The three stages of AI in innovation

We’ve all heard a lot about AI recently, fuelled by the arrival in our daily lives of OpenAI’s ChatGPT. Generative AI (GenAI) was unheard of 12 months ago (aside from people that worked in the AI sector) and is now one of the most hyped technologies in years, with further GenAI tools such as Google Bard, Microsoft Bing and ChatSonic emerging.

I won’t get drawn into the “shiny object” hype or the safety considerations around this important technology, particularly given the recent AI Safety Summit held at Bletchley Park which was dominated by GenAI talk.

Instead, I’d like to offer a different perspective – how AI is used for innovation. Innovation is integral to meeting some of the challenges the world is facing right now, from finding solutions to help address the climate crisis to how best to supply clean water in developing countries.

As a practitioner who has been working with machine learning and, more recently, AI and Large Language Models (LLMs) such as ChatGPT in the field of innovation, I know how much value it can add in this area. I’d categorise the usage of AI for innovation into three main stages.

Stage one: connecting the dots
We founded Wazoku on the belief that humans can solve any problem when supported by the right technology. We swiftly understood that basic AI / Machine Learning – could help reduce complexity and duplication in the basic process of generating and gathering ideas. It could also accelerate collaboration and transparency by helping people who had similar ideas find each other in large, complex organisations and work together on ideas and innovation. The machine made the connection that the humans created.

This ability to make connections also means that any ideas that don’t work right now or for their original purpose, aren’t lost or forgotten. Instead, they live in the system memory and when, in the future, a company is looking for an idea in that area, the AI can connect that historic idea with the current need for a solution. Recently, one of our engineering customers solved an urgent issue with an idea that had first been captured more than four years previously.

Stage two: understanding the patterns
Open innovation is an approach to problem solving that involves organisations asking questions (Challenges) to and capturing the input of thousands of experts (Solvers). It is an extremely powerful proposition and can help solve a staggering different array of problems and challenges, across many industries and sectors. These include helping to create cleaner drinking water in developing countries to protecting astronauts on space walks, and many more in between.

However, knowing the nature of the thousands of Challenges they have solved as well as the specific skills and expertise these Solvers have is extremely difficult. Simple tagging and key words could never uncover the rich details required to truly understand questions such as ‘are there trends or patterns in Challenges and solutions?’ or ‘what skills, knowledge, and experience have Solvers demonstrated that is not claimed on a CV?’ This can work both ways. People could list a skill on a CV that they actually don’t have, but also, the rigid structure of a CV can lead to the vital information required on a Solver’s overall skillset being left off.

AI has helped achieve this. Through analysis of data, covering more than 20 years of solutions to Challenges, it is possible now to see precisely what skills are contained in a community, how trends in problems have changed over time, and the evolution of AI Challenges over time. It would be impossible to understand the rich complexity of a crowd without AI reading, understanding, and categorising information. This process of clustering would have taken thousands of hours of work and still wouldn’t have been anywhere near as comprehensive or effective.

Stage three: drawing the picture
The emergence of GenAI has heralded a further boost to innovation. It can assist with approaches to writing or refining problem statements or potential solutions and can also reframe problem statements to direct the question to different crowds. This becomes incredibly useful for making a technical topic more accessible to non-technical Solvers, inviting a broader range of perspectives when seeking ideas.

This will play a critical role in encouraging more diverse groups to participate in the innovation process, and will make it faster and simpler to launch innovation Challenges. Crucially, the addition of more diverse thinking should also generate more impactful ideas.
Humankind has demonstrated that the scale of some of the innovation Challenges the world is currently facing, will not be addressed without the support of technology. It’s my belief that AI can be a powerful aid to human ingenuity and creativity and can play a vital role in addressing those challenges and helping to make the world a better place for us all.

+ posts

Rosemarie Diegnan is the Chief Product and Customer Officer, and co-founder of innovation scale-up Wazoku, which works with customers, including HSBC, NASA, AstraZeneca and Enel, to crowdsource ideas and innovation.

CIF Presents TWF - George Athannassov

Newsletter

Related articles

Hybrid IT Infrastructure: Integrating On-Premises and Cloud Solutions

Blending traditional on-site tech with cloud services creates a...

Why APIs are the new frontier in cloud security

Organisations are seeing a proliferation of APIs (Application Programming...

Tackling AI challenges in ethics, data, and collaboration

We can safely say that artificial intelligence (AI) was...

The evolution of the CISO

What began as a technical innovation on the hacker...

Building Trust: Uniting Developers & AppSec Teams

The relationship between developers and security teams has typically...

Subscribe to our Newsletter