By Steven Levy
Publication Date: 2026-01-23 16:00:00
The big AI Companies promised us that 2025 would be “the year of the AI agent.” It turned out that was the year talk about it AI agents and the preparations for this transformative moment in 2026 or perhaps later. But what if the answer to the question “When will our lives be fully automated by generative AI robots doing our jobs for us and basically running the world?” is like the New Yorker cartoon: “How about never?”
That was essentially the message of an article published without much fanfare a few months ago, in the middle of the overhyped year of “agentic AI.” Entitled “Hallucination Stations: On Some Basic Limitations of Transformer-Based Language Models,” the aim is to show mathematically that “LLMs are incapable of performing computational and agentic tasks beyond a certain level of complexity.” Although I find the scientific findings difficult, the authors – a former SAP CTO who studied AI with John McCarthy, one of the founders of the field, and his…
