A machine does not create or hallucinate. It fabricates.
Illustration © 2026 AdventureLAB

When it comes to AI, there is a distinct lack of common language and understanding. Conferences, webinars, articles, and the crescendo of announcements and claims from AI experts and consultants may all have good thinking behind them, but many still suffer from disconnection, taking points from and speaking to closed, siloed groups. Across those groups, the AI rhetoric and narrative often differ between sectors, such as creative industry professionals and academics, and this pattern is repeated within each sector's sub-sections.

Without shared points of reference, the alignment between sectors, which could advance the understanding and mindful application of new technologies, breaks down, leaving gaps and asynchronicity. In this stage, confusion and misunderstandings replace shared perspective and common sense.

Generative AI misconceptions

For starters, narrowing the debate, most mainstream attention goes to the Generative AI – Gen AI – either as stand-alone platforms, Midjourney, ElevenLabs, Runway, Stable Diffusion, chatbots, ChatGPT, Gemini or the integration of AI in existing software like Adobe Creative Suite. Presented generated texts, images, music, video, speech, etc., based on data of sometimes questionable origin, such as whether something is used with the consent of its creator – see multiple lawsuits for more – prompting immediate responses based on widespread myths and narrow understandings.

One being that of SPEED; that these tools hold something like a magical button; with one push, it will create an intriguing, full-scale production, of well, anything. Unsurprisingly, this is pushed repeatedly by some companies behind Gen AI tools.

Another is that of THEFT; one cannot use Gen AI without stealing from someone. While there is a serious and widespread lack of understanding of whether one uses a tool that digs into images, words, art, music videos, etc., without the creators' consent, there are other options. Companies that abide by legal and ethical guidelines and regulations, such as Adobe's commitment to uphold the pledges of the EU AI Pact.

Yet another is that of CREATION; even the rhetoric of AI being able to create anything showcases a very limited understanding of what it means to basically work with a machine, and even what creativity is. The machine does not initiate new creations from relevance, inspiration and context through divergent thinking; it combines, replicates and generates from existing patterns and data.

The machine does not initiate new creations from relevance, inspiration and context through divergent thinking

There is the myth of LIFE itself; we tend to anthropomorphise chatbots, treating them as persons. This is encouraged by the companies behind them, as their chat replies are often generated by chatbots that respond with the personalised "I" and overly kind responses, even when being put on the spot for mistakes. This all becomes part of a conversation where every answer leads to a new question, keeping the dialogue going. One may speculate on the eerie resemblance to social media platforms, where the goal is to keep people engaged for as long as possible and, in the process, collect as much data about them as possible.

The anthropomorphic rhetoric has even made its way into critical discussions of Gen AI and chatbots, with some arguing that they produce incorrect answers or outerworldly creations, often described as hallucinations. A machine does not hallucinate. It fabricates. Sometimes the fabrication does not meet the intended standards.

There are many other myths and misunderstandings. A final one to be mentioned here is that of REGRESSION: that we, as humans, will lose our creativity to AI, and that the creative industries are doomed. That is not necessarily the case.

However, critical and creative thinking has to be valued and, as the unique human traits they are, we have to challenge ourselves to not fall into complacency and degeneration from the acceptance of mediocrity that the AI chatbots tol often respond with; everything the user asks is great, turning the user into a consumer of lesser agency. Trial and error, reflecting rather than rejecting constructive criticism and disappointments throughout the process, should not discourage anyone from their art and craft, and the value of a sense of accomplishment should never be underestimated.

Towards a common language

Many great opportunities and abilities await those with curiosity and common sense, and AI can make valuable contributions to innovative new creative workflows. Just don't work towards SPEED as the core goal, get in the know of when THEFT occurs, do not rely on AI to CREATE anything new for you, beware of the illusion of LIFE, and finally, do not slip into REGRESSION, but work actively on new ways to innovate, work and create.

It will take effort and a critical look at how these things are done now, and how they can be done in the future, if we are to unlock the next level of creativity. The results will go beyond being "enhanced by AI" and will be created in a context balancing humanity and technology. In which a shared understanding and language pave the way for a path forward based on constructive discussion and decisions based on insight, not modern mythology.

Organisations should start asking themselves whether they share an understanding of, and a language for, what AI is and what it means to them

It may seem like a grand vision, but the work for getting there has already begun. In support of this effort and to advance their own teams, organisations should start asking themselves whether they share an understanding of, and a language for, what AI is and what it means to them, in both general terms and in very close-to-home, everyday work. Preferably through actual human-to-human conversation, without the polite, fabricated definitions served up by your favourite chatbot.