Croteam’s narrative team isn’t mincing words on generative tools. In a Game Informer interview released this week, writers Jonas and Verena Kyratzes called current LLMs little more than “incredibly intelligent versions of the autocorrect” and labeled them a “plagiarism machine that makes people lose their jobs.” The pair, who have shaped the philosophical core of the Talos Principle series since the 2014 original, made the comments while discussing The Talos Principle III, the trilogy’s planned 2027 finale for PC and PS5.

Jonas described the tech as impressive on a technical level but dismissed overhyped claims from CEOs about singularity or AGI-level breakthroughs as stock-market theater. Verena echoed the skepticism, noting that placeholder AI text leaves artifacts that are “incredibly hard” to purge from a writer’s mind and flatly stated it “absolutely hinders creativity.” The studio itself avoids genAI for creative work, though Jonas noted narrow legitimate uses like word-finding aids for skilled practitioners.

Their critique lands amid broader industry pushback. Larian’s lead writer previously dismissed AI text as “a 3/10 at best,” while other voices have flagged similar labor and quality concerns. The Kyratzes frame the real problem as labor power versus capital, not the algorithms themselves, and tie the discussion to the hopeful post-scarcity world The Talos Principle III aims to depict. The series has long used artificial consciousness as a lens on humanity; now the writers are applying that lens to the tools threatening their own craft.