The Work That Built Technical Careers Is Starting to Disappear
Published: May 2026 | OpenNova
When Block cut 4,000 jobs in February, CEO Jack Dorsey was specific about why. Customer support. Compliance monitoring. Junior code generation. AI handles those reliably enough, he said, that keeping humans in those roles didn’t make sense.
That’s a reasonable operational call.
What it misses is what those tasks were doing on the side.
The repetitive work is disappearing first
Junior analysts who used to build reports manually aren’t doing that anymore — the reports generate themselves. Support engineers who spent months triaging tickets before touching production code are being handed production code on day one. Developers who used to write scaffolding from scratch are watching copilots do it in seconds.
This isn’t a coincidence across different roles. It’s the same pattern. The routine, repetitive, low-stakes work is the first layer AI handles reliably. That’s exactly what organizations want. And it’s exactly the layer where technical professionals used to learn how to work.
The repetition wasn’t inefficiency.
It was curriculum.
What actually disappears when the repetition goes
A junior developer reviewing someone else’s code isn’t just reviewing code. They’re building pattern recognition — learning what done looks like inside a specific codebase, with a specific team, against specific standards. A compliance analyst processing routine cases isn’t just closing tickets. They’re developing judgment about which edges get escalated and which get handled.
When you automate that layer, you get the output.
You don’t get the development.
Block, Cloudflare, Meta — the major restructurings of early 2026 concentrated in the same categories: customer response routing, data processing, administrative coordination, junior code work. Across all of them, the affected roles share a structural feature. They were entry points. They were where technical professionals started building the judgment they’d need five years later.
What companies are quietly doing
The shift isn’t showing up as a headcount gap yet. The current mid- and senior-level bench is intact. What’s changing is the expectation around new hires.
Companies are hiring for readiness, not potential. Smaller teams. Less tolerance for the developmental pace that used to be built into early-career roles. The same output is expected faster, from fewer people assumed to arrive capable.
A BCG analysis from earlier this year estimates 50–55% of U.S. jobs will be reshaped by AI over the next two to three years. MIT research published in April found the same pattern: not mass displacement, but task compression. The floor of expected capability is rising faster than the pipeline below it.
This is harder to manage than direct displacement because it doesn’t show up as a headcount problem. It shows up as missing experience layers — over time, in ways that are easy to attribute to something else.
The pipeline problem
Senior engineers, data leads, technical architects — these people exist because they spent years doing the junior work first. That’s where instincts get built: the sense for what breaks under pressure, what edge cases actually look like, what done means in context.
When the entry layer goes away, organizations don’t just have fewer people. They have fewer pathways for developing the people they’ll need in three to five years.
We’re already seeing early signs. Mid-level technical professionals handed scope without the foundation for it. Hiring cycles that can’t find senior talent, without fully recognizing they’ve been quietly cutting the conditions that produce it. Organizations with strong AI tooling and thinning institutional knowledge.
What this means going forward
Getting the work done and developing the people who’ll do harder work later were never the same problem. They just shared the same solution: entry-level roles doing routine work. That’s no longer true.
The traditional technical career ladder is changing faster than most companies admit. Experience accumulation doesn’t work the same way it did three years ago. Organizations that treat this as a standard efficiency play — and only notice the capability gap when it shows up in their org chart — are going to have a harder time closing it.
The companies handling this well aren’t waiting to see how it shakes out. They’re watching what’s changing in their teams, building deliberate developmental pathways that don’t depend on routine work existing, and adjusting before the gap becomes a crisis.
That’s not adaptation language. That’s just what it looks like to take the problem seriously.
OpenNova is a boutique technology partner that helps enterprises execute on cloud, data, and AI initiatives through specialized talent and embedded delivery teams. If you’re working on a scoped initiative that needs execution support, let’s talk about your project.
