AI content: Between progress and fatigue
Automated content is becoming increasingly prevalent—and with it, skepticism about its quality and relevance is growing. Companies face the challenge of integrating AI effectively into their processes while also meeting the demand for authentic content formats. A position paper by EY Switzerland highlights key developments.
The second thesis paper from auditing and consulting firm EY Switzerland examines the near future of AI in three theses.
Thesis 1: AI content fatigue is on the rise
AI-generated content is now an integral part of digital communication, but it is increasingly being rejected. The growing amount of predictable, slickly polished text means that users are increasingly turning to authentic formats such as podcasts or interviews. "Users are increasingly drawn to authentic human perspectives to understand the context and the person behind opinions," explains Adrian Ott, Partner and Chief Artificial Intelligence Officer at EY Switzerland.
When used judiciously as an editorial and quality control tool, AI can support rather than replace human creativity. When creating a new article with AI, there are now a number of ways in which superfluous jargon and AI writing styles can be completely removed, keeping the content fresh and easy to read and allowing authors to focus on their original thoughts. An entire branch of science (Human-AI Interaction, "HAX") has emerged around the interaction between humans and AI, investigating this connection and identifying ways forward.
Thesis 2: Company data as a hurdle for AGI
Development continues to progress - particularly in the direction of artificial general intelligence (AGI), which could theoretically handle any intellectual task at a human level. In practice, however, its use often fails due to the poor quality of company data or complex decision-making processes. "The dream of loading all company data into an AGI that immediately gains an overview of the company is unlikely to work as expected," says Ott. "The question is not only when AGI will arrive, but also how long it will take to integrate it into our business world in a meaningful way." Efficiency, reliability and trust therefore play just as central a role.
Thesis 3: More efficient AI models on the rise
At the same time, there is a growing focus on the energy consumption of large AI models. While powerful models previously required enormous computing capacities, developments such as the Chinese DeepSeek model show that considerable results can be achieved with fewer resources. In the long term, this could lead to a split in the AI landscape: On the one hand, highly complex, self-learning systems for research and innovation, and on the other, efficient models for everyday use.
Human-AI partnership as a key factor
For companies, this means that AI remains a decisive competitive factor, but its use must be strategic. While the demand for credible content is increasing, new, cost-efficient AI models offer opportunities for customized applications. Those who invest in the quality of their data at an early stage could benefit from technological developments in the long term.
Here the entire thesis paper can be read in detail.
