Developing Leadership

A handful of non-obvious observations land - the cascading org-math showing 25+ leaders per 100 engineers, the 'blind-leading-the-blind' mentoring hierarchy problem, and the idea that managers should orient upward rather than act as 'shit umbrellas' - but the episode is padded with rambling Google/Microsoft commentary and ends with generic 'ship things, cut through the noise' advice that contains almost nothing actionable.

“the more junior people mentored by the most junior people in terms of the leadership... the person who read the page before you coming into class trying to teach you”

“we're at what, 25, 26 engineering leaders to 100 engineers. Now, this makes no sense to me”

The org-ratio math walkthrough and the 'two modes of org change' framework (shock vs. slow boil) are reasonably fresh framings, but most of the episode recycles familiar Silicon Valley critique and the Kobe coaching analogy is well-worn; the contrarian claim that Silicon Valley has produced almost no real leaders is provocative but asserted without evidence rather than argued.

“I believe that there is so few actual real leaders in Silicon Valley that what we end up having is these dystopian, weird, baroque Rube Goldberg machines of organizations”

“there's only two possible ways in my opinion to dramatically change organizations... all at once right away like bam... and the other is going back to that frog boil pot analogy”

Both hosts are practitioners with real operational experience - Jason references managing 38 directs at Canonical, senior leadership at GitHub and Heroku - which gives grounding to some claims, but this is a two-host chat format with no external guest, and Jason's credibility is partly asserted through self-referential anecdotes ('I get the call all the time from every large big tech company') that are unverifiable from the transcript.

“at one point I had 38 direct reports and I mean obviously if you say that people are like, that doesn't make any sense”

“when I was joining GitHub I got a lot of pushback um, on a lot of different approaches things. One of them was there was a various person who was very close to the company advise said hey, you need to go and you basically need to cut 50% of all the people”

The specific org-math calculation (20 team leads → 5 senior EMs → directors → VP = ~25-26 leaders per 100 engineers) and the 38-direct-report anecdote are the episode's strongest evidence moments; however, the advice section is almost entirely vague, named company examples (Google, Microsoft, Meta) are used illustratively without data, and most claims about AI eliminating manager tasks are asserted without concrete evidence.

“So I'm at 20 team leads... let's say it's three... we've got an extra five leaders... it's not uncommon to find a director for every two or three ems... we're at what, 25, 26 engineering leaders to 100 engineers”

“at one point I had 38 direct reports”

Aizo asks a few genuinely pointed questions - the manager-as-best-coach challenge and the tactical hours-per-week question are good moves - but the conversation is largely agreeable banter; Jason's longer monologues go unchallenged and the final advice request is explicitly self-described as a cop-out, signalling the host knew a harder follow-up was warranted but didn't pursue it.

“Do you Think that the manager is always the best coach?”

“How many hours a week should a team lead spend with an engineer on their team to help get them to the best place for them for their career and for the company”