(1/2)
@NetscapeNavigator
> the Fediverse needs AI
As others have pointed out, this statement is so ambiguous as to be meaningless. If you use "AI" in this way you will be frequently misunderstood. Especially in the fediverse, where there is a strong bias against #MOLE Training that spreads to anything described as "AI".
(2/2)
@NetscapeNavigator
> I first made this point during the major spam attack that hit the Fediverse last year
I suggest saying "pattern-matching moderation tools" or somesuch, and making (or linking) more specific proposals. For example, an automated tool that looked for unusual patterns of posting, and alerted the mods to some example posts (maybe by hooking into the existing Report workflow). Maybe talk to @iftas folks like @jaz about any research they've done into the viability of this.
