How AI accelerates innovation: the power of 'recombination'

Innovation and artificial intelligence (AI) are inextricably linked. But how can AI actually contribute to innovation? A recent scientific paper from the University of Chicago offers a new perspective: AI, and in particular language models (LLMs), appear to be ideal for “Recombinatory Innovations” — innovation that occurs by combining existing ideas and insights.

What is “Recombinatory Innovation”?

Many innovations are not completely new discoveries, but are created by cleverly combining existing knowledge. This is what scientists call recombinatory innovation. Consider how the smartphone brought elements of a camera, a phone, and a computer together in one device.

According to the research, people often have difficulty initiating this form of innovation independently, for two reasons:

1. Insights develop rapidly, making it difficult to keep track of everything.
2. The amount of information is overwhelming, making it difficult to make the right connections.

How AI is accelerating innovation

This is where AI comes in. Language models such as ChatGPT, Gemini, and Copilot can:
- Process large amounts of information and quickly filter relevant insights.
- Make hidden connections between knowledge from different domains.
- Suggest new applications, based on existing knowledge.

This means that AI doesn't so much generate completely new ideas, but can help combine and restructure existing knowledge, accelerating innovation.

AI as a Catalyst for Science and Business

This principle is not only relevant for scientists, but also for companies that want to use AI for innovation. Imagine:

• In healthcare, AI can combine existing medical research to identify new treatment approaches.
• In product development, AI can analyse trends and customer needs and link them to existing technologies.
• In marketing and strategy, AI can help combine successful ideas from different industries into new growth opportunities.

‍