If Elena Verna's piece arguing that IC work is the new career flex landed in your inbox last week and left you with some version of "wait, did I make the wrong call," hold that thought. Two things crossed my feed last week that I keep thinking about together, and I want you to read them side by side before you decide what to make of either.

The first is Verna's piece. She recently stepped back from leading a growth team at Lovable to go back to IC work, and she's making the case for what she calls the High-Impact Individual Contributor: an ex-leader who can now do the work of a whole team solo, powered by AI. You can read her piece here: IC work is the new career flex.

The second is Julie Zhuo's post on X. She surveyed dozens of senior Bay Area tech managers and found that every single one of them is expected to do hands-on AI work alongside their management responsibilities, with one category to avoid: critical-path product work, because you'll end up doing a poor job at managing or a poor job at delivering.

Both pieces got real traction. Both are saying something true. But neither one says what I think they're actually pointing at together.

The gap in Verna's argument

Verna's piece is compelling, and I want to be fair to the argument before I push back. The HI-IC is a real phenomenon. AI does enable a single person with the right background to operate at a scale that previously required a full team. The economics are changing fast enough that some of the old headcount logic really is becoming obsolete.

But there's something her piece doesn't address: what made any of it possible.

She describes what she can do at Lovable: running campaigns, managing partnerships, coding features, shipping product updates on her own. Things that used to require a PM, a designer, and an engineering team. What she isn't accounting for is the operating system she brought back with her. The prioritization instincts. The ability to understand what actually moves a metric versus what just feels productive. The judgment about which projects are worth pursuing and which ones aren't. The capacity to see the whole board rather than just the next task.

Those are skills she built by managing teams, setting direction, and learning to lead across organizations. The HI-IC she's describing is the product of having spent years in leadership. Verna went back to IC work with a fundamentally different operating system than she had before she started managing, and the difference between her and a traditional IC at the same job title is entirely downstream of that time spent leading people.

What Zhuo gets right — and where it stops short

Staying technically engaged as a manager is not optional anymore. Managers who can't meaningfully engage with how their teams are using AI, who can't pick up a tool and understand what it can and can't do, will lose credibility fast. The expectation is shifting and shifting fast.

The four categories Zhuo names: internal efficiency tools, quality-of-life fixes, vision artifacts, celebration stories. They're a useful guardrail. But they describe one type of appropriate manager hands-on work, not the complete picture. All four are fundamentally about enabling the team. Which is real and important. And also too narrow.

As an analytics manager, I've been building what I'm calling an Insight Studio for my team: an AI system trained on past analyses, feedback loops, and output templates that lets my analysts do the refinement work with the model before anything reaches me for final review. The goal isn't to do the analytical work myself. It's to reduce the rounds of feedback we burn on data QC / validation, narrative of insights, and structure of recommendations so the team can support more analyses per quarter without me becoming the bottleneck. That's Zhuo's category, and she's right about it.

But I also continue doing something else. Right now we're building a customer journey analysis for our small business segment and I’m leading the framework: what the journey looks like, which high-value actions to analyze, what the model needs to answer, and what metrics to use. But that required going deep: analyzing each product in the portfolio, its features, its lifecycle, understanding the data well enough to define which signals actually matter. That's hands-on technical work, not just strategic framing. My analysts then take that foundation and do the data discovery, the modeling, the actual build. By the time their predictions are running, I'm close enough to the data that I'm co-owning the output, not reviewing something I've been detached from. The line is intentional. I own the foundation and the strategy. They own the execution and the depth. That's what it looks like to be in the weeds as a manager without becoming the IC.

That's not in Zhuo's four buckets. And it matters, because there's a growing narrative that managers don't produce, that “they just coach”. That's wrong. The best technical managers I know own and deliver strategic work directly: work that requires the organizational context, the cross-functional relationships, and the accumulated judgment their ICs don't yet have. The real distinction isn't team-enabling work versus everything else. It's whether you're doing the work that belongs to you by virtue of your role and context, or doing the work that belongs to your team because it's more comfortable than developing them through it.

What both pieces are actually pointing at

The traditional career path was already under pressure before AI: IC to manager, moving progressively further from the hands-on work with every promotion, on the assumption that leading and doing were fundamentally in tension. AI is accelerating the end of that model. Not because management is becoming obsolete, but because the expectation of what a technically engaged manager looks like is shifting faster than most org charts can keep up with. What's replacing the old ladder isn't a cleaner version of it. It's something more fluid: professionals who can move between IC and management modes depending on what the opportunity requires, not because they never chose, but because they built both well enough to move deliberately between them. Elena Verna is showing you what that looks like when you've actually done the full work.

— Andrea

P.S. Not sure where you actually are in the transition? I built a free 5-minute assessment that shows you exactly which IC habits are still running the show — across how you execute, communicate, and delegate ownership. Nine questions. Immediate results. You can find it here.

If someone forwarded this to you and you want to subscribe, you can do that here.