Hopeful & unsettled: coaching, schools, AI
Two conferences and one book last week. On Monday it was The Scottish Education Futures Festival at the Botanics where we discussed how Scottish education might look in 2040; on Thursday, Coaching for Good: Keeping Humanity at the Heart of Change at Edinburgh University Business School explored how coaching might evolve; throughout I was reading Jamie Bartlett’s How to Speak to AI - and this rapidly developing technology was a looming presence on both days.
I ended the week feeling both hopeful and unsettled, with predictably more questions than answers. At the education conference, we kept coming back to the importance of community, and to the attributes and skills that might matter most in an uncertain future - creativity, curiosity, ethical judgment, the ability to ask effective questions rather than produce the ‘right’ answer. On Thursday, the focus was related: what does human coaching offer that AI can’t?
Bartlett highlights the ability to use language effectively as the skill that will matter most in an AI world - articulating with precision and nuance and applying the same criticality to the responses that are generated. Being able to craft effective prompts is one outcome of this. But being able to do that relies on a complexity of thinking, an acute awareness of language, a self-awareness, and having sufficient knowledge to engage meaningfully. In education, these are attributes that can and should be developed across a range of contexts. They are rooted in culture: in the environment people create and inhabit together, and in the extent the adults around young people are themselves curious and willing to be uncertain or wrong.
There’s a tension then in any assumption that AI literacy should primarily be developed by using AI. Language has long been the defining characteristic of humans - it’s why we often see the sites that use LLMs as more than just machines - and language skills are developed in conversation, by reading and writing, by being with humans and by being human. The social dimension of learning is real - community matters. If we go straight to the tool, we risk undermining the very capacities the tool requires.
The unsettled feeling also came from a clear warning about assuming too quickly that there are things that humans can do that AI can’t. What I’ve learned this week is that these are few and diminishing rapidly. There’s also the uncomfortable possibility that if we hold too tightly to our ‘special’ status as coaches and teachers we risk an opportunity to democratise access. An AI coach can reach people who wouldn’t otherwise be able to access coaching, and Sam Isaacson’s presentation on Thursday showed how this might be done ethically and thoughtfully.
Research suggests, though, that people still trust the human more, particularly in complex decision making. They feel the human understands something the machine doesn't. While this might be objectively illogical, it perhaps also speaks to our hopeful belief in each other - in all our fallibility - and to the simple recognition present when in the company of another human. A coaching mindset, in schools, as elsewhere, can help make these connections. At its heart this is about approaching each other with curiosity, without judgement, and with a belief in another’s potential for growth. It’s a fundamentally hopeful stance and one that starts with the human in front of us, not the problem they present.
