3 ms·

Yeah, I'd argue the issue is largely one of awareness. There simply aren't good established mechanisms for even competent managers to notice the people prevent

Yeah, I'd argue the issue is largely one of awareness. There simply aren't good established mechanisms for even competent managers to notice the people preventing a problem. The manager only has so much mental bandwidth, and part of being a good subordinate is not consuming too much of that. There's an ideal of "the team runs itself" with the manager only providing overall direction/resources/big picture stuff. While this can be great for team execution, by definition it insulates managers from the nitty-gritty of design/development, which is where most of the preventative measures happen. However if you try to remedy this and bring the manager into the fold, you invite micromanaging and harm execution, as well as preventing the manager from doing more managerial things. It's when things go wrong and heroes emerge that people get raises and promotions, if simply because the manager can point to those highly visible efforts as rationale for why this person gets more. Also in large orgs a manager might not know all their direct reports personally so those incidents make the person stick out, like how a professor knows the kids who sit at the front of the class the best. I suppose the ideal would be a technically experienced manager who understands how to promote good design practices and lets his team do so, but on sheer educated faith with no metrics. Given that metrics are the lifeblood of large-org management careers, the incentive structure isn't there to produce such people en-masse.

xiphias2 7y ago
I'm not sure if micromanagement is that bad for the final product, as long as the manager has deep technical expertise and respects the people who he micromanages. Steve Jobs and Elon Musk are famous/notorius micromanagers (though they have some problems with the ,,respect'' part). At the same time as they understand the problems with the product development deeper, they are able to plan more iterations ahead. One example is the autopilot hardware. About 10-12 years ago I was working at Google, sitting in a self driving car in Mountain View when it was a secret project. I was already at that time amazed that the car could do some offroading tricks that I wasn't capable of, and I was quite certain that within 5 years it would be self driving without safety driver. Now it's 2020, and it seems to me that in no more than 10 years Tesla will win the autonomous race, as Elon understood that the computer needs lots of deep learning instructions (more than what NVIDIA could provide) and HD maps and lidars are not just unnecesary, but they are the wrong direction, as the visual sensing problem needs to be solved for safe self driving. In Google there are so many product managers at this point, that they themselves are actively hindering the product development itself, as they don't want to get into the details of the real problems.

xiphias2 7y ago

I'm not sure if micromanagement is that bad for the final product, as long as the manager has deep technical expertise and respects the people who he micromanages. Steve Jobs and Elon Musk are famous/notorius micromanagers (though they have some problems with the ,,respect'' part). At the same time as they understand the problems with the product development deeper, they are able to plan more iterations ahead. One example is the autopilot hardware. About 10-12 years ago I was working at Google, sitting in a self driving car in Mountain View when it was a secret project. I was already at that time amazed that the car could do some offroading tricks that I wasn't capable of, and I was quite certain that within 5 years it would be self driving without safety driver. Now it's 2020, and it seems to me that in no more than 10 years Tesla will win the autonomous race, as Elon understood that the computer needs lots of deep learning instructions (more than what NVIDIA could provide) and HD maps and lidars are not just unnecesary, but they are the wrong direction, as the visual sensing problem needs to be solved for safe self driving. In Google there are so many product managers at this point, that they themselves are actively hindering the product development itself, as they don't want to get into the details of the real problems.