I got into management because I saw myself as a player-coach. I could create software and make my teams more effective.

The most rewarding part? Shielding creative people from bad direction and then watching them delivering results. Success gave me more influence to shape process and continue to help my teams be their best.

Chris Dixon’s recent, high-profile shift from “Management” back to “Engineer” made me pause. He’s a true champion of the possible who had an outsized impact on my early career, and his move forced me to think about my own direction.

And it raises an interesting question: What happens to middle managers when AI changes everything?

The Shield and the Amplifier

I’ve spent my career playing two roles that seemed contradictory but were actually complementary.

The Shield: Protecting engineers from terrible ideas and organizational chaos. When the C-suite wanted to rewrite the platform in the hot new framework they’d just heard about, I absorbed that directive and translated it into something that wouldn’t destroy six months of progress for a passing fad. They want it right now, when the right answer may be next week or never.

The Amplifier: Carrying technical truth upward when executives didn’t want to hear it. Engineers would tell me about architectural constraints, technical debt, or operational issues that leadership needed to understand. When executives pushed back, I could handle the blowback on my team’s behalf. I actually learned to enjoy and seek this conflict.

This worked because I stay technical enough to understand the engineering reality while speaking the language of business outcomes. When I work for stronger organizations, I spend more time directly creating value. In weaker organizations going through transitions, I had to focus on process.

It’s not that the work was different. It’s that organizational conditions demanded different things from me.

What AI Actually Changes

Here’s the optimistic realization: AI isn’t going to put junior engineers out of work. Instead, it’s going to demand that leaders who rely on the status quo step up their game and provide truly meaningful direction and deliver outcomes.

Think about all those times engineers gave me feedback about context, direction, or operational issues that I had to deliver to executives. All those conversations where I translated technical constraints into business language. All those moments where I buffered the team from poorly thought-out mandates.

AI makes that translation less necessary. When individual contributors are amplified by AI, they can articulate problems and solutions far more completely. The gap between “knowing what to build” and “building it” narrows. Engineers will be able to communicate directly with decision-makers because the need for a translator—a middle manager—will shrink. What AI can’t deliver is to ensure that this feedback is actually listened to, and not just heard. Having people like Chris on the team will ensure that this feedback will be listened to.

Automations to facilliate direct communication between customers and creators will eliminate the middleman just as interenet marketplaces allow nearly any product to be sold directly to customers.

This isn’t a threat to middle managers. It is a threat to executives who’ve been outsourcing their understanding. There’s simply no need for a middle manager to educate them when our team can directly create value and communicate with customers, investors, and the CFO.

Before AI, my workload was simply too broad to find the time for deep focus that is necessary to create software. AI lets me park my context when I need to task shift and shortens the time nececssary for me to get back into a productive flow state.

The Evolution, Not Extinction

Moving up to the exec level meant I needed to delegate more of the personally rewarding work of creating and focus entirely on budgets and processes. I did it because that’s what “career progression” was supposed to look like. But I always knew the best parts of my job were when I could still contribute technically while multiplying team effectiveness.

AI brings that back.

Instead of spending time on:

  • Fighting to get engineering concerns heard
  • Translating technical decisions
  • Absorbing bad direction
  • Attending hours of status meetings

We can focus on:

  • Being force multipliers
  • Building psychological safety
  • Solving complex problems
  • Creating value

The middle managers who spent their careers doing digital transformations and helping leadership understand software development cycles won’t become obsolete. We have exactly the right understanding to turn strategy into execution, and now we are freed from the translation work that was always a means to this end.

What This Demands from Leadership

Here’s my take on current leadership: the future means we can simply go around leaders who don’t step up.

When executives could rely on middle managers to translate and buffer, they could remain technically distant and aloof from accountability. AI removes that luxury. Leadership will need to:

  • Develop genuine understanding of technical work (AI makes this more accessible than ever)
  • Provide meaningful strategic direction
  • Create environments where technical expertise is valued at all levels
  • Understand the cost and value of what they’re asking for

The organizations that win won’t eliminate middle managers. They’ll empower us to do what we’ve always done best: multiply the effectiveness of talented people while contributing directly to outcomes.

Organizations that focused so hard on survival mode during transitions—losing their long-term focus—discover too late that competitors who kept investing in their people are now unstoppable. I’ve been in both kinds of companies. The difference isn’t resources. It’s leadership that understands what they’re building.

Back to What Matters

Freeing middle managers isn’t about going backward or some mythical horizontal org. It’s about returning to what made the role valuable in the first place—before it got buried under layers of organizational dysfunction.

We became managers to multiply impact, not to leave technical work behind. We took on the translation burden because someone had to bridge the gap between engineering reality and business decision-making.

AI doesn’t replace that value. It redistributes it. The translation work becomes less necessary. The force multiplication becomes more possible.

So yes, I’m optimistic about my future as an engineering leader. Because the future looks a lot like what drew me to management in the first place: staying close to the craft while helping talented people do their best work.

Skilled middle managers won’t be replaced, how we deliver value will evolve.

Inspired by Chris Dixon’s career evolution and my earlier musings on my role in empowering the next generation I get older.

Cover image: Original public domain image from National Library of Medicine

#Leadership #Engineering #AI #Management