often misdirected

We absolutely know that getting a handle on our learning data is key – but it goes far beyond simply mining out what you have and drawing a conclusion. “Making stupid inferences without understanding what we’re looking at is why we’re tied in knots when it comes to data,” says Donald. He showed the following quote to illustrate the point. Crude? Possibly. Fair point? Certainly.

“If we’re going to derive helpful conclusions, we need large, clean datasets, and to build the expertise to draw sensible conclusions”. The part about knowing what we’re trying to achieve with our data – the why, if you will – is blindingly obvious, but you’d be surprised how often people dive in without this crucial sense of direction.

3. Human decision making is a battle between elephant and rider

In stark contrast to the plethora of content about technology, Julie Dirksen thoroughly entertained the room when she brought learning back to its fundamental focus (people) in her session: ‘Talk to the elephant – design for behavioural change.’

“When we think about behavioural change, we really need to interrogate the reasons that people do or don’t take action”, she explains. “Let’s be honest, having the knowledge isn’t the thing that’s going to make the change between current state and future state.” Here are some of the potentially relatable scenarios she backs this assessment up with:

“What’s the connection?” she continues. “Well, people already know the facts (and risks), but they don’t change. Telling them more emphatically or louder isn’t going to do anything – and that’s so often what we do with ‘training’.”

So here’s how our brain splits down:

Who makes the decision? The rider or elephant?

“We make very different decisions for our current self and future self,” she continues. “Let’s think about the alarm going off – you know you need to get up (rider) but right now you’re so comfy and you want to stay put (elephant). So who wins?”

Julie boils decisions about behaviour down to a simple equation: (Value – Effort) x Social = Action

She explains that the further away the reward or consequence of your ‘elephant’ (innate, animal brain) leading the decision making process, the less you perceive it – or indeed the less important it becomes to you at the moment of decision making. This is called ‘hyperbolic discounting’, and it starts as soon as ten minutes after the point of intervention.

If we translate this elephant/rider set up to the learning process, it really forces us to interrogate what we’re putting in to learning interventions; and how we’re satisfying the need for instant gratification or demonstration of consequence.

“What we’re looking at is the need for more visceral learning experiences”, concludes Julie. “If you’re working in an environment where the consequences of your actions aren’t immediately obvious, it can be tough to illustrate the rationale for changing behaviour.” One example here was the need for ward staff to wash their hands to prevent the spread of infection. Another example was our hugely wasteful use of disposable paper products and the negative impact on the environment. In both cases, she talked about the consequence not necessarily emerging at the individual level, but rather at a group, or systemic level – which again removes the immediate cause/effect feedback. In both cases were illustrations of behaviour change programmes that visualised the potential impact of seemingly small actions; green paint was used in one campaign to show the spread of infection in a hospital, and an immersive VR experience showing tree-felling was used to show what the cost of disposable paper really is.

“In order for learning interventions to be successful, we need to ‘bring the elephant along with the rider’”, she says; “we need to recognise the need to make learning experiences tangibly meaningful, and understand that as humans we are wired in a way that requires immediate feedback for our actions.” It’s certainly food for thought in terms of the way we design learning interventions within wider behavioural change programmes;

A big thanks to everyone that came along to see us at eLN yesterday. We will be making a copy of the slides from Meg and Alexandra’s afternoon presentation available in due course. In the meantime, if you’re interested in learning more about our research, take a look at our next insights report launch event here.

Previous Article
Learning as a Performance Enabler
Learning as a Performance Enabler

The start of a new year is the perfect time for learning teams to look ahead and explore new ideas and tech...

Next Flipbook
L&D Agents of Productivity
L&D Agents of Productivity

Why learning and development is key to unlocking UK productivity.