The Age of AIdeation

The frequent obsession with measuring AI-fueled productivity misses much of the true value of generative AI and related technologies (genAI). Sure, you can (and should) automate a range of tasks you perform, such as auto-generating marketing email messages. (Remember “macros”?) But once you’ve found a way to speed up or offload a number of activities, where will you continue to get value from these new tools? 

The answer: Ideation powered by AI. AIdeation.

Any focus on productivity makes economists happy, who think the only road to economic growth is if we workers continue to do more in the same amount of time. But most workers I know don’t wake up each morning and wonder how they can work even harder than they already were. Sure, we all want to have more time in our work days. But isn’t it a greater joy to come up with some creative solution to a problem that’s had you stumped for a long time?

Prompting a generative AI program — text, images, presentations, video — lets you see a range of ideas you might never have come up with. That doesn’t mean you necessarily need to use the output. But sparking new thoughts, new avenues, new insights — that’s AIdeation.

Suppose you have an idea for a startup. Ideally, that thesis would include a problem to be solved (say, deforestation), a possible solution (using drones), and a business model (somebody has to pay for it, even if it’s a non-profit). You go to zigzag, plug in your one-line thesis, and in about two minutes out pops a draft version of a business prototype, including a "lean canvas" illuminating the problem, the solution, key business metrics, and potential types of customers. It even lists existing alternatives — which includes the first company I know of to build exactly this business, Dendra Systems (born as BioCarbon Engineering, in Singularity University's former incubator program, where I was a mentor). You wouldn’t take the output from zigzag and pop it immediately into your VC pitch deck, but you’d have useful input to help structure your thinking.

What about Group AIdeation? Suppose you’re co-hosting a brainstorming session with a dozen or so collaborators. You’ve written down everyone’s ideas on big PostIt board paper on the walls, and now it’s time to come up with the group’s next steps. You type all of the group's input into Anthropic's Claude2, and you ask Claude: What are the three most important steps this group needs to do Next? (I did this in a recent brainstorming session about Social-Emotional Learning for organizations.) Sure, I could have done that synthesis myself, but it would have been affected by my human bias. Sure, genAI models are infused with the same human biases — but in this case the software identified several patterns I hadn’t seen.

As a recovering journalist, I can tell you that for me, AIdeation is also a great tool for solving the Blank Page Problem. Simply asking ChatGPT to give you 10 riffs on an idea, or prompting Midjourney to generate a few dozen variations on an image, can catalyze a range of new thoughts. Upload a few pages of your own unstructured ramblings, and you can ask Claude to suggest a distillation of the themes it detects.

But while AI software is creating new content, it's not true creativity. Humans ideate, AI software iterates. A recent UC Berkeley study comparing humans (including young humans) and genAI software showed that the applications did much better with knowledge-based tasks, but humans were better at coming up with out-of-the-box creative answers. (In a few years, though, we will likely have to re-visit this equation.)

AIdeation has little to do with traditional notions of productivity. Research suggests that innovation often comes from the ideation rate in organizations. The more ideas you can generate — funneled into team processes where people sift through and workshop ideas together — the more innovation results. That’s supported by a recent study in the National Academy of Sciences journal PNAS. Reviewing over 4 million digital artworks from over 50,000 AI-assisted creators, there was a 50% increase in output the first month using the tools, and double that in the second month. And approval ratings for their works also increased.

PNAS calls this "generative synesthesia": Using genAI to spark greater human expression. 

I’ve watched this process countless times with our NextCoLabs.io community, guided by my co-founder Michael Morrissey: Creative professionals and business strategists from around the world like Roy C. Vella , Brad Rochefort , and Doug Hohulin come together three times a week to spark off each others’ ideas using the latest tools. The focus isn’t on productivity: It’s on creativity, group learning, and applying the tools to real-world problems.

And that’s why the lasting advantages of genAI programs will likely come from AIdeation. 

What Can You Do Next?

• Just start. Rather than staring at a blank piece of paper, you can start with any of the free genAI text tools, typing words, sentence fragments, dissimilar thoughts — and ask a genAI program what common ideas could be seen, or request all of those thoughts to be synthesized and distilled into a single thesis. Or into a dozen theses.
• Ask the software for help starting. State the problem you’re trying to solve. Ask for suggestions where to start. Prompt the genAI program: “You are an expert on innovation. What are other ways I should think about this problem?” Curt Carlson, former CEO of SRI International, famously ends many online posts with the question: “What am I missing?”
• Review, Edit, Synthesize, Re-synthesize — and Humanize. Keep shaping the content you create. The software doesn’t care how many times you ask it to do something differently, make big or tiny changes, or ask it an endless series of questions. But make it your own. Not only will software-generated errors inevitably creep in, but research shows that academics using genAI often include words that genAI programs favor (“meticulous”) — not only a dead giveaway they didn’t write it alone, but polluting our culture.
• Keep HitL — Humans in the Loop. Many of the AI tools shine best when melding text or images from a variety of disparate sources. This is a great use of AIdeation as a team sport: Each person brings a unique perspective, and while one human might not detect the common threads, a genAI program might. Just make sure you have human collaborators with whom you can also spark your creativity.

-gB Gary A. Bolles

I’m the author of The Next Rules of Work: The mindset, skillset, and toolset to lead your organization through uncertainty. I’m also the adjunct Chair for the Future of Work for Singularity University. I have nearly 1.5 million learners for my courses on LinkedIn Learning. I'm a partner in the consulting firm Charrette LLC. I'm a senior advisor to aca.so, building the Community Operating System for organizations. I’m the co-founder Next CoLabs , a global thinktank of AI creatives, and of eParachute.com. I'm an original founder of SoCap Global, and the former editorial director of 6 tech magazines. Learn more at gbolles.com