AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

In an interview with The Pragmatic Engineer, Imprint CTO Will Larson reflects on how engineering management expectations have changed since the 2010s. He argues that middle managers may need to return to line management and rebuild hands-on technical skills as companies reassess leadership roles and AI alters software work.

Will Larson, CTO of Imprint and a longtime engineering executive, says the expectations placed on engineering managers have shifted since the 2010s, with middle managers facing pressure to reskill and, in some cases, return to hands-on technical or line-management work. In an interview published by The Pragmatic Engineer, Larson connected the change to different business conditions and the growing role of AI in software development, raising questions about how engineers advance to executive leadership.

Larson’s central argument is that management paths shaped during the long period of low interest rates and abundant technology investment may no longer fit current company priorities. The interview describes altered expectations for both engineers and managers, though it does not provide industry-wide data measuring the scale of the shift. Larson recommends that some middle managers return to line management and reskill, drawing on his own effort to refresh technical capabilities.

The conversation also examines how AI could change engineering work and bring more technical decision-making back to senior leaders. The interview presents this as a developing change in roles, not as evidence that AI has already eliminated management jobs. Larson and the interviewer discuss whether management is becoming a more precarious career path and whether some people may need to reconsider their plans for advancement or retirement.

Larson draws on experience at companies including Yahoo, Digg, Stripe, Uber, Carta and Calm. He recounts Digg’s troubled v4 rewrite, which launched amid declining traffic and business pressures. According to the account, the redesign alienated users and had performance and reliability problems; Larson says his team traced technical issues to Python memory leaks. He does not attribute Digg’s eventual decline to the rewrite alone, saying competition from Facebook and Reddit would have threatened the company regardless.

At a glance
reportWhen: Published in The Pragmatic Engineer; th…
The developmentThe Pragmatic Engineer published an interview with Will Larson about changing expectations for engineering managers and the effects of economic and technological shifts on leadership careers.

Management Roles Face New Demands

The discussion matters to engineering managers and the companies that depend on them because it challenges the assumption that management is a one-way move away from technical work. If employers increasingly expect leaders to make or evaluate technical decisions, managers who have been removed from day-to-day engineering may face pressure to rebuild those skills. That could affect hiring, promotion and how organizations define the value of middle management.

Larson’s comments are an experienced executive’s perspective, not a forecast proven across the technology sector. Still, his account points to a practical career question: whether managers can adapt their responsibilities as businesses change, rather than rely on a stable ladder from engineer to manager to executive. AI adds uncertainty to that question, since its effects on team size, technical work and leadership responsibilities are still unfolding.

Amazon

engineering management books

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

From 2010s Norms to Today

The interview contrasts current expectations with Larson’s recollections of earlier workplaces. At Yahoo, where he worked from 2008 to 2010, he says he had only two one-to-one meetings with his manager. He also describes a server-approval policy associated with cofounder David Filo that, according to Larson, encouraged informal exchanges of unused equipment between teams. These anecdotes illustrate how management practices and incentives can differ across companies and eras; they do not establish how common those practices were across the industry.

Larson also recalls leading an organization of about 250 people at Stripe, where some senior engineers sought more engaging assignments. He says he sometimes told them, “I only have suffering to offer, but this suffering is what’s going to make the business succeed.” The comment reflects a tension between individual interest and business priorities that remains relevant to leadership: senior technical staff may want stimulating work, while managers must direct effort toward problems the organization needs solved.

“I only have suffering to offer, but this suffering is what’s going to make the business succeed.”

— Will Larson, as quoted in The Pragmatic Engineer interview

Amazon

technical leadership training courses

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

How Broad Is the Shift?

The interview does not establish how many companies are reducing middle-management roles, requiring managers to return to technical work, or changing promotion criteria. It also offers no quantified evidence about AI’s current effect on engineering headcount or the management career ladder. Larson’s views are informed by his own experience, but the source does not present them as a measured industry consensus.

It remains unclear which managers would benefit from returning to line management, what technical skills employers will expect them to regain, and whether those expectations will persist. The source also does not specify how quickly AI will alter teams or how those changes will differ between companies.

Amazon

AI in software development tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

How Teams Adapt Their Roles

The next test will be how engineering organizations translate these arguments into hiring, promotion and team-design decisions. Managers and technical leaders will be watching whether companies ask leaders to contribute more directly to technical choices, and whether AI changes how engineering work is assigned. The interview does not announce a new industry policy or a scheduled follow-up milestone.

Larson is CTO of Imprint, and the source says he is hiring for several engineering manager roles there. It describes those positions as part of an effort to use agents to rethink how software teams work, but provides no further detail on the roles or the results expected. For now, the extent of any broader change in management careers remains an open question.

Amazon

line management for engineers

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What did Will Larson say has changed in engineering management?

Larson said expectations have shifted since the 2010s and that some middle managers may need to return to line management and rebuild hands-on technical skills. The interview offers his perspective rather than a sector-wide measurement.

Does Larson say AI has eliminated engineering management jobs?

No. The interview discusses AI as a force that may change engineering roles and bring technical decision-making closer to leadership. It does not report that AI has eliminated management jobs.

What happened during Digg’s v4 rewrite?

Larson recounts that Digg’s v4 relaunch alienated users and faced performance and reliability problems. He says his team traced technical issues to Python memory leaks, while also arguing that competition from Facebook and Reddit posed a broader threat to Digg.

Is middle management becoming a less secure career path?

The interview raises that question and presents Larson’s concerns about changing expectations. It does not provide data showing how widespread or permanent any increased insecurity may be.

Source: rss

HALLOWEEN

Halloween Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

Hugging Face And TorchTPU: Bringing AI Closer To TPU Hardware

TorchTPU brings TPU device support to Hugging Face’s Transformers, Diffusers and Kernels workflows, with performance claims that remain vendor-reported.

Building Resilient Systems With Sam Newman

A Pragmatic Engineer podcast episode features Sam Newman on resilient distributed systems, microservice trade-offs and AI-assisted development.

How The RoR Creator Reignited Debate Over Coding By Hand’s Future

At Rails World, David Heinemeier Hansson said 37signals would rely on agents for nearly all its code, renewing debate over hand-written software.

Java Sealed Classes And Exhaustive Pattern Matching

A Java developer’s report explains how sealed types enable exhaustive pattern matching—and why adding a subtype can break API consumers.