Centaur vs. Cyborg: Which Creative User of AI Are You?
Exploring the Dynamics of Human-AI Collaboration in the Creative Realm: Strategies, Pitfalls, and Triumphs.
The rapid evolution of AI tools, especially large language models like ChatGPT, offers a wealth of potential for creative professionals. Yet, with potential comes complexity. The uneven capabilities of AI have ushered in a “jagged frontier,” challenging us to discern where AI can enhance our work and where it might trip us up.
The Jagged Landscape of AI: Opportunities and Challenges
A study with consultants at Boston Consulting Group1 demonstrated that AI excels at creative tasks (like developing new products), increasing productivity by over 40%. On the flip side, an over-reliance on AI decreases performance for things like strategic analysis, a task designed to be outside AI's capabilities.
The frontier is “jagged” because AI's abilities are uneven across knowledge domains and task types. If you’ve ever asked ChatGPT to do a word count, you’ll know that what appears to be an easy task might actually be quite challenging for an AI system. As the study notes:
"The shape and position of the frontier are vital to understanding the impact of AI on work...Tasks that appear to be of similar difficulty may either be performed better or worse by humans using AI."
This ambiguity around what falls inside or outside the frontier poses pitfalls for creative professionals. AI may fail in unexpected ways or produce plausible but incorrect outputs (confident hallucinations). However, when used for precisely the right tasks in a specific manner, it can reap untold benefits.
Are You a Centaur or a Cyborg?
When using AI in your creative workflow, which archetype resonates more with you, the Centaur or the Cyborg?
Centaurs strategically divide their work between the human and artificial realms based on relative strengths. They may leverage AI for writing while reserving human judgment for analysis. Centaurs actively determine what falls inside the frontier. They rely on experience and judgment to determine when AI augmentation makes sense. They know their strengths and when to leverage human versus machine abilities.
Cyborgs tightly integrate with AI, combining capabilities even at a subtask level. They continually experiment with AI by assigning customized personas or requesting detailed editorial changes. Cyborgs represent complete human-AI synthesis. They have an experimental mindset. They push AI capabilities forward but are limited by the technology's current state. This requires tenacity and continuous problem-solving.
Most professionals will oscillate between these roles based on their projects' demands. But as AI continues to evolve at break-neck speed, understanding when to use which approach will be a vital skill.
Best Practices: Traversing the Jagged Frontier
The ambiguity of the AI frontier requires a careful approach. Here are some strategies:
Embrace the Unknown: The AI frontier is ever-expanding. Abandon preconceived notions and approach each task with fresh eyes. “Don’t be the ass in assumption.”
Human Judgment is Irreplaceable: While AI can aid in various tasks, critical thinking remains an inherently human domain. Use AI as a tool, not a replacement.
Develop Frontier Navigation Skills: Learn to recognize when AI can enhance your work and when it might mislead. This skill will help you avoid pitfalls where AI may overextend or underperform.
For instance, if you're hoping to see a rendered image of a cyborg alongside a centaur, AI might not be up to the task just yet (but hey, prove me wrong in the comments!).
Both Midjourney (above) and DALL·E could not wrap their pretty little algorithms around the idea of a Centaur and a Cyborg in the same image.
Final Thoughts
Creative professionals are at the forefront of AI's innovative potential. While the journey may present challenges, those willing to experiment, learn, and adapt will shape the future.
Just remember:
As a Cyborg, you never want to be more robot than human. It’s important to know your intent and retain your agency.
As a Centaur, when you toggle back and forth between tasks, it’s more enjoyable to be the human torso than the horse’s ass.
Be careful out there!
For more, check out my 7 Habits for Highly Effective Prompting and learn about integrating Conversational AI into your creative practice!
This material has been presented to companies (Skilshare, We The Collective), creative agencies (Design Agency, 123w), and universities (George Brown, University of Toronto). Here’s what people are saying:
This is like flipping creativity on its head and finding a benefit that no one else is thinking about in education. — Jenna Marinucci Cebo, Curriculum & Learning Innovation Coordinator, Ontario Institute for Studies in Education, University of Toronto
Jason is a generous instructor and his step-by-step playbook is your takeaway tool and starting point for harnessing the power of this brainstorming/creating accelerator on your own. You'll be glad Jason taught you how to fish these new waters. — Lynne Valeriote, UX/Content Strategist, Copywriter, Consultant
Jason unlocked a new superpower for my company: the ability to aim and focus the AI to my specific advantage. — Israel Diaz, Chief Creative Officer, Sunday+Night
If you’re interested in educating your own company or creative team, let me know.
Dell'Acqua, Fabrizio, Edward McFowland III, Ethan Mollick, Hila Lifshitz-Assaf, Katherine C. Kellogg, Saran Rajendran, Lisa Krayer, François Candelon, and Karim R. Lakhani. "Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality." Harvard Business School Working Paper, No. 24-013, September 2023.
The Cyborg and Centaur analogies hit the nail on the head – it's all about striking a balance between what machines can do and our own creative mojo. Giddyup.
Hopefully, we'll use a tamed horse to build our centaur… This is the main danger, I think, relative to the halting problem. Do the horse know when to stop? Why would he stop? The Jewish culture defines the existence of an ideal point of equilibrium between chutzpah & shame. What if we build a machine with a super-chutzpah & zero shame? It seems that we are already on this path. Even if the pure logic of a situation may produce some “ideal” solution, like the infamous “final solution” enforced by the Nazi mob during WW2, non-psychopathic humans could anytime ask themselves : am I not going too far? Such feeling is rooted in non-logical ethical values & may supersede the mind, using simply our “common sense”. In all the examples I heard of blatant errors committed by some AI, what seems to be always lacking is the most basic “common sense”. How can we code that? Is it possible to code that?