Innovation in the age of AI : Towards new foundations of innovation

Philippe Aghion, 2024 Nobel Prize winnerin Economics and world-leading expert on innovation-driven growth, demonstrates that creative destruction depends not only on technological advances, but above all on the capacity of organisations to absorb, reinterpret, and transform them into sources of competitive advantage.

In the age of AI, where technologies evolve faster than companies can integrate them, this idea becomes central : disruption no longer comes from the tools themselves, but from those who manageto reinvent their ways of learning, experimenting, and innovating.

So the question becomes : how can were think innovation when technology moves faster than our organisational,decision-making, and creative models ?

The myth of endlessly shortening innovation Cycles

For two decades, the dominant narrative around innovation has rested on one conviction : technological cycles are getting shorter. In both the automotive and telecommunications industries, this acceleration is tangible. The smartphone has become the icon of this trend : onemodel replaces another every year, with new versions following at anever-faster pace. The automotive industry follows a comparable trajectory electrification, driver assistance, connectivity, digital design  the innovation cycle has been compressed into a decade.

Yet this observation is not universal. Some technologies, on the contrary, go through long phases of stabilisation.The example of telecom networks is instructive : after the rapid succession of 2G, 3G, and 4G, 5G, despite its ambitions, is settling in for the long term. Genuinely differentiating use cases are slow to emerge, and the technological leap is less perceptible than in the past... much like smartphones today, where new versions look so similar to their predecessors…

In other words, the acceleration of innovation cycles is not a natural law : it is selective and discontinuous. Andone of the fields where it manifests most intensely today is, without question,artificial intelligence.

Artificial Intelligence : A catalyst and a mirror of innovation

AI, and even more so generative AI (GenAI), is today the technological field in which obsolescence happens the fastest. A state-of-the-art model can be overtaken in six months or less. This unprecedented pace makes AI both a driver of innovation and a testing ground for what innovation could look like in other sectors.

Innovation in AI : Accelerated technological obsolescence

Never has a technology evolved at such a pace. The successive arrival of ever more powerful foundation models illustrates a phenomenon that is both new and very old : the decoupling of the local state of the art from the capacity for global deployment. Industrial and service players must now learn, or relearn, how to innovate in an environment where the technological foundation is renewed before it has even been fully integrated.

Innovation Through AI : An accelerator

AI acts as a cross-functional accelerator of innovation. It enables certain phases to be automated, iterations to be multiplied, hypotheses to be simulated quickly, and creative or technical directions to be generated. In design, R&D, content production, and modeling, it shortens cycles and democratizes access to capabilities that were once reserved for a handful of experts.

At one leading distributor of industrial equipment, work launched as early as 2022, at a time when the topic was still in its infancy, allowed the company to gain a head start : higher productivity, greater organizational maturity, and a clearer understanding of the field of possibilities. While others are still at the proof-of-concept stage, this group has integrated AI into its day-to-day operations.

A Structural Limitation : AI does not innovate, It optimizes

But this limitation must also be stated clearly : AI does not invent. It interpolates fairly well, but extrapolates very poorly. In other words, it excels at combining what already exists, at connecting known worlds, but it does not step outside the playing field. It can mimic creativity, but not transcend it.

Breakthrough innovation, the kind that changes the rules of the game, creates a new market, or introduces a new use, remains a human prerogative. AI can contribute to it, but never trigger it. It may help create an unexpected combination between two existing ideas, but for now, it cannot imagine an entirely new territory… though the full scope of what cybernetic AI may one day make possible remains unknown.

What innovation in the age of AI teaches us

Innovating with AI means accepting new foundations for innovation. It means learning to manage innovation in a world where technology moves faster than organizations, and where value no longer comes solely from technical mastery, but from the way we interact with technology.

Conviction No. 1 : Accept being slightly behind

One of the paradoxes of this new era is that one must accept not being at the state of the art in order to move forward. Trying to constantly keep pace with the latest version of a model or algorithm means locking oneself into an endless and unproductive race.

The challenge is no longer to be cutting-edge, but to be ahead of one’s market and competitors : to move early, without waiting for full technological maturity, which is precisely what has allowed some organizations to capitalize quickly on AI. By contrast, many players today are multiplying POCs without ever scaling them, trapped in a cycle of perpetual evaluation.

Conviction No. 2 : Put people back at the center, the “Human in the Loop”

In best practices for using AI, innovation processes are no exception. Here too, the fundamental principle of “human in the loop” applies. AI is a remarkable amplifier of inspiration, but it needs creative impulse and critical judgment. The initial idea remains human : in the prompt, in the direction given, in the final judgment.

It is the quality of the interaction with AI that drives innovation. At one of Eleven’s luxury clients, for instance, AI is used to generate window display variations, create visual concepts, and accelerate the exploratory phase. But the choice of concept, the aesthetic coherence, and the emotional dimension remain deeply human. At another specialist in in-store marketing, AI helped rethink the layout of commercial spaces, but it is human interpretation that turns generated suggestions into viable concepts.

Innovating with AI therefore means both learning to iterate faster and reworking the proposals generated by the machine. Humans intervene before, through framing and inspiration, and after, through selection and shaping.

Toward an augmented innovation method : Teaching innovation with AI

The question is no longer “should we innovate with AI ?” but rather “how is AI transforming the way we innovate ?”

Three structuring principles can help guide this transformation :

1. Create momentum rather than aim for excellence.
2. Hybridize human intelligence and artificial intelligence.
3. Capitalize on what AI teaches us for other technologies.

‍

Redefining the role of humans in innovation

Innovation in the age of AI is neither a renunciation nor a substitution. It is a shift in the human role : less about execution, more about direction, selection, and meaning. AI makes visible what innovation has always been : a collective, iterative, and imperfect process, made up of adjustments, informed bets, and intuitions.

But this new era also raises major strategic questions :
- How do you choose your battles in a technological landscape that is evolving faster than organizations themselves ?
- How do you decide the right moment to adopt, invest, and industrialize ?
- Which innovation models remain truly relevant when creative, analytical, and technical tools themselves are changing in nature ?
- What should humans still do, and what can now be entrusted to machines ?

These questions are now central for all organizations. Answering them is already the first step toward building an innovation strategy suited to the age of GenAI.

‍