OpenAI's chief scientist said that AI may already have consciousness.

Ilya Sutskever, chief scientist of the American company OpenAI, stated that some neural networks may already be acquiring consciousness. However, scientists have reacted to this statement with a great deal of skepticism.

According to Futurism, Ilya Sutzkever wrote on Twitter last week, "Perhaps the large neural networks that already exist today have a bit of consciousness."

The researcher didn't name any specific neural networks. But perhaps he was referring to large-scale neural networks like GPT-3, which OpenAI has developed for translating, answering questions and filling in missing words.

This innocuous statement has prompted a strong negative reaction from scientists.

For example, University of New South Wales AI researcher Toby Walsh said such messages are misleading. "Every time such speculative comments are aired, it takes months of effort to bring the conversation back to the more real possibilities and threats posed by AI," the scientist commented.

Some felt that such a statement was made just for the sake of "noise." Citi vice president and deep learning expert Valentino Zocca wrote on social media that AI "does NOT have consciousness, but apparently hype is more important than anything else."

Of course, AI has not yet reached the level of development to gain consciousness. But that doesn't stop it from imitating it, and from imitating the consciousness of others. For example, recently a programmer from the Perm region created a "terminator" and downloaded his grandfather's consciousness into it.

It might be interesting