Will AI disrupt the leadership coaching industry?

I've been having an intriguing time experimenting with ChatGPT on a number of fronts the past couple of months, as I'm sure many of us have. One of those fronts for me has been the field of leadership development, with an ongoing series of generative AI experiments alongside some adeptive self-coaching applied with the HBR Guide to AI Basics for Managers and Dr Tomas Chamorro-Premuzic 's excellent I, Human: AI, Automation, and the Quest to Reclaim What Makes Us Unique.

Some of those experiments have included:

• Feeding ChatGPT a whole bunch of my own leadership journal entries and inviting it to summarise, categorise, analyse, commentate, critique and recommend actions and learning/development activities in response, for individual entries as well as collections of them
• Coaching the AI to coach me, giving it specific leadership paradigms, frameworks and models to mediate the analysis and advice it provides, as well as particular personal learning preferences to inform how it responds to me and with what
• Getting it to roleplay different personas as a leadership coach as it responds to my inputs and questions (Blackadder was a particularly entertaining one, and it was even able to distinguish between the more slapstick buffoonery of Series 1 and the richer sarcasm of Series 2-4, as well as the different historical eras, calling for quite different language and tone)
• Probing it to explain its sources and rationales
• Relentlessly pushing it to think outside of the box and come up with genuinely novel analysis and suggestions (through things like 'mash ups')

And so much more. Experimenting with ChatGPT can get quite addictive and in some ways I'm almost thankful for the 'Plus Subscriber' GPT-4 limit of 25 responses per 3 hours - as a form of self regulation!

There are certainly still a lot of limits to what the AI can do and how reliable it can be, not to mention some dangers when it gets trapped in its own logic loops or doubles down on something too generic. It does make mistakes, and some interactions can absolutely end up somewhere close to 'uncanny valley.'

But overall I have to admit its potential is impressive. Really impressive.

Based on these experiments and the question I framed in the title (will AI disrupt the leadership coaching industry?), the answer for me at this point is probably "it depends." On a number of factors.

If we go with that baseline definition of disruptive innovation (that it refers to 'innovation that transforms expensive or highly sophisticated products or services—previously accessible to a high-end or more-skilled segment of consumers—to those that are more affordable and accessible to a broader population') then the answer is probably yes. Anyone can now access a form of information, advice and basic coaching that is relatively inexpensive, extremely well informed, with a good capacity to customise and personalise what it tells you (based on what you tell it).

On the other hand, if we consider leadership coaching (as with leadership itself) both an art and a science, then AI still has a very long way to go on the 'art' of things. The elicitation, appreciative inquiry, emotional intelligence, keen sense of perception and lateral thinking that really excellent leadership coaches can bring to bear is something that AI can't really compete with. Yet.

Even on the science side of the equation, having a useful coaching experience with AI currently requires the user to have a good capacity to coach the AI to coach the person (so to speak) - knowing what best to give it in the way of input (for example 'casing' current or past leadership experiences or 'casting' future situations and challenges) with at least some meta-knowledge around principles and approaches that mediate different interpretations of leadership and adult learning and development theory and practice.

We also need to acknowledge that, as human beings, many of us come to the coaching equation with quite complex backgrounds, needs and requirements. Leadership can be a very lonely gig. And a stressful one. It's not always easy to find the sort of safe space with another person where you can be truly vulnerable and be genuinely understood and challenged in the best possible ways. (Depending on your perspective, AI isn't or actually is a good option based on some of those factors.)

And (let's face it) there is in some cases a bit of ego-massaging that goes with being able to access personalised executive coaching at exorbitant hourly rates that others can't access or afford. Having an overpaid leadership coach is another way to assure some leaders that they are special.

At the very least, leadership coaches are going to have to come to terms with the fact that the more informational aspects of the role (the 'algorithmic' elements, if A = B then C) are pretty much already subverted thanks to AI and it is the novel, creative and uniquely human elements of leadership and the coaching compact that represent the 'safe ground' to continue doubling down on and excelling at.

But I think there are two warnings for professional leadership coaches there.

One, don't be so sure you aren't using an algorithmic approach already in a lot of your coaching, or that your coachees won't be capable in the future of developing the skills to 'coach the AI to coach.'

Second, the AI horse has already bolted from the stable and is learning and developing really fast; some of the novel, creative and 'uniquely human' in something like leadership coaching may not be as unobtainable or far away on AI's horizon as you presume they are.

What do you think, as a coach and/or coachee? Is leadership coaching ripe for AI disruption?

This is a?Leader TWIG?- the concept of (a)?growing something new?(a new awareness, skill or 'branch' to what you currently already know) but also (b) becoming equipped to 'catch on', realising or suddenly understanding something that is in fact right in front of you in the performative leadership moment (from the Gaelic 'tuig').