Learn to Manage AI Before It Manages You
Pragmatic Advice for Creative Professionals
The pace of AI is accelerating, transforming knowledge work. To stay competitive, creative professionals must proactively build real fluency with these tools before they become inscrutable black boxes. This article provides practical guidance on prompting techniques to steer generative AI while retaining creative autonomy. There’s also a quick plug for my course on collaborating with AI toward the end.
The AI Wave is Crashing The Creative Party👋🏻🌊🎈💥
Just last week, I consulted with a creative executive who was pitching for an exciting new project. In one afternoon, using AI tools, we accomplished what normally takes a seasoned creative team weeks:
A strategic approach from various stakeholder viewpoints
An analysis of the creative brief
Colour palettes and mood boards
Logo ideas and applications
A sales deck with persuasive imagery
Our experiment proved that with the right expertise, a few “AI Whisperers” could match a small agency’s output at 20x the speed.
The key was not just letting AI do the work, but actively directing it to augment our existing creative skills. We maintained intent and direction until we achieved the desired effect. For now, machines don’t have true creative desire—they still need human guidance.
But things are changing incredibly fast...
[AI] has started to display 'compounding exponential' properties…
[The] next few years will yield progress that intuitively feels nuts.
— Jack Clark, co-founder of Anthropic, the makers of the AI Claude
The Chess Player’s Request 👑♟️🌾♾️
The pace of advancement in AI is dizzying. While some claim we've reached an AI plateau, companies like Google, Microsoft Anthropic, and OpenAI make major announcements almost daily.
There's an ancient parable1 that vividly illustrates this exponential growth:
The story goes that a king granted the inventor of chess any wish. The game maker asked for a simple reward—just one grain of wheat doubled for every square on the chessboard.
The king laughed, knowing his silos overflowed with excess harvest. But the wish quickly bankrupted the kingdom. The grains doubled rapidly, exceeding all the wheat in the world!
Now imagine exponentials repeating exponentially. Forget grains of wheat—the numbers quickly outpace all the atoms in the known universe! Of course, there are real-world processor speed limits, but AI may eventually solve those, too2.
So while progress may seem incremental, we'll soon hit an acceleration point that “intuitively feels nuts”3 in scope and scale, as Anthropic’s Jack Clark colourfully describes it.
Now’s the Time to Become an AI Whisperer
Regardless of existing technical skills or lingering skepticism, all creative professionals need to start actively experimenting with AI now before these technologies get too far ahead of them.
The more you know how to communicate with these powerful tools, the more you can preserve your autonomy and manifest your intention4. Studies in the use of conversational AI have shown that knowledge workers experienced a dramatic productivity boost (66%!)5—even without training and while using dated algorithms from 2022.
Think about what you could achieve over the next few years with AI augmentation and some optimization training!
7 Habits for Highly Effective Prompting
Over the last year, I’ve codified the most effective prompting practices, which I use every single day. These practices strengthen critical thinking, elicit diverse perspectives, provide constructive feedback, and help me actively direct the AI tools toward my creative vision.
Format: Framing clear needs and parameters so AI understands the exact requirements for a response. What is being created? How long should it be? What are the specifications? Do you have any examples? Consider also Bloom’s Taxonomy6 and writing an effective “How Might We”7 challenge statement.
Chain-of-Thought8: Asking AI to detail its reasoning to strengthen logic and critical thinking, but also to see other jumping-off points to potentially pursue.
Role-Play9: By assigning different personas, creatives can elicit more specialized, tailored responses from AI. A writer could ask for feedback from an AI art critic to get an entirely different perspective. Or a designer could request UI suggestions from an AI user experience expert to get targeted insights.
Tree of Thoughts10: Prompt variations to elicit diverse perspectives, explore concepts broadly, and expand possibilities.
Reflexion11: Asking an AI to critique its own work provides a constructive reality check. Just like reflecting strengthens human creativity over time, forcing AI to self-reflect sharpens the quality and reasoning of its responses.
Evaluation: Creative pros can leverage AI as an impartial evaluator to dispassionately weigh options when ideating. When asked politely, the AI will create a Decision Matrix12 to help rate and rank concepts against various criteria.
Mimicry: Give the conversational AI your unique voice. Let it become your own personal ghostwriter. Have it interview you (I call this AME: Ask Me Everything) about your thoughts and ideas, then compile them into the most effective Format. (And now we’ve come full circle!)
As you can see, by applying these habits, the AI doesn’t come up with things. You come up with things. It hasn’t experienced what you’ve experienced, and therefore cannot make the same connections and inferences you can.
Conversational AI is a powerful accelerant, an incredible augmentation device, and an unmitigated co-creator when you want a quasi-sentient wall to bounce your thoughts and ideas off of.
—Jason Theodor, founder of More Better Different
In my practice, I am always directing and curating and shifting and shaping things. I am never comfortable with the default answer, the vanilla response. I don’t need to accept the median as the message, especially regarding AI. I can steer this ship in any direction I choose, and take a fraction of the time doing it.
Creative professionals should approach AI as a power tool: It requires training and practice for safe and effective use. Without developing real fluency, creatives risk replacement as AI advances: if not from the AI itself, than from those more proficient in its use. Through hands-on practice and mentored guidance, professionals can gain the skills needed to steer these tools with purpose, while retaining their creative autonomy.
The bottom line is evident – manage AI before it manages you. Don't leave your skills to chance. Start actively learning and experimenting with these technologies today.
My Course
Over the past six months, I’ve talked about creative AI augmentation and empowerment to companies like Skillshare, Sunday+Night, DesignAgency, We The Collective, and 123w. I’ve shared my work with students and staff at George Brown and the University of Toronto. I’ve presented these ideas at AI conferences like FITC’s Spotlight on AI. All the while, AI advancements keep shifting the earth beneath our feet.
If you, your company, or your team are interested in learning more about conversation AI and the best practices for creative professionals, check out my upcoming course or set up some time to chat.
My motto has always been do more, think better, be different—and with AI that is doubly exponentially so.
Wheat and chessboard problem. (2023, September 27). In Wikipedia. https://en.wikipedia.org/wiki/Wheat_and_chessboard_problem
Kim, T. (2023, March 22). Nvidia CEO Says AI Can Overcome the Death of Moore’s Law. Barron's. Retrieved 2023-10-23 from https://www.barrons.com/articles/apple-stock-iphone15-demand-621d5518
Spool, J. M. (2013, December 30). Design is the Rendering of Intent. UX Articles by Center Centre. Retrieved from https://articles.centercentre.com/design_rendering_intent
Nielsen, Jakob. (2023, July 16). AI Improves Employee Productivity by 66%. Nielsen Norman Group. Retrieved from https://www.nngroup.com/articles/ai-tools-productivity-gains/
“The three studies were very different and yet arrived at the same results. This vastly increases my faith in the conclusions. Any single study can be wrong or misleading for several reasons. But credibility shoots through the roof when different people find the same thing in different domains with different research methods. The lead authors of these three case studies were from Stanford, MIT, and Microsoft Research, respectively. They studied customer-support agents resolving customer inquiries, business professionals writing routine documents, and programmers coding an HTTP server.”
Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond (2023): Generative AI at Work. National Bureau of Economic Research working paper 31161. https://www.nber.org/papers/w31161
Shakked Noy and Whitney Zhang (2023): Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence. Available at SSRN: http://dx.doi.org/10.2139/ssrn.4375283
Sida Peng, Eirini Kalliamvakou, Peter Cihon, and Mert Demirer (2023): The Impact of AI on Developer Productivity: Evidence from GitHub Copilot. Available at Arxiv: https://arxiv.org/abs/2302.06590
Bloom's taxonomy. (2023, October 16). In Wikipedia. https://en.wikipedia.org/wiki/Bloom%27s_taxonomy
How Might We. Ideo Design Kit. Retrieved 2023-10-30 from https://www.designkit.org/methods/how-might-we.html
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q., & Zhou, D. (2022). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. ArXiv. /abs/2201.11903
Maurya, R. K. (2023, September 9). A Qualitative Content Analysis of ChatGPT's Client Simulation Role Play for Practicing Counseling Skills. https://doi.org/10.31234/osf.io/vwuxh
Yao, S., Yu, D., Zhao, J., Shafran, I., Griffiths, T. L., Cao, Y., & Narasimhan, K. (2023). Tree of Thoughts: Deliberate Problem Solving with Large Language Models. ArXiv. /abs/2305.10601
Shinn, N., Cassano, F., Berman, E., Gopinath, A., Narasimhan, K., & Yao, S. (2023). Reflexion: Language Agents with Verbal Reinforcement Learning. ArXiv. /abs/2303.11366
What Is A Decision Matrix? American Society for Quality. Adapted from The Quality Toolbox, Second Edition, ASQ Quality Press. Retrieved 2023-10-30 from https:// asq.org/quality-resources/decision-matrix
Really useful and practical summation Jason!