Innovation is Divergence (Part 2): How AI Amplifies Creative Thinking (and Why It's Not a Replacement for Humans)

AI will be part of our world. For now and forever. We have to learn how to co-inhabit this world with it.

AI is what economists call a “general purpose technology (GPT)”. Yes, yes — chatGPT and all Large Language Models (LLMs) are transformer-based technology (GPTs) and are a GPT. The?irony.

A general purpose technology is like electricity and the internet. You can use it any which way you want. In fact, all the uses you can use it for are still not fully known. You have to discover them and enable new business models around them. The only way to do that is to experiment and learn.

We have to acknowledge that AI will have biases, it will hallucinate and there is potential for misuse. However, the only way to learn about this is by engaging with it and figuring out how to work with these limitations and challenges.

At the same time, Generative AI is an exponential technology—in the sense that it is not growing by 5% or 10% per year, but growing at 5X and 10X on various parameters. This makes it challenging to predict its potential. All of this combines to make it relevant that we try to learn.

Today, I am starting a new experiment that will try to infuse AI into everything I do. I have been following closely the developments for more than a year and have been enamoured by the rate of change and possibilities. chatGPT, Google Gemini, Claude AI, Bing, Perplexity, Grok AI and Llama. I have tried them all and exploring all of them. It is very clear that we are entering a space where Ai + Humans will be powerful.

For the past few months, I have been exploring strategy, innovation, and design with a touch of Drucker. The goal has been to share frameworks, ideas, methods and approaches to make decision-making in organisations better. Now we need to take this to the next level. If not for anyone, for me to learn how to combine this capability into this kind of work.

Ethan Mollick has been a pioneer in this work and with his colleagues have been pushing the boundaries of AI and then helping all of us learn from that. His book Co-Intelligence on the power of LLMs and Gen AI is a good starting point. He provides a range of principles, including Always invite AI to the table. This is the plan here.

Every Monday, I write an article in the newsletter. We have talked about the importance of core-competencies, dissent in decision-making and how innovation is a discipline. Now, most Thursdays, I will invite AI to the table. What does that mean? I will take the essence of the article on Monday and see how AI can help or not help. As we explore the capabilities of AI we need to also understand its limitations. Whether its inability to do something, or hallucinate or provide nonsensical responses. At the same time we will know it’s capabilities.

This week in the innovation series we look at Innovation is divergence. Today, let’s build on that.

I said:

At its core, innovation requires us to create multiple options. To be creative, to try, to fail, to learn and to create new ways of doing things.

Ethan writes in his book, “Co-Intelligence”, the idea of the Alternative Uses Test (AUT) used as an ability to test the creative capability of humans.

One such test is known as the Alternative Uses Test (AUT). This measures an individual’s ability to come up with a wide variety of uses for a common object. In this test, a participant is presented with an everyday object, such as a paper clip, and is asked to come up with as many different uses for the object as possible. For example, a paper clip can hold papers together, pick locks, or fish small objects out of tight spaces. The AUT is often used to evaluate an individual’s ability to think divergently and to come up with unconventional ideas. (emphasis mine)

This divergent thinking is key for innovation.

Let’s try something. In my innovation workshops I use this all the time.

Come up with creative business ideas with $100 and a bicycle. You have 2 minutes to come up as many as ideas as you can.

Stop reading. Take the 2m and do this exercise.

Did you try? How many did you come up with?

Let’s do the same with AI. In 4.6 seconds this is the output. Google Gemini 1.5 Flash is used here.

Business Ideas with $100 and a Bicycle:

Simple:

1. Mobile Bike Repair: Offer on-the-go bike repairs in busy areas, parks, or bike paths. Use the $100 for basic tools and supplies.
2. Bicycle Delivery Service: Deliver small items or food within a local radius. Use the $100 for marketing materials and a delivery bag.
3. Bicycle Tours: Lead short, themed tours (e.g., historical, scenic, food-focused) in your local area. Use the $100 for maps, brochures, and snacks.
4. Rent-a-Bike: Offer short-term rentals for tourists or locals who need a bike for a day. Use the $100 for a simple rental system and signage.

Complex:

1. Sustainable Food Delivery: Partner with local farmers and restaurants to deliver fresh, locally sourced produce and meals on a bike. Use the $100 for marketing and delivery containers.
2. Mobile Photography Studio: Offer on-location photography services for events, portraits, or product shots. Use the $100 for basic equipment and props.
3. Urban Art Tours: Combine cycling with street art appreciation, showcasing local murals and artists. Use the $100 for tour guides and promotional materials.

Outlandish:

1. Human-Powered Drone Delivery: Create a custom bicycle attachment that launches small drones for fast deliveries within a limited range. Use the $100 for materials and prototype development.
2. Bicycle-Powered Electricity Generator: Develop a system that uses pedal power to generate electricity for charging phones or small devices. Use the $100 for materials and engineering assistance.
3. Mobile Coffee Roastery: Set up a bike-powered coffee roaster and sell fresh, artisanal coffee directly to customers. Use the $100 for roasting equipment and marketing.

We can build on this and just go crazy. As you can see what can seem very hard for humans can be extremely easy for Gen AI. Part of this is the vastness of the training data, part of this is the hallucination and prediction ability and part of this is pure computational power.

Humans in the Loop

It’s not enough to think about divergence on its own. This is where we need to use another principle — to bring humans in the loop.

For the above ideas, what will suit our context? What can we build on?

Bicycle delivery service is already being utilised by delivery companies around the world. So that’s a practical one. Bicycle tours are the same and work in some areas.

Urban art tours can work in specific contexts for specific customers and can be a good side hustle. Human powered drone delivery is slightly silly but may work in an emergency context.

The ideal thing will be iterate on these, use the ones we like and build on this. This is how we bring the human into the loop.

Generative AI is also images and video. In this sense, imagining an idea becomes easy. We are visual animals. Seeing makes a whole lot of sense for us. What used to be extremely hard will become supremely easy now. Llama 3.1 to the rescue for this.

Human-Powered Drone Delivery:

Bike-Powered Food Truck:

What if?

The what if questions have become easier. What if I make the budget $1K, or add two people and $2k. Here are two examples from the two whatifs.

1. Bike-Powered Food Truck: Create a mobile food truck that utilises a bicycle for power generation (or at least as a unique selling point), serving locally sourced and sustainable dishes.
2. The "Bike-powered" Concert Series: Organise intimate, outdoor concerts featuring local musicians and performers, using a sound system powered by a generator attached to the bikes.

If innovation is divergence, AI for divergence is a great addition. And humans + Ai is a great step forward for innovation.

Tell me how you are using AI and where it has worked or not for you. If you are keen to learn more about adding AI to your work, reach out to me.