I asked AI how to evaluate ideas

In the ever-changing landscape of creativity and innovation, we are constantly striving for new and ground-breaking ideas. Over time, the role of AI, such as ChatGPT, in evaluating these ideas has piqued our curiosity. We want to know if AI can be more than just a tool – can it be a guiding companion in our quest to evaluate ideas?

As we venture into this world, we reveal that AI offers both promising prospects and limitations in the innovation process.

Understanding that innovation is a process

Innovation is not a linear journey; it is a complex, multifaceted process that involves different stages and non-linear thinking. Divergent thinking, brainstorming, evaluation and out-of-the-box ideas are all part of the journey. The innovation process is about exploration, experimentation and problem solving rather than a strict, step-by-step goal-oriented approach.

Two distinct parts to consider: idea exploration and idea realization

When we turn to AI to evaluate ideas, we find that it makes a similar mistake that humans make when evaluating ideas. This is not surprising since AI is trained on human behavior and text. The most common mistake is to focus solely on idea realization while overlooking the importance of idea exploration.

Idea exploration

During idea exploration, we must evaluate factors such as the surprisingness of an idea, the uniqueness and the creative height. As Albert Einstein once noted, imagination or creative thinking is often more important than intelligence in the pursuit of innovative solutions. This phase requires us to think outside the box and embrace new possibilities.

Idea realization

However, AI can be a valuable resource during idea realization. It can provide insights into feasibility, targeting, market potential, scalability, cost-benefit analysis and more. It is a powerful tool in the logical perspective of which ideas are feasible.

Creativity is not a linear process

The innovation process is not unidirectional but rather a circular journey. The realization of an idea may reveal new challenges or problems, prompting a return to idea generation and exploration. Here we must make sure that we do not confuse the evaluation criteria for idea exploration with those for idea realization. The ability to seamlessly move back and forth between these phases is key to an effective innovation process.



Sometimes our instincts and intuition say more about an idea than any evaluation criteria can. When both our gut feeling and a logical evaluation match, it is often a strong indicator of a promising idea. In our exploration of AI’s role in this process, it becomes clear that while AI excels in logical aspects and understanding the realization part of ideas, it can struggle to identify the potential of out-of-the-box ideas and above all to evaluate their creative value. This becomes especially visible if you ask an AI if your idea is good or bad. It can only reason based on existing knowledge, which in itself can be valuable, but it has no opportunity for gut feeling given new approaches that are not yet documented and trained into the AI.

In conclusion, AI, such as ChatGPT, can be a valuable assistant in the innovation process. It excels in providing logical insights and evaluation criteria for idea realization, making it a valuable tool. It may even contribute to creative thinking, but is not very helpful in understanding the balance between surprise, news value and realizability, especially in the context of idea exploration. Effective innovation is a dynamic blend of creative exploration and logical evaluation, with both human and AI contributions having their unique roles to play. Teaching an AI to understand creative heights requires constant learning, unlearning and selective learning in leaps and bounds of innovation. It is an exciting future we are facing, where AI needs to enter the field of innovation to solve the climate challenges, but where there are a lot of challenges for AI as well.