Jo Debecker is President and CEO of Akkodis.
getty
The future of engineering will not be defined by AI alone. It will be defined by how humans and AI work together.
AI is a human-created technology. It becomes valuable when paired with domain knowledge, judgment and accountability. Humans with AI can do more than AI can do on its own.
This is not a story about replacing engineers with agents. In complex engineering environments, the real value proposition is human ingenuity together with machine precision. That is how organizations create better products, accelerate innovation, shorten time to market and unlock new business models.
The future of engineering is not an “or” story. It is an “and” story.
The biggest shift I see is that human-AI collaboration is becoming more structured and intentional.
Organizations can use AI to automate many steps in a process. But at some point, they need a human control point that checks the outcome. As AI takes on higher levels of work, that control point may move up the ladder. But it does not disappear.
You still need domain expertise, interpretation of output, final decision-making and validation. This is especially true in complex and high-risk environments such as defense, the public sector, life sciences and aerospace.
That does not mean every workflow should be slowed down by human review. It means organizations need to design the collaboration model deliberately. Where can AI act autonomously? Where does a human need to validate? Where does the organization need deterministic automation instead of generative AI?
I believe LLMs are being overused today. If something is deterministic and you need a guaranteed outcome, don’t use an LLM. Build a script. With LLMs, there will always be some possibility of drift or hallucination. Automate where that makes sense. Use generative AI where it actually adds value. Then build in human judgment at the right control points.
AI is already moving into software, simulation, documentation and development. For engineers, this changes the way they work, but it does not replace them. They will work alongside AI to develop, test, design and document.
What becomes more important is creativity, innovation and critical thinking. Engineers will still need deep digital skills because someone needs to understand what AI has built, why it produced a certain output and how to fix it when something goes wrong. Critical thinking is the key area where human added value is crucial.
Accountability also remains human. If you want your plane to be compliant, you need an engineer to stamp it. If a self-driving car causes an accident, the answer cannot simply be “The AI did it.” In product and systems development, responsibility has to remain clear.
The human role will also include selecting the right AI model for the right job. If you ask me which LLM I use, my answer is: It depends. The role of the engineer is to select the right model, contain it, ground it and customize it for the industry where it needs to run.
Aerospace and defense are often at the forefront of hybrid engineering. They are highly regulated, safety-critical industries under extreme pressure to move faster, improve connectivity and shorten time to market. Human control points are essential in trusted and regulated environments.
Aviation is a useful example. An airplane has a life cycle of 25 to 40 years, and AI can help optimize maintenance, anticipate breakdowns, analyze sensor data and ultimately reduce time on the ground. But because safety matters, humans still need to oversee the process. Nobody is going to accept “The plane crashed, but it was AI’s fault.” When the stakes are high, people look at AI differently.
In traditional IT, we had the four-eyes principle. If someone made a change in production, a second pair of eyes checked it. Human-AI interaction in engineering is the modern version of that principle. It puts human oversight at the right control points to make sure what has been done makes sense.
A truly effective hybrid engineering team is hybrid in more than one way. It combines humans and AI, but it also brings together engineers, data specialists, domain experts and AI agents as one integrated unit.
The better model is more like a multidisciplinary scrum team, where different experts and AI agents look at the problem end to end. The ability to connect technical, business, data and engineering perspectives will help teams get results faster. And if something doesn’t work, they will learn and fail faster. Failing fast is always better than failing slow.
Engineering itself is also becoming more virtual. Teams are increasingly simulating systems, testing software first and building physical models for final validation. That creates more opportunities for AI and automation but also makes multidisciplinary collaboration more important.
If AI adoption is scaling, that is a good problem to have. But it means governance becomes critical. Organizations need reliability, consistent outcomes, ethical use and clear business value. If AI does not create value, stop it.
Governance, guardrails and human control points cannot be added later. They have to be built in. Engineering, data and business teams should all be part of the oversight model so context is clear, performance is tracked, drift is avoided and value continues.
Upskilling is equally important. Successful AI implementation should create more value for companies, and that value will create new work. The task is not to reduce people. It is to reskill them and redeploy in future skill areas.
AI fluency will become a basic skill, almost like reading. People will need to understand how AI works, how to prompt it, how to provide context and how to think in systems.
Too much of the AI conversation is black and white. Either you are a disruptor or you are being disrupted. Either AI replaces people or companies fall behind.
But reality is usually somewhere in the middle.
The future of engineering will be hybrid. AI will help engineers move faster, simulate more effectively, test more intelligently and make better decisions. But human expertise will remain central where judgment, accountability, domain knowledge and oversight matter most. AI must ultimately be deployed to augment human experience and elevate the potential of people and organizations.
Knowing where to use AI, where to use deterministic automation, where to place human control points and how to build cross-functional teams around the work will be the recipe for success.
The future of engineering is human ingenuity together with machine precision.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。