Artificial intelligence is not a magical shortcut to quick gains, but a tool that must be used deliberately. An industry expert panel warned that success rests on companies’ preparedness, people, and sober expectations. The hype around “AI” often mixes up terms and promises more than it can deliver—especially in manufacturing.
Prerequisites and Realistic Expectations
A company “built on AI” doesn’t make sense on its own; AI is a tool, not a goal. Implementing it requires a financially robust company and, above all, people who understand the technologies and know how to ask for what will benefit them. The basics are to define what we want to achieve, identify internal partners, and enable them to build the necessary skills.
There is no universal formula for calculating ROI—it depends on the specific use. Marketing may promise “a lot of bang for a few euros,” but full-fledged deployments tend to be costly and the payback is more long-term. The decisive factor is the choice of approach: whether to build your own solution or use cloud services and off-the-shelf models that can lower the cost of getting started. A sensible path runs through a clear use case, a pilot, and measurement of real benefits.
Generative AI in Industry: Between Hope and Disappointment
In industry, generative AI in the form of chatbots may bring more disappointment; its benefits are better suited to marketing or office analytics. Real value in manufacturing comes from other approaches: generative design, which looks for shapes that save material and energy, or digital twins, thanks to which design and testing are handled virtually. These are development tools that shorten time and reduce prototyping costs, not “wizards” that will solve manufacturing on their own.
We need to distinguish between automation and AI. A lights-out production line is an example of state-of-the-art automation, not artificial intelligence; it would be AI if the system itself proposed changes to the product or the process. The key is to stop conflating terms and focus on specific technical benefits, not the “AI” label. Not everything that calls itself that is groundbreaking for industry.
Competitiveness, Education, and Advice for CEOs
Europe is often slow in technology, overregulated, and invests little in research, which is already evident in competition with the USA and Asia. AI alone will not save us; what matters is long-term investment in development and the ability to create our own products, not just manufacture others’. In many cases today, the question of survival is more important than quick quarter-to-quarter returns.
People are key: without strong mathematics and physics we won’t get far. Collaborations with schools exist, but interest fluctuates; it is equally important to prepare current employees so they don’t discourage young people and can integrate them. For CEOs: define a clear goal and use case, decide between the cloud and an in-house solution according to risk and capacity, create an internal team to work with suppliers, and implement gradually with measurement of benefits. And always separate marketing promises from the reality of the shop floor.
