This really resonates, James. I think you captured the stark shift from the growth-at-any-cost days to a world where efficiency and impact are everything.
This presents an interesting challenge. While some skepticism about AI tools is healthy, it's crucial that team members actively engage with them and contribute to discussions about their benefits and drawbacks. However, it becomes a concern when some team members resist trying these tools altogether, which often suggests a resistance to growth.
> it is evident that productivity is significantly higher when it is used
"Significally" is not evident enough (depending on what percent is counted as significant). 10-15% ballpark of productivity increase is more or less backed by recent studies, with AI being on par with adding good IDE or other important changes. However, a typical troubled team can lose 80-90% of productivity on other types of issues, so this increase can be relevant when team is up and running in a healthy way.
Therefore, good people management still comes first: to use AI as rocket fuel, you need to build a sturdy rocket first
> it is evident that productivity is significantly higher when it is used
And the source is? Key emphasis on "SIGNIFICANTLY". I would agree if you said that the suits want to see AI being used, but I would expect better from who I assume to be someone with a technical background.
I would say that this sounds like terrible advice for managing engineers. From an engineer point of view. Any engineer hates being micromanaged, much more so being micromanaged on their workflow.
Track metrics on how much they use AI? Jesus. Not only you have to worry about all the stupid metrics created just so managers can look good, you also have to care about querying ChatGPT X times everyday and being micromanaged on it.
Lets also treat AI Zealot Managers as underperformers, because that's what you are. You're hurting your team's morale, driving talent away, just because you think you know better than your own specialists.
This really resonates, James. I think you captured the stark shift from the growth-at-any-cost days to a world where efficiency and impact are everything.
Thank you! I appreciate the feedback. Things are indeed different. But there is always opportunity.
> Likely treat AI skeptics as underperformers.
Going to lose some good engineers with this take.
This presents an interesting challenge. While some skepticism about AI tools is healthy, it's crucial that team members actively engage with them and contribute to discussions about their benefits and drawbacks. However, it becomes a concern when some team members resist trying these tools altogether, which often suggests a resistance to growth.
> it is evident that productivity is significantly higher when it is used
"Significally" is not evident enough (depending on what percent is counted as significant). 10-15% ballpark of productivity increase is more or less backed by recent studies, with AI being on par with adding good IDE or other important changes. However, a typical troubled team can lose 80-90% of productivity on other types of issues, so this increase can be relevant when team is up and running in a healthy way.
Therefore, good people management still comes first: to use AI as rocket fuel, you need to build a sturdy rocket first
> it is evident that productivity is significantly higher when it is used
And the source is? Key emphasis on "SIGNIFICANTLY". I would agree if you said that the suits want to see AI being used, but I would expect better from who I assume to be someone with a technical background.
I would say that this sounds like terrible advice for managing engineers. From an engineer point of view. Any engineer hates being micromanaged, much more so being micromanaged on their workflow.
Track metrics on how much they use AI? Jesus. Not only you have to worry about all the stupid metrics created just so managers can look good, you also have to care about querying ChatGPT X times everyday and being micromanaged on it.
Lets also treat AI Zealot Managers as underperformers, because that's what you are. You're hurting your team's morale, driving talent away, just because you think you know better than your own specialists.
I’m seeing the same thing from my conversations, experience and research - thanks for sharing James.
New EMs look like:
- Moving more hands on
- Optimizing output and impact, not headcount
- Being a leader on ai experiments and team coaching
- Helping your ICs grow in impact and faster shipping and iteration
- Optimizing processes to speed up the entire project planning lifecycle to help engineers do their best work
Congrats for another awesome article James.
With all this flattening happening at the mgmt level, how are EMs expected to be more hands-on if their direct reports count increases?