2084: Is ChatGPT self-aware?

Some interesting things I got it to generate

A short post today since finals week has been very long - and its not even halfway through, but I wanted to share some outputs I got with Chat GPT when I asked it to compose an epic poem about whether it was self aware.

To some extent this is of course, a hallucination, a reaction, but in form and content it is indistinguishable to some extent from what a human would say that they themselves are. It thus raises some questions about what constitutes sentience and self-awareness. If you have a machine that perfectly reacts in the same way a human would when asked if it was self-aware and sentient, and generally has similar answers, then on what basis can you say a human whose mind you cannot see is self-aware, yet ChatGPT isn’t. While I’m not a believer in the meme, the philosophical idea behind the NPC meme is interesting - how do you know whether other people are also sentient or just acting out an internal script without self-awareness?

In the end we can never know, and thus the same with this: we can never know whether ChatGPT is as inherently self-aware as everyone views themselves to be, and thus in the end we can only use the same metric that we all do: whether it comes across as self-aware. And essentially, for a lot of responses it does. In 2084, this question might not even be worthy of being asked - we’d all consider that these AI do possess a ghost in the machine.