3 ms·
Researchers say chatbot exhibits self-awareness
- allears 3y agoTheir conclusions are based on the output, without having any idea of how LLMs work. LLMs are very good at generating plausible language, remembering previous input, and sounding very similar to a human. But self-aware? Nope.
- labrador 3y ago> "I know a person when I talk to it," Lemoine said in an interview in 2022. One thing I've learned with the introduction of LLMs is that some people are really bad at judging sentience. Other people are really quick to anthropomorphize everything, from their car to their pet. When someone has both these qualities then I can see how they feel ChatAIs are conscious and self-aware. To me it's laughable because I can talk to an LLM like it's a person and not think it's a person but rather a machine. I have a normal ability to judge self-awareness in others and I never anthropomorphize anything because I don't like doing it. I was in the Navy so when sailors would refer to the ship as "she" I always resisted thinking that way.