Apple co-founder Wozniak: ChatGPT does not understand human nature and may make terrible mistakes

On the evening of February 13, Beijing time, it is reported that Apple co-founder Steve Wozniak (Steve Wozniak) recently said that ChatGPT is indeed impressive, but it can also make some terrible mistakes.

ChatGPT is an artificial intelligence (AI) chat robot developed by OpenAI. As soon as it was launched, ChatGPT immediately became the focus of global attention. In a short period of more than two months, it has accumulated more than 100 million users.

While it has won widespread praise, it has also raised some questions. Wozniak, co-founder of Apple, said in an interview with the media a few days ago that ChatGPT is really impressive and very useful to human beings.

But at the same time, Wozniak said, it can also make serious mistakes. "the problem is that while doing good for us, it can also make terrible mistakes because it doesn't understand human nature," he said. "

Wozniak also compared the potential harm of ChatGPT to the artificial intelligence technology of self-driving cars. At present, artificial intelligence cannot replace human drivers, he said. Wozniak: "it's like you're driving a car, and you know what other cars might do because you know people."

OpenAI CEO Sam Altman (Sam Altman) also admits that ChatGPT has "biased flaws" and that sometimes it may give biased responses, including racism and sexism.

Of course, this is not just a personal problem for ChatGPT. Earlier this month, Google's equivalent Bard made a mistake in answering questions about the Webber Space Telescope, wiping more than $100 billion off Google's market capitalization in a single day.

Despite skepticism about the technology, ChatGPT has become more and more popular. OpenAI, the developer of ChatGPT, recently received an investment of about $10 billion (currently about 68.1 billion yuan) from Microsoft, which plans to introduce OpenAI's artificial intelligence into products such as Bing search engine and Office.