AI Library

1. Introduction

AI strengthens every phase of the End-to-End Innovation Process, often in fundamental ways. Think of it as an interplay between HI (“human intelligence”) and AI (“artificial intelligence”). How the two work together will keep shifting as the technology advances.

The pace of AI progress is stunning, and it’s a huge opportunity for how we shape innovation work. AI now delivers results, on certain tasks, that humans alone could never produce in the same time. The clearest examples: analysing large, messy datasets in customer and market research, and building visual models and prototypes. AI, however, works backwards. It draws on past data that already exists — words, cases, numbers that someone has already written down — and predicts what most plausibly comes next. That makes it excellent at everything repetitive and computationally heavy: summarising, translating, restating, filling in the obvious. But it has no mechanism for going beyond its inputs.

Human thinking runs the other way. You start from a theory — a forward-looking causal idea about how something could work — and that theory tells you which data matter and which experiment to run next. Where the data are thin or point the wrong way, a theory lets you act anyway: Kelvin and LeConte had the evidence on their side when they declared human flight impossible, and the Wright brothers had a causal theory of lift, propulsion and steering, plus a wind tunnel to generate the data nobody had yet. This gap between what you believe and what the data support is not a bug in human cognition; it is where new knowledge comes from.*

AI won’t replace the innovator, the designer, the creative mind who does more than execute tasks efficiently — who convinces people, through passion and conviction, to believe in a new business model, service, or product. AI won’t break down organisational barriers for you. It won’t build high-performing teams or rally them around a shared mission. Success still depends on these human skills. Now and in the future. And current AI tools rely on predictions and probabilities based on large datasets, which is different from human thinking usually influenced by causal logic and reasoning. Nevertheless, we embrace the technological advancements and use them extensively.

In practice, we split AI use into two kinds of tasks.

• Creative tasks: Any task that fundamentally shapes the outcome of innovation work stays, first and foremost, with human intelligence. Not because humans always do it better — but because our creative skills, as designers, are what make the real difference. We need to believe in our idea and fight for it. We need to be involved emotionally and that takes “skin in the game”. We stay involved and keep shaping the core of our innovation ourselves.

  However, that doesn’t mean we skip AI here. It means we work in three steps:

  • Step 1: Build a first version using your own creative skills — for example, the core elements of a business model or the storyline for a marketing concept.

  • Step 2: Use AI tools to improve, expand, or challenge that first version from new angles.

  • Step 3: Review what AI produced in step 2, then use your creative skills again to integrate what is useful. Your idea can change radically at this point — even should, if the input calls for it. But you stay the designer of your own innovation.

• Supporting tasks: Any task that is repetitive or where the past is a fair guide and enables creative work gets full AI support. This is pure efficiency. AI beats human ability at analysing large datasets, pulling out key insights, tracking down missing information, or even building digital click dummies and prototypes. It’s a genuine advantage for instance to generate five landing page variants at the push of a button, each pitching a new business model to a specific audience. But only once a designer has already shaped the story behind that business model — that’s a creative task. Supporting tasks rarely demand deep emotional investment from the designer. Using AI here doesn’t replace the Business Designer. It makes us faster and smarter.

For both types of tasks: Let your AI read this knowledge base and use proven prompts and briefings that deliver results that are highly useful for your work. Let's go.

2. Use Cases

We have developed an ever-growing library of "AI Skills" that take on or simplify important tasks along the End-to-End Innovation Process. These can be called via API on request and used in your own projects. The clever part: each skill draws on our years of experience and accumulated material to improve the results.

2.1. Phase I

Phase I: Picture of the Future & Strategy

2.2. Phase II

Phase II: Playgrounds & Exploration

2.3. Phase III

Phase III: Portfolio & Evaluation

2.4. Phase IV

Phase IV: Validation & Refinement

2.5. Phase V

Phase V: Execution & Go-to-Market

*Source: Felin & Holweg, "Theory Is All You Need: AI, Human Cognition, and Causal Reasoning", Strategy Science (2024).