Laurence Bastet interviews Christian Auriach
Publication : September 2025.
LB: In your opinion, what is the biggest obstacle to innovation in business?
CA: For me, there are two main obstacles.
👉 The first is the culture of the business plan.
In business, decisions are based on rationality, KPIs, and forecasted results. But innovation, by its very nature, does not respond to an already expressed need or a clearly identified target. It requires other, more exploratory, more open logics, and for this, there are increasingly mature methods, exploited in particular by unicorns or future unicorns.​
👉 The second is the weight of history
Many companies carry what I call a “history of false innovations”: past projects that did not create value and left traces in the collective memory. Yet the reality is encouraging: in ten years, the success rate of true innovation leaders has increased tenfold. The difference? Some innovate, others simply modernize.
Successful pioneers are willing to invest in a portfolio of projects, take calculated risks based on statistical logic—in short, truly innovate. They leverage the state of the art in innovation strategy and management, from open innovation approaches to modern real options-type valuations (see the excellent website strategic-finance.eu), including effective monitoring. The others, the followers, are content to imitate. Driving your digital transition is a modernization effort with a view to optimizing your cost/perceived quality ratio, not an innovation whose objective is to create a new offer or asset capable of boosting your pricing power.
LB: How do your students react to the rapid pace of innovation?
CA: “I distinguish two types of audiences.
👉 Technical profiles
Engineers, specialists, etc..., who master technology, overall welcome innovation (technological or not) with curiosity and enthusiasm.
👉 Other profiles
Others are less comfortable with technical subjects. I sometimes speak to real arithmetic-phobes. Their reaction is different: sometimes an aversion, or even a visceral fear. They fear losing legitimacy, having to step aside for hard sciences . Many ask themselves: “What is the role of a non-engineer in an era where AI is becoming everywhere, regardless of sector, level, or profession?” My conviction is clear: what matters is not technical mastery per se, but the ability to create value.
Let's take an example: the field sales representatives. Tomorrow, they won't disappear, but their role will evolve. They will become strategic sales representative, ensuring the reliability of data and capable of using it as a business lever. “Non-engineers” didn't disappear during the web revolution. They won't disappear in a post-AI era, but their role is changing.
Another example: last June, I gave a talk to interim managers. We built a data repository step by step to feed an AI-powered tool. The tool's objective: to conduct a strategic business situation analysis, focused on the value chain. You don't need to be an engineer to do this. However, you do need to master a wide range of economic models and know how to assess the relevance, reliability, and timeliness of data—the new gold. Never forget that AI is a program built with data. The founder of the company Scale AI, for example, understands this well. He devotes all his energy to creating datasets for the AI industry.
LB: How do you train students to think critically?
CA: The answer to this question can be very long. But here are some basics.
👉 AIs learn from specialists, and replace generalists
I always tell them, “Start with your own reasoning, then ask the AI to complement, extrapolate, challenge.” An excellent prompt is the description of a human work (method and result), to be extrapolated. Doing the opposite would be a mistake, because AI has no context by definition. On the other hand, it is a formidable tool for testing, comparing, and enriching one's own (human) analysis. To achieve this, you must choose a fine specialty and cultivate it constantly. It's the price you pay to stay relevant. What does the future look like? 9 billion humans, 9 billion specialties.
👉 understand how it works, at least in broad outline
Knowing the 5 main branches, the 6 main levers and the 2 main biases of AIs, understanding the differences in design and potential for technical and functional evolution between them, is the minimum that any new "honest human" must learn.
In short: critical thinking means keeping control, not delegating your thinking to the machine. Asimov and his laws of robotics were right: in short, humans must remain in control of the on/off button.