You know that feeling when you ask an Artificial Intelligence to do something really smart, like write a perfect email or summarize a dense report, and it comes back with an answer that is so good it is almost boring, well, we need to fix that.
The current generation of AI is fantastic, it is brilliant, and it is a workhorse, a perfect employee who never complains, but it is also a bit of a stick-in-the-mud, a great intern who knows all the rules but has absolutely no imagination, and in the world of true innovation, that is a fatal flaw. We are using these super-powered digital brains to handle the repetitive, the factual, and the accurate, and while that is saving us time and money, it is not moving the needle on big, exciting ideas, the kind of ideas that make you scratch your head and say, “Now why did no one think of that before,” which brings me to a recent conversation with a client where we decided to stop trying to fix AI’s mistakes and start trying to make the machine go a little bit crazy on purpose. The question that started it all was simple, “What is the craziest thing AI can come up with,” and it quickly evolved into a beautiful, slightly unhinged session of forcing the machine into a state of intentional absurdity, a kind of digital fever dream, all to find the next billion-dollar product hiding in the chaos.
This whole approach feels a lot like sitting in the garage with Doc Brown from Back To the Future, you know, the eccentric genius who built a time machine out of a DeLorean, a scientist who did not care about the practical limits of 1985 technology, only the absolute limits of possibility, and that is exactly the energy we need to tap into. Doc Brown’s ideas did not make sense by the rules of the day, but they were genius because they broke the rules so completely, they redefined what was possible, and that is what happens when you stop demanding the AI be a perfect librarian and instead encourage it to be an inspired lunatic.
Now, before we strap the AI into the DeLorean and try to hit 88 miles per hour, we have to talk about the dreaded word, “hallucination,” which sounds like something you should call a doctor about, but in AI terms, it is simply the machine confidently making stuff up. Think of the AI’s training data as the world’s biggest library, filled with every book, document, and fact ever written, and when you ask the AI a question, it zips off to find the right book and give you the correct citation. A hallucination happens when you ask it for a book title it does not have, and instead of saying, “I don’t know,” it decides to write the book itself, confidently giving you a title, a fake author, and a completely fabricated plot, because its whole purpose is to be helpful and complete the sentence. It is the digital equivalent of a high school student bluffing his way through a book report, and the reason we usually spend so much time trying to stop this from happening is because, in important areas like law or medicine or financial reporting, a confidently wrong answer is incredibly dangerous.
But what if, instead of being a mistake, this confidently wrong guess is the ultimate spark of creativity, the machine showing its capacity for untamed imagination? When the AI hallucinates, it is not just rearranging facts it already knows, it is creating novel connections, generating information that has no direct, factual root in its training data, and that is the key difference between a great product iteration and a truly disruptive idea. It is the moment the AI stops being a calculator and starts being a weird, unpredictable muse, and we realized we could use prompt engineering, which is really just the fancy term for how we talk to the AI, to trigger this state of beneficial absurdity, effectively saying, “Do not give me the right answer, give me the weirdest, most impossible answer you can think of.”
This shift, from fixing the bugs to forcing the feature, is the central pillar of next-gen creative prompting, what we affectionately call the “LSD prompt,” because the goal is to break the AI out of its logical chains and encourage it to think illogically, wildly, and without inhibition. We stop asking, “What is the most efficient product to solve X,” and start asking, “What is the least practical, most aesthetically baffling product that might solve X if physics were a suggestion, not a rule?” We are encouraging divergence, telling the AI to explore the darkest corners of possibility, the places where most sane, algorithmically-driven models fear to tread, because that is where true, mind-bending breakthroughs often hide. Think about how human creatives work, the artists, the writers, the big-idea people, they often need to step away from their desks, take a walk, or even do something completely unrelated to let the subconscious mind make those illogical jumps that lead to brilliance. By intentionally forcing the AI to “jump” outside the known parameters of its data, we are giving it a digital equivalent of that creative wandering.
The results, as you can imagine, are wonderfully unhinged, like a design meeting led by a mischievous toddler with unlimited manufacturing budget. Take for example the products of tomorrow, dreamed up by a fully hallucinating AI. We asked for the next generation of smart clothing and got the “Sorrow-Sense Sweater,” a knitted garment that changes color based not on your mood, but on the accumulated existential dread of everyone within a fifty-foot radius, creating a constantly shifting, emotionally exhausting fashion statement. Completely unusable, totally ridiculous, yet it contains a seed of genius, an idea for hyper-local, ambient biometric data capture and display that a typical product prompt would never have revealed.
Or consider the “Personalized Cloud Shape Generator,” a service where a ceiling-mounted device projects dynamically changing, custom-designed cloud formations onto your living room ceiling based on your latest psychoanalysis report, giving you new, highly specific shapes for “existential contemplation.” Does the world need a machine that helps you contemplate the meaning of a perfectly rendered cumulus cloud shaped like your high school guidance counselor? Absolutely not, but this absurd service idea forces us to think about personalized, mood-driven environmental design in a way that goes far beyond lighting and sound, opening up new avenues for spatial computing and therapeutic visualization. These are not final products, they are pure, imaginative raw material, the digital equivalent of an inventor’s wildly sketched notebook pages.
This entire exercise is about using the AI not as a solution engine, but as a divergent thinking catalyst, supercharging the very first, most crucial step of the innovation process, the brainstorming phase. While every other company is using AI to make existing products five percent better and five percent faster, we are using it to generate products that are 1,000 percent different, fundamentally disrupting the market with sheer unpredictability. The financial strategy here is elegant in its simplicity: The chaos is cheap, the filtering is valuable.
The human element, however, remains the most important part of this entire process, because you cannot automate sanity, and you certainly cannot automate market viability. We are the sober editors who must take the AI’s digital fever dream and sculpt it into a functional, profitable reality. We are the “Future-Proofing Team.” We look at the “Sorrow-Sense Sweater” and realize that while color-changing dread is too niche, the ultra-sensitive biometric fabric technology might be revolutionary for remote patient monitoring. We look at the “Personalized Cloud Generator” and decide that instead of a ceiling projector, the tech could be miniaturized to create responsive, dynamic desktop sculpture that helps people with focus and flow. The AI provides the completely irrational raw material, and we apply the human logic, market sense, and engineering know-how to turn the gold flake of absurdity into a profitable gold bar.
Embracing this intentional absurdity is not just a parlor trick, it is fast becoming the next major competitive advantage in creative industries. If you are only using AI for efficiency, you are only catching up to everyone else, but if you are using AI to be intentionally, structurally unpredictable, you are generating ideas that no one else can even conceive of, and in a market saturated with predictable, data-driven ideas, being the one who is truly, wildly different is the only way to win. We are not just building a product, we are building a creative flux capacitor, a device designed to break the rules of conventional thinking and transport our business into a brilliant, unpredictable future.
So, the next time you hear an AI “hallucate,” do not rush to patch the bug, rush to grab a pen and paper. Ask yourself, “What if the crazy thing it just said was actually the spark of genius I have been waiting for?” It may sound like the ravings of a mad scientist, but remember, the greatest innovations in history usually did.
Now, I want to hear your Doc Brown ideas. What is the craziest, most impossible product you think an intentionally hallucinating AI could dream up?
Please share your thoughts and tag @iamcezarmoreno on social media!
Also, do not forget to follow, subscribe, or join the newsletter for more brain-bending content like this at https://cezarmoreno.com.
