When I was a content marketing executive at MindLeaders in 2012 I had big ideas. One day I sketched out on a piece of paper something that might bring everything I then felt we needed to learn into one place: a consumer-grade experience for any kind of learning content to play, but embedded alongside a space for notes and reflection, all linked to people and goals. To my astonishment, the following week I was shown something the company had built already that resembled it: a holistic skills platform (this is 2011, remember!):
Assessments, career planning and learning in one place, connected by a personalisation engine based on skills. The idea was compelling. Powered then by fixed logic and not machine learning or generative AI, it would nevertheless have been a self-organising skills platform. Many would attempt it in the years after.
Yet twelve years on, we are still pursuing this dream. Today plenty of smaller platforms and some larger ones have the capability to integrate all the skill functions. Indeed, the claim of being a ‘complete learning platform’ is so common that it has become rather unhelpful. Despite all this, I have yet to see it work in practice for a large organisation. Instead, big players have emerged in each of the skill categories. Workday and SuccessFactors dominate HR. Intelligence and role matching is the domain of Eightfold, Gloat and Techwolf. EdCast and Degreed lead in learner experience.
Each skill platform only weakly echoes the functionality of those in the adjacent categories. And each treats skills differently. The struggle now is to integrate these systems in a single experience. It is hard to do this. In fact, it often seems like we haven’t moved a single step closer to the original dream of unification.
It’s worth reflecting, though, on what has been achieved in those twelve years. In the field of learning experience platforms we have accomplished things like:
Alongside this we have made significant strides in the automation of mandatory training content: the category of learning management systems. Systems like Saba (acquired by Cornerstone) and Intellum introduced incredibly sophisticated ways to launch and track pathways of content to different audiences in granular ways.
Instantiating these ideas at the level of technology and platforms, creating the expectation of a platform that ‘just works’, is a significant achievement. Perhaps because of this success, the continued shortcomings of our platforms feel painful:
And now, of course: we have generative AI.
Although limited to the fairly high-level knowledge (not much beyond what is immediately accessible via Google and Wikipedia), tools like ChatGPT, Claude and Gemini are capable of teaching the basics of almost anything.
That’s what first struck me about gen AI: you could use it to try to do things you could not do before, and not just to do something you can already do faster.
So why not use AI to pursue the dream of a self-organised skills platform?
The dream feels more relevant than ever.
Talent development teams are faced with greater expectations and fewer resources than ever. It is not uncommon to have a single person responsible for learning technology, content and events for several thousand people in several languages.
What if they had access to a technology for learning that really was turn-key?
Google doesn’t require manual intervention to list a new website (as many of its early competitors did: Elon Musk’s first internet business was little more than a directory). The key innovation was to crawl and rank the web automatically. Self-organising.
Self-organising systems underpin personalisation. Think of the mind of a child learning basic skills from her sense-impressions. The child has her own constantly shifting goals: to eat, to drink, and above all, to play. But the concepts of the things before her emerge on their own. Amazingly, solid categories develop to organise the amorphous mass of images, sounds, smells, tastes and tactile sensations. Things are filtered out. Language and grammar form an underpinning structure. The child’s personalised learning engine emerges from the self-organisation of her world.
Here is what a self-organising learning platform for organisations needs to do.
And the self-organising system takes over:
In Filtered, we can now leverage our own indexing algorithm to make a learning system self-organising in exactly this way.
So what do you do next? That’s the important part here. The tech - algorithms and AI - allows the ecosystem to be self organising: always indexed, discoverable, and therefore personalised.
But on its own, this is not going to be aligned with business goals. We do that through the main customer input of defining skill priorities. By adding alignment with business goals through the definition of skills priorities, the self-organised skill platform is always channelled in support of business strategy. Understanding and updating those priorities, and measuring progress, becomes our new area of focus, not manually organising content.
Get in touch if you would like to explore that.