AI Isn't The Reason You're Not Creative Anymore
Obviously every single speaker at Denver Startup Week was legally required to mention AI at least once in their message (joke).
However, this is my thought process around the consistent AI vs Creatives clash.
We were already in a Creativity Crisis
A 2010 Newsweek cover story, "The Creativity Crisis." drew on research by psychologist Kyung Hee Kim, who analyzed about 300,000 scores on the Torrance Tests of Creative Thinking, the most widely used creativity measure since the 1960s. Scores had risen for decades, but started declining around 1990. The steepest drops were among young children (kindergarten to 6th grade) and in "originality" and "elaboration," meaning the ability to produce unusual ideas and develop them in detail.
The "efficiency culture" explanation
Anyone in corporate America is VERY aware of this.
The argument is that several forces pushed people toward optimizing and away from exploring over the same period:
Work. The early 1990s brought business process reengineering, lean management, Six Sigma, and later KPI dashboards and productivity software. These descend from Frederick Taylor's "scientific management" of 1911. The shared idea is to find the one best way, measure it, and eliminate waste. That works well for repeatable processes. But creative work is inefficient by nature: it needs dead ends, slack time, and ideas that don't pay off. An organization that treats all unmeasured time as waste squeezes out the conditions creativity needs. (Insert the Office episode where Dwight was tracking Jim's time at the office and listed socializing and lunch breaks as time wasters)
School. In the US, No Child Left Behind (2002) and the testing era that followed rewarded right answers on standardized tests. Recess, arts, and unstructured projects often got cut to make room for test prep. Kim and others point to this as a key cause of the childhood decline.
Attention. After smartphones arrived around 2007, idle moments like waiting in line or lying in bed stopped being idle. Neuroscience links mind-wandering (the brain's "default mode network") to making novel connections. Boredom used to be a free creative incubator, and it largely disappeared.
Taste. Algorithmic feeds reward what already performs well, which nudges creators toward the same formats, hooks, and aesthetics. Culture got more efficient at giving people what they already like. Aka - why every city now looks the same. In-and-Out isn't just a California treat, now I can go down the street in Colorado and get an animal style burger. Every tik-tok post has a link to an amazon item to purchase. And people stopped making weird art and now make art that makes money, is non offensive, and appeals to the masses.
The "malnourished" metaphor means that none of these forces killed creativity outright. They just steadily underfed the capacities it relies on: tolerating ambiguity, sustained daydreaming, finding problems rather than only solving them, and developing personal taste.
Why AI "exposes the deficit"
1. AI automates exactly what efficiency culture trained us for.
For decades, the rewarded skills were producing competent, formatted, on-template output quickly: the standard report, the best-practice slide, the correct answer. Generative AI is extremely good at that.
So workers whose value was "reliable, efficient output" are finding that this was the most automatable part of their job. What stays scarce is original framing, judgment, taste, and knowing which question to ask. Those are the capacities that got underfed, so the deficit becomes visible and economically painful.
2. AI can deepen the deficit if used as a shortcut.
Early research points this way, though it's still preliminary:
A 2024 study in Science Advances found that AI-generated ideas made individual short stories rated more creative, but made stories more similar to one another. Each person improved while the group became more homogeneous.
A 2025 MIT Media Lab study on essay writing with ChatGPT found lower brain engagement and weaker recall of one's own work. The authors called this "cognitive debt."
The worry is that AI becomes the ultimate efficiency tool, skipping the messy middle of thinking where original ideas actually form.
Daniel Pink posted this week on LinkedIn that he hand writes all of his first drafts.
Because "writing is how I think".
Which completely makes sense. Writing is how I process and think. It's why even in grad school I took notes by hand instead of typing them. Somehow the physical act of writing forces my brain to work harder for each word.
The core claim: AI as an amplifier, not an equalizer
Creativity researcher James C. Kaufman's recent work argues that AI doesn't replace human creativity but can amplify existing differences in skill, judgment, and expertise.
people with higher creativity and intelligence benefit more from AI, and overreliance may hinder skill development and learning. In his words, "AI doesn't suddenly make everyone equally creative."
A creative task with AI has three stages:
Framing - What problem are you actually trying to solve, and what would a good answer look like?
Generating - Producing options. This is where AI is very strong.
Selecting and refining - Recognizing which output is good, which is generic, and how to push it further.
AI mostly speeds up stage 2. Stages 1 and 3 depend on taste, domain knowledge, and judgment, which are the capacities the previous answer described as underfed. So a person with a sharp question and a good eye gets a lot of useful material and picks the best of it. A person with a vague question gets a lot of plausible, average material and can't tell it apart from the good stuff.
During Denver Startup Week, there were a number of speakers and companies discussing Quantum Field Theory in Curved Spacetime (QFTCST) as applied to Fault-Tolerant Quantum Information Processing.
I had ZERO idea what they were talking about.
They might as well have been speaking Japanese.
I asked AI to explain it to me and it did.
This isn't my area of expertise - so honestly I have ZERO idea if AI did a good job or just confused me further. I have to talk to someone in the actual field or find a researcher to find the truth. I have no way of knowing if AI is giving me an accurate answer.
The evidence
Visual creativity. A March 2026 study in Advanced Science compared artists, non-artists, and an image-generation model. Human artists outperformed the AI overall, but structured human guidance substantially improved the AI's outputs. The researchers tested this directly by prompting the AI under two conditions: one with a lot of human input and one where it guided itself. The quality of what humans put in showed up clearly in what came out. Wiley Online LibraryWiley Online Library
Workplace field evidence. In a telemarketing field experiment by Jia and colleagues, AI assistance increased employee creativity on average, which raised sales, but the effect was strongly skill-biased. A later analysis of that study concluded that abundant AI output without matching expertise helped those already equipped to use it, while adding pressure on those who weren't. Lower-skilled agents received no differentiated scaffolding or extra training to handle the added complexity. Harnessing Abundance: A Generativity Perspective on Human-GenAI Collaboration +2
The ceiling of raw AI creativity. A study of more than 100,000 people, published in Scientific Reports in January 2026, found that generative AI can now beat the average human on certain creativity tests. However, the most creative humans still outperformed the models, and that gap was even larger among the top 10 percent of creative individuals. AI produces a strong average. What rises above that average still comes from people. ScienceDailyScienceDaily
Sure, On routine, well-defined tasks, AI often levels the field.
The best way to reconcile these findings is that it depends on the task. When "good" is well-defined. AI gives everyone the expert playbook and narrows the gap. When "good" is open-ended, such as original ideas, persuasion, or art direction, judgment becomes the bottleneck and AI widens the gap.
BUT Amplification can turn into atrophy. A mirror only reflects what you bring, and if AI does all your framing and selecting, you gradually bring less. This is also where it connects back to the creativity-crisis story: AI raises the value of human-centric capabilities while making them harder to develop.
How to Be Creative and Use AI
People usually use AI go generate ideas. But Creative freeze is rarely a lack of ideas. It's usually a threat response. Your brain treats a blank page, an open-ended question, or a chance to be judged as a risk, and the safest move is to do nothing or do the expected thing. (Thank you three decades of efficiency culture for this reflex: get the right answer, avoid waste, don't be wrong in public. AI adds a new version of the freeze: when the tool can produce something instantly, why risk your own attempt?)
So getting unstuck isn't mainly about idea-generation tricks. It's about three layers: what you believe about yourself, the state you're in, and what you actually do.
Lever 1: Belief — creative self-efficacy
What it is: Your confidence that you can produce something original in this situation. Researchers call it creative self-efficacy - one of the strongest predictors of whether people attempt creative work at all.
Why it matters for freezing: If you believe "I'm not a creative person," every open-ended task feels like a test you'll fail, so you avoid it or default to a template. People with low creative confidence hand the framing to AI, and their skills weaken further.
Fight it.
Redefine the scale. Kaufman and Beghetto's "Four C" model separates big-C creativity (genius-level), Pro-c (professional), little-c (everyday), and mini-c (personal insights that are new to you). Most people freeze because they measure themselves against big-C. Creativity in daily work is little-c: a better way to run a meeting, a sharper framing of a problem, a new recipe.
Treat creativity as learnable. Research on "creative mindsets" finds that people who see creativity as a skill that grows persist longer after setbacks than people who see it as a fixed trait.
Collect small mastery experiences. Bandura's strongest source of self-efficacy is doing the thing and succeeding, even at small scale. Confidence tends to follow action rather than precede it.
If you really want to nerd out on this lever - I highly recommend reading The War of Art. It will kick your butt into gear.
Lever 2: State — safety and approach motivation
What it is: The emotional and cognitive state you're working from. The key distinction is between approach motivation (moving toward something you want) and avoidance motivation (moving away from failure or judgment).
What the research shows:
Being watched and judged suppresses creativity. Teresa Amabile's experiments from the late 1970s onward showed that expecting evaluation lowered the creativity of people's work, especially their originality. (aka maybe turn your social media account to private so you aren't worried about what friends and family think of your weird art)
Creative cognition involves two modes working together. Neuroscience research, notably Roger Beaty's work, shows creative thinking involves cooperation between the brain's default mode network (spontaneous, associative idea generation) and its executive control network (evaluating and directing). Freeze often happens when the evaluating mode takes over too early and blocks generation. This is why leader Jim Collins (and many many more) do walking meetings or schedule time to be unproductive. It's why most of your best ideas come to you when you're in the shower or working out or painting .
Fight It:
Separate generating from judging. Spend time producing ideas without evaluating them.
Lower the stakes of early attempts. Label work as "draft 0," "sketch," or "for my eyes only." You're removing the evaluation threat.
Protect some idle time. The mind-wandering the default mode network relies on needs unstructured moments. aka the boredom that phones took away.
Lever 3: Behavior — agency through small, repeated action
What it is: Turning intention into a repeated cycle of making, testing, and adjusting. This is where "active creative agency" actually happens.
What the research shows:
Quantity drives quality. Dean Simonton's research on creative careers found an "equal-odds rule": a creator's hits appear roughly in proportion to their total output. Originality is largely a numbers game played over time.
Specific plans beat intentions. Peter Gollwitzer's research on "implementation intentions" (if-then plans like "When I open my laptop at 9, I'll sketch three rough options before checking email") shows these considerably improve follow-through compared with general goals.
Progress fuels motivation. Amabile's later diary studies of knowledge workers found that making progress on meaningful work was the biggest driver of positive inner work life. Small, visible wins. (this is why my approach of lets change everything all at one time usually fails within 24 minutes and my dad's approach of consistent, daily progress infuriatingly wins in the long run)
Constraints often help. (David Epstein has en entire book on this coming out called Constraints) A blank page is paralyzing; "three ideas in ten minutes using only these materials" isn't. Many studies show moderate constraints improve creative output by narrowing the search space.
Fight It:
Start before you feel ready. Make the smallest possible version: a one-line thesis, a thumbnail sketch, three bad options.
Set output goals, not quality goals. "Ten headlines" rather than "a great headline."
Ship small and get feedback. Each cycle builds skill (improving Lever 1) and makes judgment feel less threatening (improving Lever 2).
How the three levers interact
These three levers reinforce each other. Action creates small wins, which build self-efficacy. Stronger self-efficacy lowers the sense of threat. A safer state makes the next action easier. The freeze is the same cycle running in reverse: low confidence leads to a sense of threat, which leads to avoidance, which leads to no wins, which leads to even lower confidence.
That's also why you can start from any lever. If your confidence is the weak point, start with tiny actions. If anxiety is the problem, fix the conditions first. If you keep planning without making, commit to an output quota.
Connecting it back to AI
Each lever corresponds to part of what you bring to the tool:
Belief determines whether you frame the problem yourself or give the framing away.
State determines whether you explore AI's options with curiosity or grab the first acceptable answer to avoid thinking.
Behavior determines whether you keep practicing your own craft or let it weaken.
A simple practice combining all three: before any AI-assisted task, spend five minutes writing your own rough take, without judging it (Lever 2) and focusing on quantity over quality (Lever 3). Then use the AI to challenge and expand it. Over time, those attempts build the creative confidence (Lever 1) that makes the next input to the "mirror" stronger.
Do the framing yourself first.
Write down what you actually think, what you want, and what "great" would look like before you open the tool. A specific, opinionated input gets a specific output; a vague input gets a generic one.
Use AI to widen your options, then choose with your own judgment.
Ask for many variations, then choose, combine, and reject them yourself. The selecting is the creative act.
Protect some unassisted practice.
Keep a share of first drafts, sketches, and brainstorms AI-free so your input quality keeps improving rather than declining.
Treat taste as the skill to invest in.
Reading widely, studying excellent work in your field, and getting honest feedback all make the "reflection" better.
In one sentence: the research supports the idea that AI amplifies human input on open-ended creative work, but not on every task, and not permanently if you let it do your thinking for you. perfect example? me practicing this by yapping on this blog :)
If you want to use the little checklist I created ( along with AI obviously) - feel free to steal the one I made.
If AI can help me create something good in the world that didn't exist before - I'm okay with it.
Comments