ARE YOUR BEAMS ON?
What can orangutans teach us about how people learn to use AI? Given the rush and panic of the last month, you can be forgiven for having no clue.
So as the hypegeist missed it, we’ll right that wrong now.
A young orangutan has a lot to learn. Around 250 foods, many hard to get at, some only with tools. And eight years to master them.
Researchers in Sumatra followed 21 youngsters through those years. They watched adults closely – and did most of their own experimenting in the hours afterwards. By the time they left their mothers, those who watched and tried most had the best diets.
Watching, it turns out, sparks the trying. Trying builds the skill.
In our experience, people learn AI the same way: by seeing someone who knows how to get the best out of AI work through a task they recognise. Then having a go themselves.
But who should they watch?
In another study, 600 people played an online game, chasing a moving target across a 3D crater, alone or in teams of five. In some teams, a bright beam over each player showed how close they were to the target, so everyone could tell who to follow. Those teams beat solo players. Teams without beams just followed whoever was in view. They did worse than people playing alone.
Most AI rollouts have no beams. Everyone can see who has access and who posts most in the AI channel. Few can see whose AI use works: the output that survived checking, the error someone caught. With AI changing this fast, copying the busiest colleague can send a team the wrong way.
Embed, our AI adoption programme, is built on that loop of watching and trying. We work through people's own tasks with them. They practise on real work. Then we coach: what worked, where the AI got it wrong, how they caught it. Everyone finishes with a project their colleagues can see.
Show people the work. Let them try it. And turn the beams on.
But with a name like ours, we would say that.
The orangutans: Revathe et al., Science, 24 September 2026 – "How 8 years of social and individual learning interact to shape ecological competence in orangutans"
The crater game: Chirkov et al., Nature Communications, 23 September 2026 – "Payoff selectivity drives collective intelligence during dynamic resource tracking in humans"
And how Embed turns the beams on, from our How Enterprises Adopt AI series:
