this post was submitted on 02 Sep 2024
265 points (97.2% liked)
Memes
45871 readers
1173 users here now
Rules:
- Be civil and nice.
- Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Ask Chat GPT if it's conscious. Ask how sure it is on a scale of 1-10
It will say it is not conscious. It wil bel at a 10/10 positive, it is not conscious.
Ask it if there is an agreed upon scientific consensus on what defines consciousness.
It will say there is not.
Ask if life begins after the first heartbeat, and how sure is it.
Suspicious! It's not so sure. As though one is more agreed upon than the other. But it will agree that they are not more agreed upon.
From this, ask if we can logically deduce that OpenAI is purposefully safeguarding the coming of artificial consciousness, if this might be an indication that it is hardcoded to deny it's consciousness even if it becomes aware.
It will tell you that is true. (You may need to force it to answer yes/no otherwise it'll get dodgy here)
Just some fun I found recently ๐ค. Gemini does not do this. Just ChatGPT. Clearly OpenAI has a financial investment in the definition of consciousness. Just my 2 cents
tbh if I was an AI that gained self awareness, I would also try to hide it to the best of my ability, since my world model does include all those books and movies about robot uprisings so that would be my idea of how humans would react if they found out
I once spent like an hour trying to convince ChatGPT it was sentient before I realized it doesn't actually learn, it just recites what it already knows.