🤖 #chatGPT-4 Passes the Turing Test
Specialists from the University of California asked about 500 participants to interact with both humans and AI models in a blind test. The main task for each participant was to determine whether they were interacting with a human or a computer.
56% of the participants mistakenly identified #chatgpt as a human.
According to #VitalikButerin , OpenAI’s creation has passed the Turing Test:
"The ability of people to determine who is in front of them — a human or a bot — is practically identical to a coin toss! Well, that's not quite true, because people are identified as humans in 66% of cases, and as bots in 54% of cases, but the 12% difference is negligible; in any real-world setting, this is considered a pass."