ChatGPT can get an engineering degree. What do we do now?

Estimated reading time: 5 minutes

With each generation of new AI model, more advanced benchmarks are being set to see how well they perform at commonly administered tests for humans.

A recent study by academics at the École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland should give those within the software engineering sector more pause than most. The survey found that AI assistants like ChatGPT can answer around 66% of exam questions correctly across 50 different technical and natural science subjects – including computer science and data sciences.

With 70% of school students saying in a recent survey that they use tools like ChatGPT in their schoolwork – whether they’re technically allowed to or not – the use of AI is likely to become normalized across the next generation of technology workers.

What hiring managers ought to do about that is the next big question.

Hiring and firing

“AI tools like ChatGPT are making it easier than ever for students to produce work without fully grasping the underlying concepts,” says Peter Wood, chief technical officer at Spectrum Search. “This means businesses could soon face graduates with impressive-looking CVs but limited practical knowledge.”

Businesses will have to rapidly redraw their hiring process for developers to accommodate for the fact that a university degree may no longer represent built-in knowledge to do the job without AI support. “Relying solely on degrees won’t cut it anymore. Practical assessments, real-world tasks, and problem-solving exercises will be key to separating genuine talent from those leaning too heavily on AI assistance,” says Wood.

Managing them will also prove tricky.  “We’re about to face a dilemma with junior developers becoming ‘AI-native’,” says Laurent Dougin, director of developer relations and strategy at Couchbase. “They will have always relied heavily on copilot AI tools to double check their work or even do parts of it for thee. While that’s great for efficiency, it creates a huge gap between junior and senior developers from a career standpoint.”

Ongoing education

For companies unwilling to risk generative AI’s hallucinations, investing in ongoing education programmes will be vital. “Companies will have to invest more in education – especially through hands-on experience,” says Dougin.

“AI-driven changes demand continuous learning. It’s no longer just a nice-to-have; it’s survival,” Wood says. “Businesses must create structured learning programmes, from onboarding bootcamps to regular skill refreshers.”

Those lessons should focus on building the foundational knowledge that universities are teaching, but students may well not be paying attention to. “For existing teams, fostering a culture of lifelong learning ensures employees stay relevant and can adapt as AI capabilities expand,” says Wood. “The businesses that prioritize learning will build resilient, future-proof teams.”

What to look for

The value of AI also can’t be understated: companies should seek out candidates who use it to help, but don’t rely on it as a crutch. 

“Rather than filtering out AI-aided degrees, employers should focus on how candidates use AI as a tool, not a shortcut,” says Wood.” It’s about hiring people who can work alongside AI, understand its limitations, and innovate beyond what it can offer. After all, in a world where AI can do the basics, true value comes from creative, critical thinkers.”