⏱️ Slide Timing: 1 min - Summary of Nishant Virmani's CIO.com article (26 Oct 2026); title, argument and practices are his - Topic: what AI assistants quietly take away from engineering teams - Usual worry: juniors lose mentoring - This briefing flips it: seniors, and the organisation, may lose more

⏱️ Slide Timing: 2 min - The fear figure matters most: the driver is social, not convenience - teams that punish "dumb questions" accelerate this - Small sample (30 observed, 131 surveyed), so read it as a signal, not a census - Nobody decided to stop asking seniors; it happened one private chat at a time ❓ Ask: "When did someone last ask you why your system works the way it does?"

⏱️ Slide Timing: 2 min - The organisation pays seniors to hold the architecture together - so their loss is the costlier one - From the inside, fewer interruptions feel like a pure gain - The rest of the briefing unpacks the right-hand column

⏱️ Slide Timing: 3 min - Job 1: the "why" behind a design almost never reaches the wiki - nobody writes a design doc after the fact unless it breaks again - Job 2: after three years owning a service, its rough edges feel obvious - Job 3 is the overlooked one: teachers know you learn by explaining - seniors used to get this workout dozens of times a quarter, for free

⏱️ Slide Timing: 3 min - Recognition: "that line looks wrong" when it's on screen - Recall: "if this times out, what happens three services away?" from memory at 2 a.m. - A junior asking "what if this times out?" used to force seniors to trace that path weekly - Without the rehearsal, the first live trace happens during the outage

⏱️ Slide Timing: 2 min - Aimed at engineering leaders: your green dashboard can be a lagging liability - Teams can coast on accumulated knowledge for months before it shows - By the time it shows in a postmortem, the habit has been gone for quarters

⏱️ Slide Timing: 4 min - DORA shows a correlation, and says its data can't explain why - lost senior context is the author's explanation, not DORA's finding - Stability covers failed changes and how long recovery takes - Anthropic trial: 52 engineers on an unfamiliar library, so it's about learners, not seniors - The junior still needs the senior, just later and more expensively - from upstream teaching to downstream code forensics

⏱️ Slide Timing: 2 min - The problem isn't missing information, it's missing practice - Incentives matter: if directors are rewarded only on velocity, coaching loses to the sprint board - So the fix has to be built into work that already happens

⏱️ Slide Timing: 4 min - Code review: junior explains what the code assumes; senior explains why the boundary exists - Whiteboard: where the explanation stalls is where memory has already decayed - Postmortem: unscripted explanation under peer questions is the real test of understanding - 90 days: high early velocity but every structural change still routes to one veteran - AI is speeding up surface work while the core stays in one head

⏱️ Slide Timing: 3 min - The argument is plausible, but its statistics support nearby claims, not the central one - Mechanism rests on learning-by-teaching research (the "protégé effect"), mostly from students - If junior hiring keeps falling, the questions disappear whatever tools juniors use - Practical upshot: teach juniors to ask AI "explain why", not just "write this" ❓ Ask: "Which of these counterpoints applies most to your team?"

⏱️ Slide Timing: 2 min - Keep AI assistants; put the human friction back deliberately - One action for this week: pick a critical workflow and explain it on a whiteboard to a colleague - Notice where you hesitate; that's where to start

⏱️ Slide Timing: 1 min - Start with Nishant Virmani's article: this deck summarises it, and it links to the original sources - DORA report is the one to bring to leadership: it ties AI adoption to stability numbers - Anthropic study is worth reading if you mentor juniors who rely on AI

⏱️ Slide Timing: 1 min - These back the Counterpoints slide; none are cited in the original article - DORA 2025 updates the 2024 finding the article relies on - Canaries paper is the strongest data on the shrinking junior pipeline