Artificial intelligence that can deceive people: GPT-5!

As the launch of GPT-5 approaches, the results of a new test have been published on how effectively artificial intelligence systems can imitate humans. This test reveals how successful modern artificial intelligence tools are at producing human-like text and language tasks. This test about GPT-5 could be a turning point in the development of artificial intelligence.

GPT-5 successfully passed the Turing test and managed to fool people

The Department of Cognitive Science at UC San Diego conducted a controlled Turing Test involving GPT-4, GPT-3.5, and the 1960s simple rules-based chatbot ELIZA. In this test, participants tried to guess whether the person they were chatting with was a human or an artificial intelligence after a five-minute conversation.

In testing, GPT-4 was identified as human 54% of the time by participants, compared to 50% for GPT-3.5. Both models performed much better than ELIZA (22%), but fell behind real humans (67%). These results show that GPT-4 makes humans guess correctly above chance when it comes to detecting AI. This shows that today’s AI systems are capable of deceiving people.

Analyzes revealed that participants relied on language style, socio-emotional factors, and knowledge-based questions to determine whether they were speaking to a human or an AI. GPT-4 and GPT-3.5 were programmed with instructions to adopt a “young person” persona and use slang, even making occasional spelling mistakes. This enabled the AI ​​to exhibit a more human-like style of language.

The interface used in the test was similar to a typical messaging app, with the AI ​​sending messages with a delay based on the number of characters. This method prevented the AI ​​from giving away that it was not human by responding too quickly.

These results, obtained before the launch of GPT-5, show how far artificial intelligence technology has come and what potential it offers in the future. Artificial intelligence is further advancing in human-machine interactions, increasing its ability to mimic people’s language and behavior.

With the release of GPT-5, the impact of artificial intelligence on humans is expected to increase even further. This latest Turing Test once again revealed the capacity of artificial intelligence to deceive people. These developments about “GPT-5” give important clues about the future role and potential of artificial intelligence. This rapid evolution of artificial intelligence can bring revolutionary changes in human-machine interactions.