Tatejo: Difference between revisions
added images |
|||
| Line 4: | Line 4: | ||
== Relations With Other Users == | == Relations With Other Users == | ||
=== d9eoead === | |||
=== Catha'vi === | === Catha'vi === | ||
| Line 9: | Line 11: | ||
</blockquote> | </blockquote> | ||
=== Fern === | |||
== Ban == | == Ban == | ||
| Line 17: | Line 21: | ||
== TatGPT == | == TatGPT == | ||
On June 8th, 2025, [[Catha'vi]] pitched to [[Fernsy|Fern]] the idea of training a [[wikipedia:Large_language_model|Large Language Model]] off of Tatejo's message history, and running it as a bot in the [[OWOT Discord|OWOT Discord server]]. Fern agreed to this on two conditions: | |||
# That the bot be entirely constrained to the <u>#trash</u> channel. | |||
# That Fern should be able to audit the LLM's source code for security checks. | |||
Catha'vi promptly began work on the project, and TatGPT began spitting out its first nonsensical messages by June 11th, 2025. Though it must be noted that this first run of output was only marginally more unhinged than Tat's original dialogue.<gallery widths="150" heights="150"> | |||
File:Tat rant 1.png|Attempt by AI to break down Tat's rant into a graph(''June 8th, 2025'') | File:Tat rant 1.png|Attempt by AI to break down Tat's rant into a graph(''June 8th, 2025'') | ||
File:Tat rant 2.png|Attempt #2 | File:Tat rant 2.png|Attempt #2 | ||