Search Results
6/28/2025, 12:59:14 AM
>>713832819
I used to but once you become fully aware of the limitations you start to chafe against them and it makes it harder to suspend your disbelief. When a newer, fancier model comes out the limitations change with some general expansion of capability, but you'll find those eventually too
Are there any current free LLMs that will tell the user "no, you tried but failed" if you set up a scenario where your abilities are limited and then do something well beyond what you're capable? That was my biggest problem with AID, NAI and Claude+SillyTavern when I tried them; you could say you're a blind quadruple amputee fighting a tarrasque and, no matter how much I worded it in the memory/author notes that the character should fail at tasks they can't accomplish, if I wrote "I try attack the tarrasque" I would invariably succeed. That's what makes LLMs bad at text adventures in my experience, but I acknowledge I might be retarded.
I used to but once you become fully aware of the limitations you start to chafe against them and it makes it harder to suspend your disbelief. When a newer, fancier model comes out the limitations change with some general expansion of capability, but you'll find those eventually too
Are there any current free LLMs that will tell the user "no, you tried but failed" if you set up a scenario where your abilities are limited and then do something well beyond what you're capable? That was my biggest problem with AID, NAI and Claude+SillyTavern when I tried them; you could say you're a blind quadruple amputee fighting a tarrasque and, no matter how much I worded it in the memory/author notes that the character should fail at tasks they can't accomplish, if I wrote "I try attack the tarrasque" I would invariably succeed. That's what makes LLMs bad at text adventures in my experience, but I acknowledge I might be retarded.
Page 1