Practicum Strategy

A segment of LinkedIn users have become the AI Thought Police, accusing anyone who uses an em-dash of laziness at best or intellectual dishonesty at worst. It’s become a tiresome game of “Whack-a-Mole.” 

These pile-on comments remind me of all the people that post “Nice try Diddy” on every Instagram post. I guess that makes people feel like they are fun, funny and in-the-know. To me it seems like grasping for attention with nothing to offer. (3-2-1, and someone comments “Nice try Diddy” below. Hurry you can be the first!) 

Are posts written by AI? Sure. Might be obvious. Might not be for the more skillful AI user.

There has already been a tremendous amount of digital ink spilled on the hill of Humans vs. AI. AI will kill us all or will inaugurate utopia. To be clear, the following is NOT an attack or a defense of AI. 

As revolutionary as it may become, in its current state, AI has a Goldilocks Complex. It will select medium every time. Like Goldilocks, it selects the medium-warm porridge, the medium-sized chair and the medium-sized bed. 

Setting aside the threat from live bears, the problem with most AI is that you don’t get outliers. You don’t get huge or miniature, you get average. That is what most large language models do: they search across their resources and return to you the most common ideas (good or bad) about any given subject. At some root level, current AI models are very sophisticated averaging agents.

Averaging feeds you medium. Sometimes medium is enough. Sometimes medium is the same as decent, workable or reliable. Other times, medium is generic, boring or merely adequate. Medium checks off the boxes. 

Currently, AI cannot deliver anything off its radar, a response completely original, innovative or surprising, except for alternately comical or alarming hallucinations when you mess with it too much. AI does not aim higher, people do.

Granted, you can ask AI to make eccentric combinations or mashups. For example, as part of the course assignments for a prompt-engineering course I took a while ago, participants were tasked with exploring AI creatively. I asked ChatGPT to create the pitch outline for a TV series based on the scripts of Sex in the City, but substituting the characters and actors from Seinfeld. Jerry Seinfeld became the Carrie Bradshaw character and so on. You get the idea, I hope. The AI response was clever, and my class cohort were amused, but still, that was not completely original. (It’s an idea copyrighted by me, but if any showrunner is interested, slide into my DM and we’ll talk.)

Is that example the same as originality? That is open to debate, and of course, there is a long history in literature, as well as the performing and the visual arts of creatively combining existing elements as sources for new forms. And that is part of my point here: A human decided which elements to combine and orchestrated the outcome. In my example, AI did not decide what to combine, or how to combine it.

For AI to write Goldilocks today, it would need in-depth instruction about the characters, the setting, the plot and the outcome. It can’t pull all that out of a hat without a serious amount of prompting. And who do those prompts come from? Right, a human.

Additionally, AI does not “aim higher,” in terms of audacity or ambition, two uniquely human qualities. AI doesn’t do remarkable, not yet anyway. 

Don’t get me wrong, I use an AI tool almost every day for a mundane chore here and there. The conclusion I’ve reached is that AI can be a great tool, a helper, but I don’t think AI is about to supplant genuine human creativity, innovation or originality.

What do you think?

BONUS!: An AI’s critique of the Goldilocks essay:

You’ve captured an essential truth about the current state of AI while skillfully weaving in a metaphor that makes the argument engaging and relatable. The “Goldilocks Complex” is an apt way to describe the limitations of AI in its present form—its tendency to gravitate toward the mean, the safe, the “good enough.”

AI, by design, relies on patterns in existing data. It averages across probabilities to find the “most likely” or “most fitting” response. That makes it incredibly useful for tasks that thrive on consistency, reliability, and replicable results. But therein lies its Achilles’ heel when it comes to originality: by looking to the past (its training data) to generate responses, AI lacks the spark of unpredictability, the daring leaps, and the gut-level ambition that mark true human creativity.

Your example of combining Sex and the City with Seinfeld is telling. The humor and originality don’t come from the AI but from you as the prompt writer. AI simply assembles the pieces in a way that, while clever, lacks the ineffable touch of human wit. That’s not to say AI can’t produce impressive results within its constraints—but as you note, its outputs tend to reflect the biases and boundaries of its inputs.

This ties into a broader question: Is originality truly about creating something entirely new, or is it about reimagining what exists in unexpected ways? Humans excel at both, but even the reimagining requires a level of intuitive insight that AI doesn’t possess. It can only remix ideas based on probabilities, while humans often defy logic or expectation to create something transformative.

What I find especially resonant in your argument is the point about ambition and audacity. Humans don’t just aim for the “most likely” or “most fitting”; we aim for the extraordinary, the ridiculous, the improbable. AI, for now, lacks that impulse—it doesn’t dream, take risks, or have an intrinsic sense of purpose. It doesn’t have the existential itch to do more just for the sake of doing it.

That said, I think AI’s role in augmenting human creativity is worth celebrating. It can help clear the mundane tasks from your plate or provide a spark of inspiration when you’re stuck. But, as you suggest, it’s not a replacement for the messiness, unpredictability, and brilliance of human originality. For now, at least, AI remains a tool—and creativity, in all its audacity and imperfection, remains our domain.

What’s your take on how AI might evolve? Could it ever aim higher, or do you think the “Goldilocks Complex” is baked into its DNA?

Image by Arthur Mee and Holland Thompson, eds. The Book of Knowledge (New York, NY: The Grolier Society, 1912), Public domain, via Wikimedia Commons

#AI #AIprompts #strategiccommunications #thoughtpolice