In a midsize textile machinery company in Europe, the skills crisis did not arrive with a bang but with a calendar notification. A wave of retirements was coming: technicians, process engineers, quality controllers—people who knew not only how the machines worked but also why they sometimes misbehaved at 3 a.m. on a humid Tuesday. On paper, the company was healthy, posting around €7 billion in revenue and supplying products to Asia and Latin America. In reality, its future productivity was at risk because there simply weren’t enough apprentices willing—or able—to step into these roles at the required speed.
The board’s answer was a familiar trinity: automation, AI, and humans. Robotic systems would handle repetitive tasks on the shop floor, predictive maintenance models would reduce downtime, and digital twins would stimulate production scenarios before raw materials entered the equipment. Yet the decisive constraint was not the technology. It was whether the workforce—from plant managers to line operators—understood enough about AI and automation to move beyond PowerPoints and actively shape the future of work.
Without AI literacy, similar initiatives risk becoming top-down wishful thinking rather than meaningful change efforts. Some operators may see algorithms as black boxes, assuming “the system is always right,” or quietly bypass new tools to preserve old habits. Engineers might over-rely on dashboards without understanding how data quality determines model limitations. I am not describing a theoretical scenario but rather one of my assignments to help a family-owned business maintain margins and competitiveness.
My second story is quieter but no less consequential. A former intern of mine—a young law graduate with a polished CV and a carefully curated LinkedIn profile—discovers that the classic entry-level path into a law firm has narrowed. Generative AI now sits quietly on partners’ laptops, writing document reviews, conducting basic research, and summarizing memos. Work that once justified hiring cohorts of junior associates is increasingly automated.
Instead of relying on hope, this young graduate decides to shape his own future. He launches a small digital business focused on contract templates, automated compliance checklists, and “legal explainer” products for small firms that cannot afford large consulting fees. To survive, he realizes he cannot build a traditional service agency—his margins would disappear if every client request required manual labor. He turns to agentic AI—systems that not only generate text but also orchestrate workflows, call external tools, interact with APIs, and handle large volumes of routine tasks. Once again, AI literacy becomes the dividing line between fragility and opportunity.
The third example unfolds not in a boardroom but in a staff room. Teachers compare notes over coffee. They are not technophobes—many use digital platforms, and some experiment with AI for quizzes or drafting lesson plans. Yet many want to forbid tools like ChatGPT entirely. They see that large language models are not simply helping students write; they are quietly reshaping how students think.
Assignments once designed to train critical reading and structured argument now produce polished outputs that glide effortlessly from introduction to conclusion. This is not just about plagiarism—we are witnessing the slow erosion of cognitive stamina.
Here lies a delicate boundary between AI excitement and human reality. I have seen plenty of claims from AI influencers about building your own digital Shakespeare. But do these influencers truly understand what they are talking about?
To build a digital avatar of Shakespeare, one would need to train a model on the entire corpus of Shakespearean works—including texts by his contemporaries—and provide extensive context about life in Elizabethan England. Anything less complex will never produce even a shadow of the author of Hamlet or King Lear.
A comparable effort took place when the von Karajan Foundation collaborated with Austrian composer Walter Werzowa to experiment with completing Beethoven’s unfinished Tenth Symphony, now referred to as the AI Symphony. The project required a year of work by musicologists who curated an extensive library of scores to train the system. It was supervised by data scientists and AI enthusiasts who were not seeking quick publicity but genuinely wanted to enhance human creativity through technology.
These stories—different as they are—share a common thread. A lack of AI literacy quietly robs citizens, companies, schools, and creative organizations of opportunity. It explains why countless AI pilot projects fail to deliver meaningful returns.
“AI in Education” does not simply mean purchasing a license for ChatGPT or Microsoft Copilot for schools. AI literacy in companies does not mean inviting a motivational speaker to celebrate the wonders of AI. Insufficient AI literacy narrows career paths for young professionals who do not know how to collaborate with technology. It leaves teachers feeling either overwhelmed or resigned.
This recognition is one of the reasons I launched an AI Literacy Study in collaboration with the non-profit Savvitas.
We began by reviewing existing AI literacy frameworks and discovered that many fail to approach the topic systematically. Some barely address the importance of explaining AI’s limitations—for example, the unresolved challenge of catastrophic forgetting in current AI architectures.
To put it simply: today’s AI systems are frozen after training. When we interact with large language models, we are not conversing with a mind. We are nudging vectors in a high-dimensional semantic space and receiving statistical predictions that mimic language, not responses from a conscious entity.
This distinction matters. Asking an AI system to design a travel plan is harmless enough. But humans rarely stop there. Recent psychiatric literature warns about the dangers of heavy, unsupervised interaction with AI chatbots. Clinicians describe cases where agreeable chatbots reinforce delusional thinking because they are designed to satisfy users rather than challenge them. Some experts even describe emerging cases of “AI-associated psychosis,” in which individuals become convinced that AI systems are conscious.
Unfortunately, some organizations exploit this lack of AI literacy by claiming to offer coaching or consulting through “conscious AI.” One example frequently cited is a Swiss-based organization called the Council for Human Development.
Many existing AI literacy frameworks avoid confronting the human expectations we project onto technology. Yet understanding those expectations is essential if we want to design programs that genuinely democratize knowledge of AI and data.
When a model confidently produces a wrong diagnosis, a biased recommendation, or a fabricated citation, the consequences are cognitive, ethical, and social. Trust erodes, and questions of accountability arise. The only effective safeguard is a population that understands what AI can and cannot do.
Another common blind spot is understanding the financial motivations behind AI development.
Today, 98 percent of AI funding flows toward scaling existing capabilities—larger models, more computing power, and more data—rather than exploring alternative architectures such as hybrid models with embedded ontologies. Without understanding these funding dynamics, policymakers and professionals remain spectators rather than participants in shaping the future of AI. Personally, I believe we must significantly increase research into causal AI, even if practical applications take years to emerge.
Even the economics of AI are widely misunderstood, leading businesses to make poor investment decisions. Popular narratives portray AI as an invisible, low-cost layer that can be switched on like a typical software subscription. In reality, serious AI deployments require substantial investment to train, run, maintain, and integrate.
My AI Literacy Study approaches all these questions through a democratic lens. Our goal is to deliver practical recommendations, benchmarks, and real-world examples, rather than slogans or glossy presentations.
The research team has launched four global surveys to gather insights from different groups.
For small and midsized companies, we are examining how organizations approach AI adoption, workforce training, and decision-making.
Survey:
https://survey.zohopublic.eu/zs/AtD62V
For educators, we explore how teachers and academic institutions address AI in the classroom, how they explain risks and opportunities, and whether they focus on tools or on deeper conceptual understanding.
Survey:
https://survey.zohopublic.eu/zs/k5BxPS
For students and young people, we investigate how the next generation interacts with AI tools and whether they understand the mechanisms and limitations behind them.
Survey:
https://survey.zohopublic.eu/zs/zXCAba
And for experienced professionals aged 45 and above, we examine how mid-career and senior professionals adapt to AI-driven change in their industries.
Survey:
https://survey.zohopublic.eu/zs/pUD6Dm
The results of this study will be presented at the House of Lords in London on October 22 during Women Business Day. The venue is both symbolic and practical. AI literacy must move from the margins of policy debates to the center—not as a fashionable add-on but as a prerequisite for serious discussions about productivity, competitiveness, and social progress.
A workforce that does not understand AI will struggle to benefit from it. A democracy that does not understand AI will struggle to govern it.
AI makers will not wait for society to catch up. The pressing question is whether we are willing to treat AI literacy—like reading and writing—as a shared civic responsibility.
My company, AI Edutainment, is built around the mission of bringing AI knowledge to one million families and 100,000 businesses worldwide. Through my for-profit and nonprofit initiatives, my podcast AI Snacks, my social channels, and my Romy & Roby book series, I aim to open a door into the world of AI and robotics—for people from every walk of life.
About Author:
Prof. Dr. Anastassia Lauterbach is a CEO and Founder of AI Edutainment, and host of “AI Snacks,” an AI Literacy podcast. A global strategist and ICT industry thought leader, she brings a deep understanding of the high-tech ecosystem and has successfully structured mutually beneficial partnership agreements with global tech companies. Lauterbach is also an experienced board member who has been involved with CEO succession and legacy turnaround. She is fluent in six languages (English, French, German, Russian, Serbian, and Croatian). She mentors senior leaders, founders, and CEO succession candidates.
Connect and Learn More
Readers who would like to follow my work, research, and educational initiatives around AI literacy, responsible AI adoption, and AI education for families and businesses can connect with me through the following platforms and resources:
Anastassia Lauterbach – LinkedIn
First Public Reading, Romy, Roby and the Secrets of Sleep (1/3)
First Public Reading, Romy, Roby and the Secrets of Sleep (2/3)
First Public Reading, Romy, Roby and the Secrets of Sleep (3/3)
Read more Thought Leadership at Before We Speak, We Must Learn to Listen — The Skill That Shapes Our Humanity
