I Did Something Bad
I’ve been thinking quite a bit lately about decision making and making mistakes.
Like everyone else on the planet, I’ve been playing with Chat GPT a bit lately. Something interesting about Chat GPT is that it is often wrong.
ChatGPT is a language model, and is generally bad at math it’s not designed to provide answers to specific questions. Fortunately, that means that skynet is a still a little ways off. For now…
I Don’t Know
What I find annoying about ChatGPT’s wrong answers is that it’s confidently wrong. It doesn’t use weasel-phrases like I think the answer is x, or My best guess is y - it states correct and incorrect answers with the exact same bravado. Chat GPT knows that math isn’t its strong suit, but it won’t admit it. That’s ok. It’s a computer.
Unfortunately, some human leaders are also happy to be confidently incorrect (look no further than American politics for canonical examples). But even in the software world, I’ve worked with leaders who would rather give a confident incorrect answer - or explain how something works incorrectly than say three words that leaders should say more often.
I. Don’t. Know.
It’s ok to be wrong. It’s ok not to know. It’s ok to fuck up. Yet so many people want to get everything right the first time. I think Weinberg’s Secrets of Consulting books are as much about leadership as they are about consulting - especially where he says:
If you can’t accept failure, you’ll never succeed as a consultant.
Which I think is fair to adapt to all knowledge work. Making mistakes and recognizing what you don’t know are opportunities to learn.
The only real mistake is the one from which we learn nothing. - Henry Ford
I am a big fan of everything Patrick Lencioni has ever written. In Getting Naked, Lencioni focuses on the idea of "vulnerability-based trust" in professional relationships. He talks about how transparency and authenticity are necessary to build trust (which he talks about extensively in Five Dysfunctions of a Team). Lencioni and I agree that the most successful leaders are those who are willing to be vulnerable and open with their teams.
Some people just like to be right. It’s difficult for them to be vulnerable, and they think that as leaders, that they need to know all of the answers. Kathryn Schulz has a wonderful talk on being wrong and why humans get stuck in this bubble of needing to be right.
Maximize Learning - Not Failure
The optimal way to maximize learning is to make a whole lot of small mistakes - reflecting and learning after all of them. I’d much rather cause a minor latency issue in a production service than mistakenly delete a production database.
Another reason for fast feedback loops (which I still don’t think a lot of people fully grok) is that they give us an opportunity to make those small mistakes in our software process rather than ship an entire feature that does nothing to solve a customer problem.
Inevitably, we’ll make bigger mistakes. Things that may be embarrassing and make us question whether we’re even in the right career. I have certainly done my fair share of (in hindsight) dumbass moves in my career. These were all followed by guilt, some embarrassment, and then a whole lot of reflecting and learning. One thing I’ve learned from my bigger fuck ups is that they’re rarely caused by a single mistake - they come from me doing multiple things wrong at once, leading to a much more significant error. While I learned a lot from these situations, it sure feels a lot better to learn from mistakes with less impact.
Decisions & Practice
A leader at my company recently sent a note with his thoughts on decision making. It was a bit rambly, but good. It reminded me of a colleague at Microsoft who was one of the most phenomenal decision makers I ever met. No matter how much - or little information they had, they could make a good decision. Of course, they were also a master at listening and always knew what questions to ask to lead them towards a better decision - and they weren’t afraid to say, “I don’t know”. There’s an art to good decision making that goes beyond the scope of this post, but it’s important that we apply learning from mistakes to decisions as well.
I’ve learned that getting good at making decisions requires the same principle as getting good at a musical instrument. Practice a lot. If you’re a leader, practice making decisions by making a lot of them. Make them in different contexts, with different levels of discussion and debate with teammates and colleagues. Make as many decisions as you can that are recoverable. Make mistakes with decisions and learn from those mistakes. Then, when you have to make bigger decisions with bigger impact, you will have to tools to get it right. And even if/when you don’t get that big decision correct, you have also practiced learning, so you can get back on your feet and try it again.
Chat GPT will go on being confidently incorrect. You, however, are a human and can admit when you don’t know something. Too many leaders think that leadership is telling people what to do, but it’s not. It’s about learning what the right thing to do is - and essentially, that’s good decision making. So go out there and screw shit up, and then use those screw ups to maximize how much you can learn and grow.
Related to leadership, I really like the "salary:decision" ratio approach. The bigger decisions people with lower salary can do, the better shape the company is. This is as I see leadership is, empowering people making the decisions which are in their influence area to make.
Regards to Brent, really like the podcast!
Hi Alan (and Brent)!
I would like to comment on this:
I’ve learned that getting good at making decisions requires the same principle as getting good at a musical instrument. Practice a lot
I would add: We should learn about decision making from books. For example I recently read
Decisive: How to make better choices in life and work
by Chip Heath, Dan Heath
Practicing what we read in books is a great combination.