The debate regarding the benefits of AI in business (and other fields) continues to rage. Its use in coaching and training has been a particular area of interest. Advocates of AI will make the following claims:
- 1:1 coaching is expensive and not scalable, and AI will make coaching available to anyone anytime.
- AI coaching will be unbiased as it will be based on relevant, consolidated data and information.
- AI can process many data points, such as word usage and analysis, and immediately present insights and action plans.
- It is efficient and cost-effective.
Throughout history the creators of every innovation have highlighted the benefits, minimised the threats and told us that their innovation will change the world! Advocates will focus on the benefits while naysayers will focus on the risks. When the first video recorder was launched advocates claimed that we would now be able to replicate the cinematic experience in our lounge rooms at 10% of the cost, while naysayers claimed that the invention would destroy the entire film industry. Neither prediction proved accurate.
Similarly, fintech innovators and investors claimed that cryptocurrency would make financial systems more transparent and traceable – in fact, it has done the opposite. Mark Zuckerberg and social media investors claimed that Facebook would bring the world together and ensure we would all have equal access to each other and to important information. While social media does make it easier to keep in touch and share information, access is definitely not equal, nor is the information disseminated via social media all factual. In fact, in many cases, social media is pushing people apart, feeding social divides and causing many people to feel more disconnected from friends, family, the community and even themselves.
So, while all the above points may be true about AI coaching, with disclaimers, the real question is ‘Will AI successfully replicate the impact of human coaching?’ Of course, we are biased. Our name is Peeplcoach and ‘peepl’ will always be at the centre of what we do. But being cheap, accessible, scalable and quick does not make something more effective, impactful, relevant or even useful. And while we do a see a role for artificial intelligence and co-piloting in coaching, we do not believe that AI will ever totally supersede human coaching.
Artificial intelligence is best used to assist with repetitive administrative tasks that require the synthesisation of complex and detailed information. And AI has been utilised, somewhat successfully, in a variety of industries and fields. However, we do not believe that AI can replace softer skills and activities – those that require less data and more creativity and intuition.
Here are the reasons why we, at Peeplcoach, believe AI coaching cannot replace human coaching.
- It often draws on false or incomplete data. AI gathers insights from accessible data. Some of that data is verified and proven, but sadly, it is getting harder to differentiate fact from fiction. Even OpenAI CTO Mira Murati could not confirm when AI ‘hallucinations’ would be fixed. (The term ‘AI hallucination’ refers to wrong or misleading results returned by AI models due to such factors as poor data sets, inadequate or biased model training and faulty assumptions made by the model.)
- Even if the data is factual or validated, it is often biased. Much of the research and information shared online is based on the insights from a very select group of individuals – white and male! Until the 1970s women were still being excluded from health trials if they were of childbearing age or using contraception. (I guess if you were between 12 and 50 and a woman you didn’t get sick or need treatment?) So the data that AI has access to is already skewed.
- The effectiveness of coaching is part art and part science. Impactful and effective coaches blend experience, intelligence, information, data and insights with human connection, chemistry, trust, intuition and emotional intelligence. While AI can try to replicate ‘humanness’, it is not human. Great coaches will share their own experiences; AI cannot make you feel heard or cared for.
- AI can’t ‘read’ a situation and change the direction of a conversation to ‘break the mood’ if needed. Sometimes, the best coaching tactics are silence, changing direction, humour or the sharing of personal experiences to create self-awareness or insight. AI cannot do any of these.
- AI can only be used where data related to previous experiences has been collated. How will it solve unique and new situations – for example, when coaching leaders as they face fresh challenges such as a global pandemic?
- AI works best with cerebral data-based tasks and situations. Many coaching conversations focus on emotional causes – that is, why we act or fail to act – rather than on the academic process of a particular action.
- AI coaching is not cost-effective. In the article ‘AI chatbots lose money every time you use them. That is a problem,’ Will Oremus highlights the prohibitive cost of AI. And a number of companies investing significantly in AI, such as Adobe, are seeing profit downgrades. Of course, costs will drop – but when?
- The Harvard Happiness Study, the longest longitudinal study of human happiness ever completed, found that the number one predictor of happiness, long life and success was the number of connections you had with people. They can be deep but not frequent, like your school friend that you only speak to a few times a year but the conversations are significant and personal. Or they can involve the regular chat, banter and checking in that you have with your florist or favourite barista. And, of course, there is the regular contact you have with your immediate friends and family. As a coach, I put in a lot of effort into connecting and building rapport. This cannot be done by AI.
We can see the benefits of AI for supporting coaches with up-to-date and just-in-time information but we cannot see it ever replacing human-led training and coaching. Relying only on AI coaching would be akin to only having sex with a sex toy. It might work functionally, some of the time, but it is not quite the same without connection, intimacy and human interaction.
We went to the source, AI itself, for the final word on coaching and AI. We asked ChatGPT ‘Why will AI coaching fail?’ So here it is, from the AI horse’s mouth. AI coaching may fail for several reasons, including the following.
Lack of Human Connection
One of the primary reasons people seek coaching is for personalised guidance and support. AI, while advanced, may struggle to provide the empathetic and understanding interaction that humans can offer. Many individuals may find it difficult to form a meaningful connection with an AI coach.
Limited Understanding of Human Emotions and Context
AI systems may not fully comprehend the nuances of human emotions, context and interpersonal dynamics. Coaching often involves addressing complex personal and professional challenges that require emotional intelligence and empathy, areas where AI may fall short.
Inflexibility and Lack of Adaptability
AI coaching systems typically rely on pre-programmed algorithms and data patterns. They may struggle to adapt to unique situations or unexpected developments that arise during coaching sessions. Human coaches can adjust their approach based on the individual’s needs and responses, which AI may find challenging.
Ethical Concerns and Biases
AI systems can be prone to biases present in the data they are trained on, which can lead to unfair or inappropriate recommendations. In coaching, where trust and confidentiality are crucial, any perceived bias or ethical issues could undermine the effectiveness of AI coaching platforms.
Privacy and Security Concerns
Coaching often involves discussing sensitive and personal information. AI coaching platforms must ensure robust privacy and security measures to protect user data. Concerns about data breaches or misuse of personal information could deter individuals from using AI coaching services.
Complexity of Human Behaviour
Human behaviour is intricate and multifaceted, making it challenging for AI systems to fully comprehend and address. Coaching often involves helping individuals navigate their thoughts, feelings, motivations and behaviours in various contexts, which may exceed the capabilities of current AI technology.
Lack of Intuition and Creativity
Coaching sometimes requires intuitive insights and creative problem-solving approaches that are difficult for AI to replicate. Human coaches can draw upon their intuition and creativity to offer novel perspectives and solutions, which AI coaching systems may struggle to provide.
While AI coaching has the potential to augment certain aspects of coaching, it cannot fully replace the role of human coaches in providing personalised, empathetic and effective support to individuals.
Sources
- The hidden and high costs of AI and ChatGPT
- History of women’s participation in clinical research
- Have we reached peak AI?
- How many jobs will be lost to AI?
- AI and coaching: embracing opportunities amidst disruption
- AI hallucinations
- Adobe’s stock slides toward worst day in 18 months, as AI story will take time
