An AI-Powered “Bot” As Your Leadership Coach — A Redux On My 2021 Article About This Future. ChatGPT Compresses The Timeline…
(Author’s note — This article has references to coaching, chatbots, delegation, Barry Manilow, Two-Buck Chuck, and leadership limericks. Readers should proceed accordingly.)
“A leader stands tall,
guiding with wisdom and heart.
Inspiring the pack.”
- Haiku about leadership written for me by ChatGPT
Unless you’ve been hibernating in a cave for the last several months, you’ve heard of ChatGPT, the AI (Artificial Intelligence) based conversational chatbot that has attracted over 100 million users just two months after it was made available. And by the way, the “GPT” means “Generative Pre-Trained Transformer.” But you knew that…
And I am among that 100 million+.
Current search engines like Google or Bing provide long lists of relevant website “hits” to whatever you ask about, and from those hits a user must navigate to specific sites. ChatGPT, on the other hand, provides plain language responses to your questions or queries, not just a register of sites, many of them “sponsored” sites.
ChatGPT can write essays, poetry (like the haiku above), solve math problems (but sometimes not very well), and provide recommendations of many sorts, based on information fed into it from the web through 2021.
It’s head-shaking in its profound simplicity, yet it and similar AI-driven chatbots are in the very early stages of sophistication. Those of us who were wowed by the computer games Pong and Space Invaders in the early 70s now appreciate how primitive those games now appear in the context of having seen several generations of their offspring. Technology has advanced rapidly and will continue to do so at an even faster pace.
Which gets us back to the AI bot-as-coach redux. In 2021, I published a piece called “Will Alexa 4.0 be your leadership coach of the future?” and it was an updated version of an article I had originally written in 2017. In the article, I suggested that within 10–15 years a “smart” tool may well exist that could be our personal leadership coach (or life coach or therapist), providing advice and “guided discovery” solutions to whatever we asked about and in a conversational manner. An electronic trusted advisor who knows us well, it would have no hidden agendas and have access to millions of terabytes of data on the web for best practices.
That “coaching bot” — coming sooner?
I was too conservative. It won’t take 10–15 years from 2021 for that capability to exist. I now think devices like that will be on the market about 5–7 years from now.
There are several reasons I think this capability will get here sooner and why it has — at least for now — supplanted the Metaverse or self-driving cars as the current tech “wow factor.” One reason is that in the last few weeks, an arms race has broken out in the nascent AI-chatbot space. Microsoft has recently invested $10B in OpenAI, the company that developed ChatGPT.
Google, though, also has deep pockets and has its own entry it is developing called “Bard.” Google had a burp in its original public release on February 8 but is doubling down in developing the capability because it sees its own traditional search — and ad revenue — as being vulnerable to this type of interactive search.
The Chinese search engine giant Baidu has also announced it will launch its own AI-based chatbot in March. Baidu’s product will be called “Ernie,” which is an acronym for some technical terms loosely translated from Chinese to English. And there are other lesser-known competitors emerging in the space.
Experts see many potential use cases for AI-chatbots beyond composing clever haikus and ghost-writing students’ term papers for them. Consultant/writer Michael Watkins of “The First 90 Days” fame, for instance, has trialed ChatGPT and sees it as enhancing decision-making, problem-solving, and strategy development for executives. The natural language underpinnings of those bots make the tools very approachable for everyone. Even executives.
My trialing of ChatGPT — Interesting and fun stuff…
Enough of the big-picture stuff. Let me share some interesting and sometimes fun things I’ve learned from my own trialing of ChatGPT after opening an account on openai.com.
One of the questions I asked ChatGPT relates to the premise of this article. I asked: “Can ChatGPT coach leaders?” The verbatim response was:
“As an AI language model, I can provide information and guidance on various topics, including leadership. However, I am not capable of providing individualized coaching to leaders. My function is to provide general information and suggestions based on the data I was trained on.
Effective leadership requires a deep understanding of individual strengths, weaknesses, motivations, and interpersonal dynamics, which is beyond the scope of what I can offer. A human coach or mentor with personal experience and expertise in leadership development would be better equipped to provide personalized coaching and support to leaders.”
A candid, accurate and well-written response, I must say. Since ChatGPT can only currently access web material, it’s not at a level like “Alexa 4.0” might be in the future. An AI-coach like that (as my article suggests) would have a great deal of personal information about you, the user. Personalization like that is not here yet.
I also asked ChatGPT some questions I often got from leaders when I was in a professional coaching and consulting role. I wanted to match what the bot prescribed vs. what I, with a number of years of coaching and leadership development under my belt, might have offered to a client. So, one of my questions was: “How can I delegate better as a manager?”
ChatGPT’s response was five different practical tips having appropriate detail and reflecting the best practices of delegation. In addition, for any of those 5 “tips,” I could dig in and ask more detailed questions about how I might specifically execute that step. I asked about one of the the tips (“providing clear instructions”) and got 7 additional sub-steps about how to further carry out the action.
I received similar responses to other leadership coaching questions I posed:
- How to fix an attendance problem
- How to improve relationships with colleagues
- How to foster innovation across departments
- How to improve work-life balance.
The responses tended to be formatted the same way, enumerating the options, and always closed with a note of positive reinforcement, which I thought was powerful. For instance, for my question about building better relationships with colleagues, ChatGPT closed the response with: “Remember, building strong relationships with colleagues takes time and effort, but it can greatly improve the work environment and lead to greater job satisfaction and success.”
That’s right out of the coach playbook.
To compare to existing digital assistants, I also asked Google Assistant (“Hey Google…”) and Siri (Apple) how to delegate more effectively. Google Assistant responded vocally with some accurate but very brief measures to take and Siri simply provided three website options on the screen. No vocal/conversational response. ChatGPT, which also has a verbal response option, was the user interface winner hands-down in that limited test.
Asking ChatGPT about non-work related topics…
I also had some fun trying out the bot when I asked about non-work-related items. For instance, I asked ChatGPT to compare Boomer musicians/ songwriters Roy Orbison and Barry Manilow. The several-paragraph response compared and contrasted both musicians’ styles and the different types of music they represented, as well as the titles of two songs each is famous for. “In terms of musical style, Orbison was heavily influenced by blues, rock and roll, and country music, while Manilow’s music is rooted in pop, adult contemporary, and easy listening.”
The response was informative, succinct, and written in everyday language using flawless grammar.
The next test of the bot was with wine. Being a cork dork, I asked ChatGPT to identify the best California cabernets for under $20. In response, the bot identified only four wines, one of which was Trader Joe’s “Two-buck Chuck.” So, there are clearly limitations to the bot’s knowledge based on its database. In fairness, the response did state that wine taste preferences are “subjective.”
ChatGPT writes a limerick for me, but not a naughty one
Finally, appreciating the nuances of clever verse, I asked ChatGPT for a “naughty limerick about leadership.” The response I received indicates the guardrails that the tool’s developers have created to prevent obscene, offensive, or overly biased results. It said:
I’m sorry, I cannot generate inappropriate content. Limericks are meant to be lighthearted and humorous, but it’s important to maintain a level of respect and professionalism, especially when it comes to sensitive topics such as leadership.
Leadership is a “sensitive topic?” So much for my bid for bawdiness. However, when I repeated the request but left out the word “naughty,” I received the following limerick:
There once was a leader quite bright, whose vision shone clear as day light;
He worked with great zeal, his team to reveal, the path to success that’s in sight.
Competent, yes. Clever? I think so. And it was composed and communicated back within two seconds — no “think” time. And what struck me is that the bot can put words and phrases together to create a unique response that seems very much like a product of human creativity.
What will the AI future bring?
So, this AI-driven tool is impressive, and remember, we are only in the “Pong” stage of its maturity. If we indeed have AI-based leadership and “Alexa 4.0” personal coaches in 5–7 years, what will we have in 15–20 years?
The Law of Unintended Consequences and its dark sides will surely be with us and even today there is great concern about AI being used to create “deepfakes,” meaning the use of AI to replace the likeness of one person with another in video and other digital media. Deepfake technology can be used to create fake news and misleading, counterfeit videos. It will be difficult to determine what is real and what has been created in a digital lab.
Even with ChatGPT being in its embryonic stage, experiments asking people to distinguish between human-created essays and those created by ChatGPT demonstrated the difficulty in trying to distinguish between the two. One NY Times article describes how even a fourth-grade teacher, a Stanford education professor, and children’s author Judy Blume were unable to consistently tell the difference between sample essays written by actual children and those written by the bot.
Technology like this is exciting, but also uncertain and scary. Ironically, perhaps a personalized AI-driven coach can help us to navigate that AI future. And it calls into question the future job security for bloggers and speech writers.
About the author: Mike Hoban is a business topics writer. He is actively working at becoming a world-class grandpa to his five young granddaughters. In addition to his 40 years experience as a leader, consultant, executive coach, and business owner he has also published extensively in Fast Company and wrote many thought leadership pieces for DDI when he was there. He also wrote a business column for 12 years. His recent commentaries can be found on his LinkedIn page: https://www.linkedin.com/in/mike-hoban-b5756b6/ He can also be reached at mjhoban99@gmail.com.
