Ethereum co-creator Vitalik Buterin claims that OpenAI's ChatGPT-4 passed the Turing test. Buterin refers to research by scientists from the University of California in San Diego. The Turing test is a situation in which a person is unable to recognize that he or she is talking to artificial intelligence and not another human being.
ChatGPT-4 passed the Turing test? That's what the research says
Vitalik Buterin has recently been speaking more and more often about generative artificial intelligence. Buterin emphasized, for example, the points of contact between the world of cryptocurrencies and generative artificial intelligence, highlighting the advantages and opportunities resulting from such a connection. This time, the co-creator of the second cryptocurrency on the market referred to the fact that ChatGPT-4 passed the Turing test, at least according to scientists from the University of California in San Diego. The Turing test is a situation in which a conversational artificial intelligence model is so human-like that people talking to it confuse it with a real human. The term was coined by the famous mathematician Alan Turing, who proposed it in 1950.
Nearly 75 years later, scientists from the University of California, San Diego published a paper titled “Humans Cannot Distinguish GPT-4 from a Human in the Turing Test.” About 500 people took part in the study. The test involved anonymous conversations with other people and generative artificial intelligence. According to research, people mistakenly thought ChatGPT-4 was human 56% of the time. Buterin summed it up by saying:
This means that people's ability to tell the difference between a human and a bot is basically flipping a coin!
Will ChatGPT-4o take interaction with AI to a new level?
Vitalik Buterin's concerns about the Turing test come at an interesting time for the AI industry. OpenAI announced a new model, ChatGPT-4o. One of the most important novelties introduced by version 4o is the possibility of virtually natural-sounding voice conversation with a chat bot. The pauses between the human query and the AI response range between 232-330 milliseconds. This time is comparable to the response in a natural human conversation. The presentation of OpenAI caused many technology analysts to compare the capabilities of the 4o model to the 2013 movie “Her”.