AI is a Dangerous Shortcut

Relying on AI for your day-to-day tasks (that go beyond simple automation) is diametral to your personal goals, and likely diametral to the quality of our social interactions.

I used ChatGPT to come up with an agenda for a Retrospective, and it did a semi-good job. I could have run the session as it was designed, but then the outcomes would not have been as impactful, and even further, it would have robbed me of an opportunity to grow (even if just a tiny bit) as a facilitator. I would just be a passenger.

I have no idea why ChatGPT selected this agenda, what its reasoning was. Of course, it doesn't have any reasoning, it just selects something and hands it to me.

I have designed and run countless workshops. I have built experience. And this experience tells me that the ChatGPT Retrospective was not that good. Without it, I would have been forced to take it at face value and use it, without modifying it (or - in this case - throwing it out all together and designing something new).

What makes us grow our skills is repetition - doing things over and over again, always trying to improve in every iteration. Striving for mastery. Putting in the hours, doing the work, understanding what we are doing and why we are doing it.

AI can be helpful in giving you pointers, or broadening your understanding of the problem, but it can never be the work itself.

I am afraid of a society where all the microdecisions are made by AI, because excellence comes from doing. And most of the things we are doing are on a microscale.

Letting AI write our letters, edit our emails and plan our meetings will not enrich our society. It will dumb us down and make us depend on a machine for even the smallest creative things.

And without the training and experience, we will fail on the big ones.

Published 2024~06~22

Link Graph

Yeah, I know, the 2000s knocked and wanted to show you their ideas about knowledge navigation, but I really like those graphs, even if they are not the most practical instruments, plus I actually developed a network-based knowledge management system called 'Serendipity' back in the day, so please stop making fun of me.