Artificial Intelligence cannot think

In an essay that published in The Verge, Benjamin Riley argues that the current boom in Artificial Intelligence is based on a fundamental misunderstanding: Language modeling is not the same as intelligence.

Discover more articles in search results.

“The problem is that according to current neuroscience, human thought is largely independent of human language – and we have little reason to believe that increasingly sophisticated language modeling will create a form of intelligence that matches or surpasses our own,” Riley writes.

The article goes on to point out that we use language to communicate. We use it to create metaphors to describe our reasoning. It states that people who have lost their language ability still have reasoning. People create knowledge when they are dissatisfied with current metaphors. Einstein’s theory of relativity was not based on scientific research. He developed it as a thought experiment because he was dissatisfied with the existing metaphor. It quotes someone who said: “common sense is a collection of dead metaphors.” It then states that Artificial Intelligence, at best, can rearrange these dead metaphors in interesting ways. But it will never be dissatisfied with the data it has or with an existing metaphor.

A different review (PDF) states that even as a language model, Artificial Intelligence (AI) is flawed due to its dependence on the internet. The languages ​​used on the internet are not representative of the languages ​​of the world. There are other languages ​​that contain unique descriptions/metaphors that are not found on the internet. For example, there are descriptions of snow types in Inuit languages, which describe properties that do not exist anywhere in European languages. If these metaphors are not found on the internet, AI will never be able to create them.

“That doesn’t mean AI isn’t useful. But it’s not even remotely human intelligence. It’s just a bad metaphor. We need a better one.”

Benjamin Riley is the founder of Cognitive Resonance, a new venture to improve understanding of human cognition and AI genetics.


Google preferences

Leave a Comment

Your email address is not published. Required fields are mentioned with *

Your message will not be published if:
1. Contains insulting, defamatory, racist, offensive or inappropriate comments.
2. Causes harm to minors.
3. It interferes with the privacy and individual and social rights of other users.
4. Advertises products or services or websites.
5. Contains personal information (address, phone, etc.).