17 tech workers share how to stay relevant as AI takes jobs

Key Takeaways

• Senior workers who resist AI tools face more risk than juniors who lack experience
• The most valuable skill may be catching AI mistakes, not prompting AI correctly
• Physical crafts and real-world work remain AI-resistant career paths

Finding a tech job has rarely been harder. AI tools have gutted entry-level positions, some companies have stopped hiring juniors entirely, and automated recruitment systems mean candidates send hundreds of applications without a single human response. The industry feels precarious. So Sifted asked 17 European tech workers and investors a simple question: what should people actually do to stay relevant?

Their answers ranged from practical (master AI agents, learn to prompt well) to philosophical (become a baker). But a few themes kept surfacing. Critical thinking beats raw knowledge. Pattern recognition matters more than following trends. And the real skill keeping people employed right now? Knowing exactly when the AI got it wrong.

Why senior workers face more risk than juniors

The conventional wisdom says entry-level workers are most exposed to AI automation. Stan Marchand, CEO of mobile app creator Rocapine, disagrees. "The most vulnerable people may not be juniors but seniors who confuse experience with knowing the best way to work," he told Sifted. "For seniors, the challenge is to put your guard down and be ready to be humbled on AI."

Jack Davies, VP of European marketing at VC firm Antler, made a similar observation after attending a recent hackathon. "There were 18 and 19 year-olds doing things with Lovable, the vibe-coding platform, that Lovable people didn't know you could do," he said. "Instead, it's going to be tough for people entrenched in the non-AI ways of doing things."

The implication is stark: decades of expertise can become a liability if it calcifies into resistance. Juniors have less to unlearn.

The skill that actually keeps you employed

David Erhun, head of comms at VC firm AENU, offered the bluntest take. "For knowledge workers like me, it's too late. So go become a baker." His brother runs Juno, one of Copenhagen's best bakeries. No AI agent is shaping cardamom buns by hand.

But Erhun also shared something revealing about his own workflow. He has had nearly 190,000 conversations with Claude Code since mid-March. "The skill keeping me employed is being able to point at exactly what the agent got wrong and send it back to fix it. Again. And again. And again."

His advice boils down to two paths: make something physical, or get very good at catching the machine's mistakes. Both require judgment that AI cannot yet replicate.

Three AI skills that will remain relevant

Lara Kennedy, head of marketing at AI upskilling platform Ivee, broke down the technical skills she thinks will matter over the next decade:

1. Natural language prompting. Knowing how to ask AI the right questions.
2. AI agents. Managing them, prompting them, running multiple agents in parallel.
3. Foundational models. Connecting your model to all your context, tools, and data. Building it as a second brain.

Kennedy emphasized that thriving now means being an early adopter and staying in the AI conversation. Long-term success depends on education, critical thinking, creativity, and strategic positioning.

Why introverts might have an advantage

Bianca Zwart, chief strategy officer at fintech Bunq, argued the most important traits for the future are critical thinking, resilience, and no ego. AI makes knowledge accessible and automates execution. What it cannot automate: spotting a flawed assumption, changing course when a bet fails, or bringing human empathy to a complex decision.

"It might actually be a great time for us introverts," Zwart said. "We are used to observing before speaking, spotting patterns, and asking the question that changes how everyone sees the problem."

Pattern recognition over trend-following

Renée Shaw, who handles brand and social at AI startup tl;dv, pointed to a skill that does not show up on LinkedIn profiles. "The thing worth getting good at is pattern recognition," she said. "Everyone reads the same newsletters and lands on the same takes. Being able to look at another industry and work out what that means for your own corner, that's valuable."

Her advice: get good at zigging where others zag. Connect two things that do not obviously go together. AI can summarize consensus. Humans spot the gaps.

Work where AI has no data

Rupert Small, CEO of AI startup Egregious.ai, offered a geographic metaphor. "Doing something in the real world, in spaces where AI doesn't have much data, easy access or a good understanding, will increasingly be where humans do their best work and thrive."

This aligns with a broader industry trend. AI systems trained on internet data excel at digital tasks. Physical work, novel domains, and contexts with sparse training data remain human territory. The question is how long that lasts.

The uncomfortable consensus

Across all 17 responses, one theme kept recurring: the tools are not optional. Jed Rose, partner at Antler, put it simply. "Lean into the AI tools and master them. See them as empowering, rather than replacing, jobs."

But mastery is a moving target. Every few months, the tools improve. What felt like expertise in January becomes table stakes by June. The arms race continues.

Frequently Asked Questions

Which tech jobs are most at risk from AI?

Entry-level coding, content writing, and routine data analysis face the highest near-term risk. But senior workers who resist AI tools may be equally vulnerable.

What skills should I learn to stay relevant with AI?

Critical thinking, pattern recognition, AI prompting, agent management, and the ability to identify when AI outputs are wrong. Physical crafts also remain AI-resistant.

Are junior or senior workers more at risk from AI?

Both face risks. Juniors lose entry-level learning opportunities. Seniors risk obsolescence if they resist new tools or assume experience alone provides security.

Should I learn to code if AI can write code now?

Understanding code remains valuable for reviewing AI output and catching errors. The skill is shifting from writing code to evaluating and debugging AI-generated code.

Source: Sifted