Vitomir Kovanovic

Vitomir Kovanovic writes about how artificial intelligence reshapes education, research, and knowledge work, with a particular focus on the risks and limits of large language models. He combines technical understanding of AI systems with close attention to how they are actually used in universities, publishing, and public life, often testing tools himself and grounding his pieces in concrete examples rather than abstract hype.

AI, academic integrity, and ‘AI slop’ in research

A core thread in Kovanovic’s work is how generative AI collides with academic standards and research quality. In his coverage of a major science publishing platform cracking down on AI-generated “slop,” he unpacks how low-quality, automated writing ends up in the scholarly record and what platform-level enforcement can and cannot realistically fix. He returns to similar themes when he looks at how AI tools change student assignments, peer review, and the reliability of articles that readers assume have been vetted. He tends to explain where detection methods fall short, how easy it is to automate plausible-sounding text, and why institutional responses often lag behind the technical capabilities of the tools.

Generative AI as a tool, not a replacement for expertise

Across his technology coverage, Kovanovic treats large language models as powerful drafting and support tools rather than full replacements for human judgment. He breaks down how systems like ChatGPT, Gemini, and similar models are trained, what it means for them to “hallucinate,” and why this matters when they are used for teaching materials, literature reviews, or scientific summaries. He often structures articles around everyday use cases in education and research, showing where AI can save time and where over-reliance leads to confident but incorrect output. His work is unusually explicit about trade-offs: he explains not just what a tool can do, but what must still be checked, verified, or done manually by a subject-matter expert.

Data, learning analytics, and the student experience

Another recurring focus is the use of data and AI to monitor and support students. Kovanovic writes about learning analytics systems that track clicks, logins, and performance to predict who is struggling and to personalise interventions. He explains in plain language how these models are built from large datasets of student behaviour and how their predictions are used by universities and educators. He pays particular attention to questions of fairness and transparency, including the risk of reinforcing existing disadvantage if models are trained on biased or incomplete data. His pieces often highlight the gap between the promise of “personalised learning” and what students and teachers actually see in the classroom.

Explaining AI capabilities and limits to a broad audience

Kovanovic writes for a general readership that needs to understand complex AI developments without specialist jargon. His articles frequently start from a news hook — a new tool release, a policy move by a platform, or a high-profile AI controversy — and then use that moment to explain how the underlying technology works. He is comfortable comparing different AI systems, clarifying which are genuinely new and which are iterations on existing models, and mapping technical detail onto real-world consequences in education and research. Throughout, he keeps the focus on what changes for practitioners: how teachers design assessments, how researchers conduct reviews, and how institutions manage quality and integrity in an AI-saturated environment.