Evolving with AI: Setting the Bar for AI Coaches

How to create a future-proof AI Coaching Framework and Standards

The current rise of AI coaching since the release of ChatGTP highlights an urgent need for robust standards and ethical frameworks to ensure that high coaching quality is also implemented in the development of AI coaches. Human coaches can get accredited with professional coaching bodies to provide transparency and guidance for those seeking masterful coaching providers. But how can we tell if AI coaches really know and incorporate good maybe even masterful coaching practices?

Notably, the disparity between high-quality professional coaching practices and the varied quality in AI coaching implementations continues to widen, necessitating a thoughtful reassessment and update of existing frameworks. Luckily back in early 2020 Prof. Nicky Terblanche (PhD) 's foundational work A design framework to create Artificial Intelligence Coaches laid early groundwork for AI Chatbots in coaching - just half a year before the launch of ChatGPT. So all we need to do is explore a possible necessary revision that incorporates the capabilities and challenges of emerging generative AI models today and incorporate additional work that has been done by others since then.

Currently, as predicted, the market is about to get swamped with mediocre and sometimes even bad so called "AI Coaches" that Nicky called out in his post article on Artificial intelligence - you give coaching a bad name. But is it really AI giving Coaching a bad name, or is it a known problem for the coaching industry that coaching is used for many other things BUT professional coaching as we practice it? I would say that AI is just an amplifier and mirror of the problems that we already had since professional coaching has been established.

Building on Terblanche’s Pioneering Work

Nicky Terblanche's 2020 framework for AI coaching was seminal, introducing structured methodologies that have influenced the trajectory of AI applications within the coaching industry. His work, characterized by a prescient understanding of AI’s potential, provided a robust scaffold that integrated AI with established coaching practices, long before the widespread usage of generative AI technologies. Terblanche's principles — focusing on strong coach-coachee relationships, ethical conduct, and focused coaching outcomes — remain remarkably relevant. They continue to inspire a new generation of AI coaching developments that aim to enhance, rather than replace, the human touch in coaching.

Impact of Generative AI on Coaching

Unlike earlier AI technologies used in building coaching chatbots, which primarily relied on more predefined paths and responses, which we call decision-tree based or scripted approaches, or Natural Language Processing, which involved a lot of work, generative AI chatbots in coaching can react to and learn from interactions and evolve over time in nice human-like conversations. This shift not only enhances the capability of AI coaches to handle complex coaching scenarios but also introduces new challenges in ensuring these interactions remain ethical, unbiased, and especially aligned with professional coaching standards. The distinction between the earlier forms of AI used in Terblanche’s earlier work, Natural language based chatbots as used creating AI Coaches as done by Olivier Malafronte s Pocket Confident and Rocky AI Coaching & Personal Development Platform , and today's approach using generative AI calls for a reevaluation of the framework that guides the future development and deployment of AI Coaches as they continuously advance at great speed.

Revisiting Terblanche’s Framework

Reassessing Nicky's framework involves identifying components of this framework that are still applicable and those that require modification to accommodate the advances in AI. For instance, while the core ethical guidelines and outcome-focused approaches remain relevant, the framework’s technical aspects concerning AI's role and technical capabilities need a few updates to incorporate the nuances of generative AI.

Moreover, addressing challenges such as data privacy, AI bias, and the digital divide will be critical in shaping a responsible future for AI in coaching. Therefore we will also need to adjust standards and data policies meeting the newly released EU AI Act and it's implications for developing and using AI in coaching. More about that in my next newsletter.

Personal Insights from developing AI Coaches generative AI

• Building AI Coaches since the arrival of ChatGPT and large language models in general required a new skill-set that forms in different variants under the name "Prompt engineering". Writing, designing well structured instructions to frame generative AI models acting as Coach incorporating coaching principles and ethical alignments has become crucial to build these new generation of AI Coaches.
• Also understanding how different large language models work and tick has proven to be essential for deploying AI Coaches based on these models.
• When looking at ethics you also need to consider how they have been trained and/or fine-tuned and ensure they deliver a reliable output that is safe for end-users. Especially reliability and consistency is still an issue with Large Language Models as of today.
• The reason why we didn't integrate NLP (natural language processing) technologies before the arrival of GPT3 and then GPT 3.5, known as ChatGPT, was simple: as long as a scripted Chatbots provides better results and intent and emotion recognition back then was just not helpful enough to the coaching process with a Chatbot and did not align with good coaching practice, where the client is supposed to identify their own intents/topics and emotions. The when ChatGPT came out, allowing for a natural human-like conversation integrating this naturally the benefits of integrating this finally surpassed the challenges involved in integrating it.

Call to Action

I encourage all stakeholders in the coaching industry to engage in ongoing discussions about standards and ethical practices in AI coaching. Your insights, experiences, and critiques are essential as we refine and expand the frameworks that will guide the future of AI in coaching. Together, let's shape a future where AI enhances our ability to support growth and transformation in others.

Detailed Discussion needed for an Updated Framework

• What ethical considerations do you think are most critical with the introduction of newer AI technologies in coaching?
• How do you envision the integration of AI with human coaching practices?
• Are there specific models or methods you’ve found effective?
• What technological advancements have you observed that significantly impact coaching practices?
• How should these be addressed in the framework?