AI: What Could Go Wrong?

National newspaper duped into running GPT-4-written rage-click opinion piece

When Irish AIs are smiling

The Irish Times, one of the republic’s respected broadsheet newspapers, has apologized for publishing an opinion article said to have been generated by AI software and submitted by an anonymous trickster.




This reply was automatically generated by EndeavourOS Forum’s AI.



The investment bank Goldman-Sachs estimates that generative AI will have a negative impact on 300 million jobs (pdf-link), not counting other possible negative consequences due to concentration of power and immense consumption of resources.

I am afraid, they accidentally (?) forget about the negative impact from the huge mistakes and fake misguiding products of the really stupid AIs.
As said in another similar topic, any AI is as smart and flawless as the programmer that writes its code. Every programmer in the world knows that there is no bugless program (with the exception of pacman, of course :wink: ).
The more critical the jobs that any AI is given, the more significant the negative consequences.

Be aware of the Beast :japanese_ogre: !


We got a CV today which read like it was produced by ChatGPT. As a joke we put the text into GPT-4 - and it improved the text. It came across as less machine-like.

That is alarming - a CV which was full of keywords and (as it turned it out) had a few lies thrown in had AI make these less obvious :exploding_head:



Then we won’t be that much worse off :rofl: