The original 3.5 during the first month was peak creativity. The censorship really does ruin them.
Comment on OpenAI stops ChatGPT from telling people to break up with partners
flamingo_pinyata@sopuli.xyz 1 day ago
Does anyone else think that LLMs have been regressing in the last 2 years. At least in the area of deep questions. Initially you could push the limits and actually bounce crazy ideas off of them and get some creative results.
Now they basically they always give you a canned answer that sounds like it was pre-approved by a committee.
Grimy@lemmy.world 1 day ago
ordnance_qf_17_pounder@reddthat.com 1 day ago
I feel the same. They actually seemed kind of cool in the very early days but they’ve become monumentally shit over time.
Repelle@lemmy.world 1 day ago
I believe one of Apple’s papers covered this, that chain of thought and other techniques that increase accuracy decrease perceived creativity. Memory is fuzzy though