sparring shows the highest ROI up front, you can noticeably improve day over day for months (sometimes years) just sparring with light instruction on technical details
after a while sparring you plateau, now you have to understand why parts of your game are not effective and then incorporate that understanding into your sparring
all arts are martial arts
Indeed, early in oneโs career, taking a course to learn how to build successful products is not a good use of time.
But after 10-20 years, *if* one wants to be world-class at *building successful products*, just shipping more provides zero returns on the learning one now needs.
@mitchellh posthog comes to mind, the non technical people not running a harness all use the chatbox. I almost never use the chatbox. the only time I use it is if I want to share the chat thread to a non technical person
this has been hackily solved for me for ~6 months and its based on the engineer, their experience, my trust level of them, and my understanding of the problem they're working on
there's a "stamp after a quick glance" threshold at some intersection of those things.
I will never trust some (juniors, people who are still learning, non-technical folks, etc.) enough to have zero other human beings review their code, but I think 6 months or a year from now its possible that there are many engineers I would have that trust with.
One thing I am hearing that is top of mind for many eng leaders
"What are we doing to deal with this ongoing increase in code review load?!"
Everyone is seeing it, no one seeing a solution, but lots of experiments from better tooling to starting to admit code review wont scale
"I want to have a podcast with awesome guests nobody knows about but I do, and I value them greatly.. but I don't want to record it"
"you mean you want to have long conversations with your friends?"
"yes"
its not new, this is something @steipete has been doing with agents for over a year "close the loop" is his direct quote from a long ass time ago
and its something that every engineer has been doing themselves to build software since the dawn of time. its just a feedback loop
What is "loop engineering" to you, anyway?
Been looking into this, and also experimenting myself... and I'm not (yet?) buying that it's a new paradigm. But I might just be missing something.
Can you share "loops" you regularly use?