A.I. Is on Its Way to Something More Remarkable Than Intelligence (www.nytimes.com)

The article argues that A.I. has already crossed into what most people would call “intelligence” and that this shift has been driven less by a fixed definition and more by interactive evidence — a point Alan Turing made in 1950 when he proposed assessing machines by whether their answers are indistinguishable from humans’ (the Turing test). Citing a 2024 YouGov poll showing growing public acceptance that computers rival or will surpass human intellect, the author contends we have already broadened our concept of intelligence in response to increasingly sophisticated systems, much as digital photography became accepted as photography. The provocative claim is that the same social-epistemic process will lead to A.I. being regarded as conscious: continued interaction with richer, more human-like models will reshape our criteria for consciousness. This isn’t portrayed as mere wordplay but as a real feedback loop between empirical capabilities and our conceptual frameworks. For the AI/ML community, the piece implies practical consequences — shifting evaluation metrics and benchmarks, new criteria for attributing mental states, and urgent ethical, design and policy questions about how to treat systems as their behavior increasingly maps onto human-like faculties.