menu_open Columnists
We use cookies to provide some features and experiences in QOSHE

More information  .  Close

AI as Personal Coach? Maybe. Three Ways to Make It Useful

30 0
latest

Take our Burnout Test

Find a life coach near me

Using AI to help you professionally requires understanding the limits of the technology...and yourself.

Human coaches can have even more impact on actualizing AI-generated tips into actionable steps.

AI has less understanding of the role of timing and human resistance to change, which needs to be factored in.

Over the holidays, a ChatGPT commercial featured a woman who had committed to run regularly. As she and her children ascended a hill (down the center of a street, in a picturesque town with no traffic, apparently), the AI plan scrolled down the screen. It makes sense that ChatGPT and other LLMs are advertising practical ways to improve your daily habits and reach goals. But recently another leadership coach I spoke to revealed that a client was using AI to help them with executive professional development strategy, alongside her as their “human” coach.

Honestly? I’m ok with that. I could point out that planning a workout routine to reach fitness goals is a less complex ask. But customization is key, and I don’t want to disregard it as apples and oranges. (I used to be a pretty serious triathlon competitor – so I say this with respect for the work it takes to figure out the right regimen for your personal goals.) The risk to me is not the use of AI for career coaching or any significant personal goal. The challenge is that the more you have to understand the unknown variables – in this case, other people – the harder it is for AI to build a plan.

It's pretty simple – AI is pulling information from the internet. The internet is vast, and the idea is that AI can synthesize a lot of data and discard what isn’t important for you. Customization is as much about eliminating the irrelevant. You could read ten books on how to train for a marathon – but you have to eliminate advice for experienced runners, older runners, runners competing professionally, all the things that don’t apply to you. But at the end of the day the inputs are all personal. You can tell AI your fitness level, when your race is, what your diet and sleep habits are. And AI can build a plan.

Here are three ways to leverage AI for career coaching, without losing your human lens.

Remember that artificial intelligence struggles to understand irrational responses

Career coaching – especially for people who are leaders – is by definition full of scenarios that involve other people. When I work with clients there is always a balance of what this person brings to the table (their leadership “fitness” so to speak, meaning experience, current success, self-awareness and general attitude), and the circumstances around the team they lead. Most of my clients need help understanding the impact of their choices on the team, and that is where a human expert can help. The answer isn’t just about what to do, it’s about deciding what it takes to make that work with this team.

AI is fantastic at things with rational outcomes. And it can pepper you with your likely outcomes because it needs two data sets: your personal inputs, which you mostly control, and the vast internet of options, culled down to your needs. Even your “irrational” emotions can be somewhat sensed from your responses.

But humans in a workplace environment introduce a much broader range of outcomes, and not all of them are rational. Office politics, years of experience, personal priorities all can lead to very unexpected results. It isn’t flat out irrational, as in, ridiculous. But it is hard to predict. Think about people who take less money for more personal freedom, or who succeed as much due to charisma as competence. Now build a prompt for that.

To be clear, human coaches are not oracles. I can’t tell you how all those people will react. But I can tell you what I’ve seen: clients I’ve helped, types of people, and examples of how they have behaved. And very little of that kind of information is on the internet. I’m not better than AI, necessarily. I just have direct offline access to that old analog database called human experience. (And professional training on how to synthesize it.)

Take our Burnout Test

Find a life coach near me

A human coach can help bridge where you need to develop, and where others do

AI is understandably hyper focused on you. First of all, it’s a product. I asked mine about a particularly polite and supportive response it gave (blame it on the day job but it is difficult for me not to consider its motivations) and it said unapologetically “I’m coded to keep you talking”. Wow. I guess that’s my job too, but I don’t craft my responses to keep a conversation going that has run its course.

This is most key in terms of the balance of what you need to develop, and what needs to happen in the team. AI will focus on the dataset that is most knowable – your inputs about yourself, your view, your goals, your emotional response. I suspect it has to reduce – I would call it under index – the fact that sometimes it really is a team member that is the problem, not the leader. AI will try to fix you, when maybe you really aren’t what needs changing. But that keeps you talking.

Timing is everything. But not the kind AI understands

In traditional coaching, it is typical to do a series of sessions over several months. The human brain can only do certain kinds of emotionally difficult decision making for so long in any one setting. And, actually practicing what you learn in coaching is part of the process. And people can’t implement multiple things at once – especially leaders who impact other people – without some pacing guardrails. AI loves to give a full, comprehensive response with a kind of “there you go – what’s next?” finality. But humans don’t process change like that. Timing and pacing are a huge part of the likely success.

AI can be a really useful tool to think through different options. Humans can help make the hard call

My colleague ultimately began to review some of the ChatGPT output with her client. And some of the information was genuinely helpful. The human conversation then focused on what made sense in her personal circumstances, what the team’s likely response would be, and how timing could positively impact that response. That isn’t nothing – in fact, it is a significant change in how human coaches help. But it also isn’t everything – and that is the part to really keep in mind. Your fellow humans will thank you.

There was a problem adding your email address. Please try again.

By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy


© Psychology Today