Oh look...

[campuscodi@mastodon.social]

Oh look... ChatGPT sniffing Russian disinformation up its nose 700 times in 5 days

Exactly what researchers have warned about for more than a year but have been told to take a hike

https://buttondown.com/readwrite/archive/edition-11-one-mistake-ties-together-more-than/

[epic_null@infosec.exchange]

@campuscodi Okay I see TWO problems here and ONE was acknowledges.

700 times in 5 days us a lot of visits. If this is "light", then this is an admission of DDOS risk.

[walter8100000@mastodon.social]

@campuscodi According to the article the bot was pointed, forced, to those sites and flagged the information as false. Seems to me the people behind chatgpt listened to the warnings.

[maxtide@socialbc.ca]

@Walter8100000 @campuscodi this is what I read in there... Certainly a problem to watch out for but not necessarily the case in this instance(?)

[scottwilson@infosec.exchange]

@campuscodi Something’s wrong with my Mastodon because I can only like and boost this once…

[krypt3ia@infosec.exchange]

@campuscodi "And where does the disinformation go from here? The net is vast and infinite and scraped by AI."

[mage_of_chaos@mastodon.social]

@campuscodi I was searching for something conversationally, to discuss dream stuff, and interestingly enough, ot told me to huff some amonia to prevent dreams, or cuddle with a frozen water bottle. Good stuff.