Weeknotes 517
I did:
Heatwave
The second order effects of things like heatwaves always fascinate me. In-person sessions cancelled, schools closed, more people on leave, people not feeling their best, more stress, maybe different decisions get made than would have in other circumstances. Anyway, this is some of the stuff I did during the heatwave:
- Did a training course in AI product strategy. We used AI as a tool to consider how changes in the AI market affects AI in our products.
- Set up a spreadsheet for doing Bayesian analysis on product outcomes. It needs more work but there’s definitely something useful in being able to express uncertain outcomes as probabilities.
- Chatted about Total Quality Management and where it could be useful for us.
- Created a little agent for answering questions about some work we’re doing.
- Me and three other product managers pitched our ideas to a group of stakeholders and got lots of good feedback.
- Presented my thinking on the future direction for one of our big products.
I read:
How LLMs affect business
Interesting post from the CEO of Buttondown about customers having used LLMs to research their product and what that does to support needs, lifetime value, etc. It’s an important consideration for every organisation. When customers come to you having already established an understanding about your business that you haven’t had much control over, it changes the nature of the relationship. This is AI as an intermediary.
The right it
Stumbled across Alberto Savoia whilst wandering around YouTube so bought his book. Despite calling Howard the Duck a failure, he has some good advice about collecting data to validate new ideas.
Bottleneck theatre
“Imagine the “bottleneck” really was engineering or coding. What does that imply about the competitive game you were playing? Are you saying the only thing holding you back was something relatively commoditized—something you could just hire more people to do? What does that say? It suggests you were competing in a world where speed of execution mattered more than insight, where everyone was building roughly the same thing, and whoever shipped faster won. But if that’s true, you better hope engineering wasn’t your bottleneck. Because if it was, and AI removes that constraint for everyone, your competitors will take you to the cleaners.”
I mean, no one ever really believed that organisations actually only have one bottleneck stopping them from suddenly being more successful, right? In many cases, the way the organisational structure is set up to operate is the actual bottleneck, and AI can’t fix that.
So a lot of this is just AI theatre. And when we talk about theatre we’re talking about the suspension of disbelief, which is to say, everyone knows it’s not real but everyone goes along with pretending it is. AI can’t fix that human trait either.
I thought:
Changing sociotechnical systems
There’s a pacing layers thing going on where technology can be changed quite quickly, changing tasks is slightly slower, structural changes happen more slowly still and people changes can be really slow. So there’s an uneven distribution of benefits of the change.
Spirals, not circles
So many diagrams explaining methods like Theory of Constraints, Agile, Scrum, etc., are circles. I get the simplicity but I think it’s too easy to interpret as repetitively going through the same process again and again but not really getting anywhere. I think spirals might explain the process better. The work flows through the same steps but each time it is more refined and closer to the target in the middle of the spiral.