
The Clothes Have No Emperor: AI is Ready, but Humanity is Not
Author(s): Sophia Banton
Originally published on Towards AI.

βIt hallucinates too much.β
βI just used it to write an essay.β
βIt canβt reason.β
βI just made my Studio Ghibli image.β
These varied reactions underscore a fundamental truth: very few topics divide people as much as AI does. For some it evokes wonder, others fear, and among some disdain.
The Entitlement of Control
Human beings conveniently have short-term memory when it comes to technological innovation. Has anyone seen a VCR lately? Yet, if Netflix goes down even for a few minutes, the internet is flooded with commentary from βdissatisfied customersβ.
This sense of entitlement to own and rule technology isnβt limited to streaming platforms. ChatGPT is amazing as long as it βobeysβ, but if it asserts an idea, itβs suddenly hallucinating. I would argue that the obsession with hallucinating AI models has very little to do with technical benchmarks in percentage points and much to do with control. The βemperorβ could be our perceived need for control over technology, and the βclothesβ are the justifications we use (like βhallucinationsβ) to maintain that control.
The Hypocrisy of Language
But what exactly is an AI hallucination? A hallucination is defined as βa perception of something that is not present or a false sensory experience.β In clinical terms, itβs when the brain creates sensory input that doesnβt exist in reality. Ironically, many critics of AI warn us not to humanize AI; yet their most common pejorative for its βerrorsβ is a deeply human, clinical term. A hallucination is a human quality.
Mathematically, AI models donβt hallucinate. They arrive at predictive responses based on pattern matching and sometimes the patterns revealed are uncanny. But in that uncanniness, there can be brilliance, humor, or error. But donβt humans do the same? We donβt call it hallucinations when a human fumbles around trying to answer a question; rather we show intra-species level grace, perhaps saying βOh, you misspoke,β or βThatβs not quite right.β
Humans learn by making mistakes. AI learns by identifying patterns and refining predictions. Both make βerrorsβ or βunintended outputs.β But when the AI canβt be 100% accurate all the time, it is deemed a failure simply because it cannot be controlled.
Shifting the Goal Post
βIs it AGI (artificial general intelligence)?β
βNo, itβs not. AGI has to do this, this, and this.β
βWell Googleβs AI just made text to video outputs with integrated sound effects, voice and music.β
βOh well, to be βreal AGIβ, the AI has to do this, this, and now this.β
Itβs not about AGI vs not AGI, AI has long surpassed our cognitive abilities. The calculator beat us in math more than a century ago.
What Intelligence Actually Looks Like
AI has nothing left to prove. It can already do more than we ever imagined it wouldβ¦ and the real question now is βwhatβs left for us to do?β
I recently read a quote where someone said,
βAI is creating art and music while Iβm doing my laundry, but I want AI to do my laundry while I create art and music.β
Fundamental to that train of thought is the misalignment between what intelligence is and what humans expect machines to be. Art and music creation are expressions of intelligence, as are writing and of course solving mathematical problems. While the public may have expected automation and industry may have hoped for digital laborers, researchers never set out to build artificial brute force workers; they set out to build artificially intelligent systemsβ¦ and thatβs what we did.
So yes, AI is creating art, music and so much more. The challenge for us now is to use AI tools to create new ideas and find novel ways to express ourselves as photographers did with cameras and cinematographers did with video and film. These tools didnβt just change existing work; they created entirely new industries, from commercial photography to Hollywood studios, spawning countless job roles that didnβt exist before. The possibilities are endless.
As for the laundry folding robot, itβs coming too. But remember humans, every machine has to be maintained and serviced. Not everything can and should be automated. Brooms and mops are still stocked in stores, Swiffers too, despite the advanced robotic vacuum cleaners and mops available for purchase. Why? Because sometimes automation isnβt necessary. Or to put it another way, sometimes you will need to sweep the corners of your house and sometimes you will need to fold your jeans.
A Tale of Two Generations
If we step back as a species we have two options. We can mourn our displacement as the βtop of the food chainβ or welcome company at the top.
We werenβt ready. We werenβt ready to see ChatGPT write an essay in milliseconds. We werenβt ready for image generators to create art in the styles of the masters, or for AI music generators to compose entire songs in seconds.
But the βweβ here doesnβt apply to βall of usβ, because our children were ready. Children of this generation donβt fear AI or mourn its arrival the way we do. The same way our thumbs adapted to typing, their cognition will adapt to having AI collaborators. They wonβt debate about whether prompt engineering is an essential skill set, they will just talk to AI. They wonβt feel unsure about using AI-generated music, they will just choose the songs they like. I have seen this in my own life. The children are ready. Us adults are not.
You know zoos were always a source of contention. Theyβre great for educational purposes some argued β¦ while others said the animals didnβt deserve to be held captive. Well today a child can put on a VR headset and see a lion in its natural habitat, use AI image generators to draw them, use AI video generators to create short films of them in the wild, while we look on and question our entire existence.
Because unlike the child we consider the job displacement behind those activities: no artist, no videographer, no sound engineer. But instead, what we have to realize is that there are new jobs behind those images, videos and experiences: AI prompt engineers, AI ethicists, data curators, AI-powered content strategists, and roles we havenβt even named yet. Itβs just that most of us arenβt trained to do them.
Getting Ready
So rather than continue to say AI isnβt ready using weak arguments that vanish like morning dew, we do as children do, and we get ready. We get ready by realizing that control isnβt necessary to thrive. The best parts of us were never the consequence of being βcognitively dominantβ on our planet, but what we built as part of our universe.
Now go prompt, speak, or simply engage with AI.
About the Author
Sophia Banton is an Associate Director and AI Solution Lead in biopharma, specializing in Responsible AI governance, workplace AI adoption, and building and scaling AI solutions across IT and business functions.
With a background in bioinformatics, public health, and data science, she brings an interdisciplinary lens to AI implementation β balancing technical execution, ethical design, and business alignment in highly regulated environments. Her writing explores the real-world impact of AI beyond theory, helping organizations adopt AI responsibly and sustainably.
Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming aΒ sponsor.
Published via Towards AI