Operator Group Chat: If AI can code, what makes engineers stand out in 2026?

Mehdi Ghissassi

CPTO at AI71; former Director of Product at Deepmind & Google; serial AI angel investor

As AI takes over much of the routine coding, the best engineers won’t stand out by typing faster — but by thinking better. What will differentiate talent is the ability to frame problems, design scalable systems, make smart trade-offs, and ship real products amid ambiguity. The standout engineers will combine strong technical fundamentals with product sense, curiosity, and the ability to orchestrate AI tools effectively. In short: not just coders, but builders who can turn messy problems into working solutions with speed and judgment.

For early-career engineers, that means your edge won’t come from memorizing tools but from owning problems and learning fast. Use AI to accelerate you, but show you understand what you’re building and why it matters. Focus on delivering real projects, asking thoughtful questions, understanding trade-offs, and developing product intuition. The path to standing out now is clear: become someone who ships value, not just code.

Sebastian Enderlein

CTO at DeepL; former VP Engineering at Personio; former Head of Cloud Infrastructure at Uber

The impact of AI on the engineering profession is undeniable. Increasingly, entry-level jobs are going unfilled because there’s a fear that AI will replace entry-level engineers. 

AI coding tools are now widely adopted, but while they can promise speed of delivery, their quality varies significantly. Around 50% of industry leaders report value, while 50% say these tools create more bugs to fix. So what does this tell us about the changing role of the engineer? 

AI has dramatically lowered the barrier to writing code, but it hasn’t lowered the bar for engineering judgment. They’re exceptional at generating code quickly, but far weaker at assessing correctness, making sound architectural trade-offs, or deciding whether you’re building the right thing in the first place. That’s why we’re seeing a split: legacy systems struggle to realize the benefits of AI, while newer companies actively shape their architecture to work well with AI agents.

Today’s great engineers must be fluent with AI as a tool, but not dependent on it. They need to know how to push an agent to move fast, and equally, where it will fail or mislead them. System design has become even more critical because clean, well-structured architectures are easier for both humans and AI to reason about. 

When hiring for engineers, we still look for deep system understanding and a strong bias toward quality and impact over volume. While AI pushes engineers to operate at a much higher level, the greatest benefits come when experienced engineers use AI to supercharge their judgment rather than replace it. Interviewers should know how to test this by asking candidates to build something sophisticated with AI in 30 minutes or less.

The fundamentals haven’t changed: strong product sense, fast experimentation, and building for real human needs still define engineering. AI just amplifies those strengths.

John Lyle

VP Engineering at Unobravo; former Engineering Director at Meta

It’s worth differentiating here between “generalist engineers in the age of AI” and “AI engineers”.

Firstly, with regard to hiring generalist engineers. Essentially, all competent engineers are now using AI in their development, so it’s part of their skill set.  Beware outlandish claims as well as skepticism – the reality is that anyone not using it is behind the curve.  Anyone claiming crazy impact, though, is also probably not doing “engineering” but vibe coding, or working in unrealistic environments. 

Interviewing has to change, as it is in many places. Take-home tests are fine (if onerous); online coding tests are dead; VC coding interviews are challenging but still doable. In-person is best but impractical. Candidates should walk through how they used AI, show their prompts, discuss their process, and show the actual result. Candidates who think deeply about how to use AI most effectively in the best context likely stand out (see here), as do those with experience in LLM engineering, who can show examples of pitfalls and challenges. It’s all about showing real tensions and trade-offs. Just as engineers don’t only write code, they shouldn’t just be using AI to write code. Whether that’s designing, spec’ing, documenting, or planning, there’s lots of value everywhere. 

Now, with regard to hiring AI engineers. We also need to differentiate between classic ML/NLP/AI engineering and LLM engineering, as they’re really quite different technologies now. That said, some common knowledge and mindset are helpful – experimentation, data engineering, data hygiene, and measurement.  

This is a fast-moving space, so it’s key to find people who are active in research and experimentation; specific technologies are less important. Consider that demonstrating experience with ChatGPT a few years ago is not the same as where it is now. As ever, the model is the tip of the iceberg, and largely outsourced, so the real meat is in evals, datasets, context building, and so on.  This means that a lot of the work is internal tools and wiring things together, so lots of generalists can be super impactful. Given all this, cross-functional, collaborative, and product-thinking skills remain super important.

Priit Kaasik

Co-founder & former CTO at Katana; ex-Microsoft

We are witnessing a rapid evolution from ‘Vibe Coding’ into true Agentic Software Engineering, accelerated by breakthroughs like Google Antigravity and Gemini 3. Consequently, the role of a developer is shifting from a writer of code to a builder and maintainer of agentic systems—a shift that will likely eventually permeate all knowledge work.

In this new reality, manual coding skills are less relevant; you could use AI to analyze AI-generated code, but why would you? Instead, the premium is on theoretical informatics and Systems Thinking. The ability to design architecture and understand complex system interactions is what differentiates a great engineer from a mere operator.

When hiring in 2026, I look for deep AI fluency. A candidate should have such an intuitive grasp of these tools that they can look at an outcome and almost identify which agent, LLM, or version was used to generate it. Beyond that, I prioritize Project Management skills and Systems Thinking. Finally, because technical baselines are leveled by AI, personality traits are paramount; I look for individuals who complement the existing team’s dynamic to ensure a diverse approach to problem-solving.

Disagree with our Group Chat? We’d love to hear your thoughts.