Search Results
7/21/2025, 4:44:51 PM
6/23/2025, 6:00:59 AM
I think the next step for frontends should be more metagaming. Right now, chatbots all fly by the seat of their pants... even the ones with the reasoning block. They're too volatile and shaped by the exact last thing you said to them. But what if you could, maybe initially, in the background and not exposed to the user, reason out a plotline using the bot's background information, and adhere to that regardless of what is happening in each response? I think doing this would result in much more interesting chats, because a lot of times it's really obvious what it grabbed from your response and how that shaped the entire chat.
6/21/2025, 9:26:55 PM
6/13/2025, 1:34:38 AM
>>40520093
Elaborate.
>>40520096
They use Neville's methods. What I'm saying doesn't negate him, it just needs to be recontextualized.
Elaborate.
>>40520096
They use Neville's methods. What I'm saying doesn't negate him, it just needs to be recontextualized.
Page 1