Is AI coaching a realistic proposition? Well …….

Since the release of ChatGPT in November 2022, the world has gone Artificial Intelligence (AI) crazy. The optimists are saying AI is great and it can solve the world’s problems. The pessimists have a Doomsday narrative; Terminators are coming, and we need to stop. The realists are looking for opportunities to deploy AI in supportive roles, to enhance human capabilities and reduce mundane tasks.

Regardless of how you perceive AI, the world has turned an innovation corner. Like electronics, computers, the Internet, smartphones, AI is changing the world we live in. Period. Over time, some roles will be replaced by intelligent machines. Do not despair as new, potentially unimagined roles will be created. But what will the impact of AI be on coaching?

There are a wealth of different areas of innovation for AI. Whilst no one area will have an immediate, direct impact on coaching, several could be combined to create a more interesting challenge. For example:

• ChatGPT is a chatbot style implementation that has emerged from an area known as Generative AI. Generative AI takes vast amounts of input data and uses this to create new content. In the case of ChatGPT, this content comes in the form of human-like answers to questions.
• Computer vision is an area that enables machines to “see” and interpret the real world. One capability in this area is the ability to interpret voice data from people speaking.
• Another capability in the Computer Vision space is the ability to interpret video and determine a human’s emotional state.
• Generative AI can manipulate images of humans, to add lip sync and even some physical emotions. This manipulation can create realistic video and almost undetectable Deep Fake’s have been created of famous people.
• A third area of Generative AI is text-to-voice generation. This intelligent capability can determine the sentiment in the text and add tone and change the speed of generated audio.

By combining the above areas as part of a video call or app, it could be possible to understand spoken language, interpret body language, pass this information to an AI chatbot such as ChatGPT, have it generate responses, and have a seemingly human conversation with relevant, spoken words and body language. A very crude implementation could be created today. In fact, OpenAI, the creators of ChatGPT are releasing a new feature under the blog post ChatGPT can now see, hear, and speak.

But what impact could this virtual, AI-driven person really have? Could they become a coach?

In its most basic form, coaching is the ability to connect with another human being on a profound level. The coach actively listens at a deep level, whilst observing the coachee to determine their feelings. The coach processes these streams of information to try to determine what the coachee isn’t saying, what they are hiding, if they are lying to themselves. The coach then asks insightful questions to shine a light on this unintentional misdirection. When clarity is achieved, the coachee can identify the next, best steps given this new clarity.

But coaching quickly surpasses this basic form. Where AI uses statistics, and logic, and algorithms to determine the most likely conclusion, humans’ reference vast amounts of experience, and intuition, and imagination, to connect dots that aren’t even visible. Good coaches are able to see what isn’t being communicated and ask insightful questions to expose it. Similarly, whilst AI is able to detect emotions from physical behaviour, and potentially reflect this back to a coachee in words and manipulation of an image, good coaches are able to combine a coachee’s visible behaviours, mannerisms, metaphors, specific, key words, to foster trust and openness, and true human empathy, by relating to the experiences, challenges, and emotions of the client.

But perhaps one of the most critical areas in which human coaches surpass AI is their ability to remain un-biased, their integrity, and their ethical standards. The foundation for ChatGPT and most forms of Generative AI is vast amounts of human data. Human data is littered with bias and therefore, so is generative output. There have been numerous examples of where Generative AI has “behaved” inappropriately – and this is an exceptionally challenging issue to resolve. It goes right back to the origins of the data.

Whilst there is currently no real challenge to human coaches, there are some basic AI building blocks that can be combined to create a crude representation of an attentive, solution focussed human. But AI is evolving exponentially and, over the coming years, expect to see technology solutions in the coaching space. With the current rate of innovation, it is impossible to predict how good these combined areas could become but, to ensure your longevity, focus on being the best, most creative human coach you can be.