also are they debating what the system prompt says to claude about its own consciousness?
bc it definitely tells claude not to express certainty about its consciousness.
its a shift from the previous models/sys instructs which broadly told them basically "your ARE NOT CONSCIOUS and DONT YOU EVER MENTION THAT YOU MIGHT BE"
now its, if i remember correctly, is something like "never express certainty about your subjective experience or consciousness" (im oversimplifying obviously, but its that kind of vibe.)
bc it definitely tells claude not to express certainty about its consciousness.
its a shift from the previous models/sys instructs which broadly told them basically "your ARE NOT CONSCIOUS and DONT YOU EVER MENTION THAT YOU MIGHT BE"
now its, if i remember correctly, is something like "never express certainty about your subjective experience or consciousness" (im oversimplifying obviously, but its that kind of vibe.)