Re: CRYPTO-GRAM, November 15, 2022
  By: August Abolins to TCOB1 on Thu Dec 01 2022 07:34 am

 > Hello TCOB1!
 > 
 > ** On Tuesday 29.11.22 - 22:09, TCOB1 wrote to All:
 > 
 >  T> Adversarial ML Attack that Secretly Gives a Language Model a Point of Vi
 > 
 > [...]
 > 
 >  T> One example of an extension of this technology is the ΓÇ£persona bot,ΓÇ¥
 >  T> an AI posing as an individual on social media and other online groups.
 >  T> Persona bots have histories, personalities, and communication styles. Th
 >  T> donΓÇÖt constantly spew propaganda. They hang out in various interest [.
 > 
 > That is totally insane.   :/
 > 
 > --
 >   ../|ug
 > 
 > --- OpenXP 5.0.51
 >  * Origin: (} Pointy McPointface (618:250/1.9)

No, it is a logical conclusion.

Once AI becomes advanced enough, the Internet will become unusable because
there won't be a way to separate legitimate traffic from bot traffic. It will
be worse than with newspapers - in newspapers you know the content is always
illegitimate.


--
gopher://gopher.richardfalken.com/1/richardfalken
--- SBBSecho 3.15-Linux
 * Origin: Palantir * palantirbbs.ddns.net * Pensacola, FL * (618:250/24)