I do notice in myself that the more I create rules for a GPT, the more I am actually micromanaging, and how impatient I get when I don't get the result I want. It's like I just threw away all my management skills, and I am playing a video game. There is a real potential that people management might suffer at least at first here. Positive note, might open up a new world of coaching --> managing those who manage AI :)
My experience with LLMs (like ChatGPT) has been shaped by the level of expectation, so many promises and quite "amazing" features that make you believe the model truly knows its stuff. The frustration initially came when I realised I’d spend so much time trying to align the model and correct its outputs.
But as I studied LLMs more deeply and understood what they really are, I began to see them like a SQL database except this time, I interact with it using natural language. This realisation helped me understand that LLMs aren’t a magical pill but rather a powerful tool for prediction, analysis, and rapid experimentation.
Now, I spend most of my time asking, "Why this? Why that?" not expecting the model to know the answers, but to help me explore what the data is saying and its many possible interpretations. AGI, super intelligence, we are so far away from it.
I do teach and help managers and leaders to understand how to best use LLMs, and still remain in control/human. And I understand the frustration some have while using chat bots.
I really value this perspective especially the line about machines organising information but not meaning.
I explored a similar idea in Post Utility Man, looking at how optimisation culture quietly devalues the human instincts that once defined leadership and worth.
Leaving it here only in case it adds to the discussion:
Thank you! Yes, we certainly need to keep humans in the driving seat. As usual, excessive reliance on innovation/automation will only make more money to the ultra wealthy and create greater inequality for the rest. We need the right balance, or this will all go very badly….!
Couldn't agree more, it's so important to remember that while AI helps with so many tasks, the essence of leadership remains profoundly human, especially when it comes to guiding and fostering genuin growth in people, not just optimizing sytems.
Great piece!
I do notice in myself that the more I create rules for a GPT, the more I am actually micromanaging, and how impatient I get when I don't get the result I want. It's like I just threw away all my management skills, and I am playing a video game. There is a real potential that people management might suffer at least at first here. Positive note, might open up a new world of coaching --> managing those who manage AI :)
My experience with LLMs (like ChatGPT) has been shaped by the level of expectation, so many promises and quite "amazing" features that make you believe the model truly knows its stuff. The frustration initially came when I realised I’d spend so much time trying to align the model and correct its outputs.
But as I studied LLMs more deeply and understood what they really are, I began to see them like a SQL database except this time, I interact with it using natural language. This realisation helped me understand that LLMs aren’t a magical pill but rather a powerful tool for prediction, analysis, and rapid experimentation.
Now, I spend most of my time asking, "Why this? Why that?" not expecting the model to know the answers, but to help me explore what the data is saying and its many possible interpretations. AGI, super intelligence, we are so far away from it.
Agreed, same here! Maybe we do need a whole new category of coaching as you say!
I do teach and help managers and leaders to understand how to best use LLMs, and still remain in control/human. And I understand the frustration some have while using chat bots.
Great to see this collaboration! Living the leadership!
Thank you Kathleen, it was great working with Andrea and I hope we have many more collaborative work.
Thank you!
Enjoy the read! Thank you Andrea and Diamantino.
I really value this perspective especially the line about machines organising information but not meaning.
I explored a similar idea in Post Utility Man, looking at how optimisation culture quietly devalues the human instincts that once defined leadership and worth.
Leaving it here only in case it adds to the discussion:
entropicart.substack.com
Thank you! Yes, we certainly need to keep humans in the driving seat. As usual, excessive reliance on innovation/automation will only make more money to the ultra wealthy and create greater inequality for the rest. We need the right balance, or this will all go very badly….!
Couldn't agree more, it's so important to remember that while AI helps with so many tasks, the essence of leadership remains profoundly human, especially when it comes to guiding and fostering genuin growth in people, not just optimizing sytems.
Definitely. We can’t risk getting to a point where people can’t do anything anymore and we rely solely on computers 😅