AI milestone: GPT‑4.5 nearly aces Turing test in new study
Does GPT chat think in a similar way to us? The media has circulated reports that artificial intelligence has passed the Turing test. Does this imply that AI might become indistinguishable from humans?
Researchers from the University of California, San Diego conducted a study which found that the GPT-4.5 language model from OpenAI was mistaken for a human in 73% of cases. The Turing test, which was designed to verify this, involved conducting a text conversation with both the chat and a human simultaneously. The study aimed to show whether participants could differentiate between a human and a machine.
Other models, like LLaMa-3.1-405B, were also tested but were mistaken for humans in only 56% of cases. Older systems, such as ELIZA and GPT-4o, were much less effective, with only 23% and 21% incorrect identifications, respectively.
The Turing test has sparked many debates about consciousness, thinking, and the limits of artificial intelligence. It remains one of the most well-known and controversial symbols in discussions about the relationship between humans and machines. It has been popularized as the ultimate indicator of machine intelligence, but the debate continues regarding what the test actually measures and whether it truly assesses machine intelligence.
The Turing test was introduced in 1948 by Alan Turing, a renowned computer scientist. Initially, the test involved four individuals: two players and an operator who were supposed to play chess with a hypothetical machine. This experiment aimed to simulate a game, and its creator believed it was an effective method to assess intelligent machine behavior.
The study has sparked discussions about the limits of artificial intelligence and its ability to emulate human thinking. Critics point to the short duration of the test and the fact that the AI assumed a specific personality, which could have influenced the results.
