Seite - 57 - in Coaching im digitalen Wandel

Chatbots as an instance of an artificial intelligence coach 57 For a chatbot to realistically mimic a human coach, we should strive to create more complex chatbots (open, long, generative-based). This borders on what is known as Artificial General Intelligence (AGI) – machines able to conceptual- ise and transfer knowledge on a par with or better than humans. In the context of organisational and life coaching, Kamphorst (2017) pro- vides us with some helpful criteria for building a realistic coaching chatbot: 1. Social ability: The coaching relationship (rapport) is considered by many the most important determinant of coaching efficacy. To build rapport, the chatbot must have the appearance of social awareness and the ability to »bond« with the client. 2. Trust and credibility: Trust has been shown to be an important predictor of a strong coaching relationship. The chatbot must therefore demonstrate expertise in order to come across as trustworthy and must be able to main- tain trust. 3. Context-aware: The chatbot must be able to tailor a conversation based on the context of the client. One of the future versions of Vicci will use person- ality type as the cue for context-aware conversations. 4. Proactive: The coaching chatbot needs to invite conversation instead of sim- ply responding to client input. This will help to stimulate reflection, a crucial element of successful coaching 5. Theoretical underpinning: It is important that the chatbot uses an underlying theoretical model, such as adult learning (e. g. Kolb) or psychology (e. g. CBT), to direct the coaching process. This is akin to a human coach whose coach- ing model should be underpinned by tried and tested theoretical models. It appears that we have a relatively strong base and helpful guidelines to show the way forward in creating realistic coaching chatbots for workplace and life coaching, but what about the ethical aspects of such an endeavour? Ethics Mustafa Suleyman, CEO of DeepMind, tells us that »the study of the ethics, safety and societal impact of AI is going to become one of the most pressing areas of enquiry over the coming year« (Suleyman, 2018). In the USA, the IEEE has released their guide on ethical AI (IEEE, 2019). In the UK, the House of Lords has set out key principles for the ethical design of AI. The Alan Turing Institute has launched a number of projects and, in part- nership with The Nuffield Foundation, has created the Ada Lovelace Institute. Open-Access-Publikation im Sinne der CC-Lizenz BY-NC-ND 4.0