[ExI] The Automathician

[BillK]

On Mon, 27 Mar 2023 at 23:45, spike jones via extropy-chat
<extropy-chat at lists.extropy.org> wrote:
>
> There is another, possibly more worrisome possibility: the AI could write whatever it wanted, quote a bunch of sources that it also created.  The referenced sources could reference the same sources the AI created.  An AI could create 15 papers, all of which reference each other.  None of these would then really be “phony” exactly, for the AI could produce all 15 sources, regardless of which one of the 15 one chose. >
>
> It leads to the question of what it means to be published.  Published by who or what?  Can an AI publish its own paper by just having it available in its own memory where anyone can reference it?  We could end up with reams of research papers generated by AI where it is difficult to distinguish from human-generated I.
>
> Oh it is sooooo damn cool to be alive today watching all this unfold.  Thank you evolution, we bless thee and praise thee, thou completely natural process.
>
> spike
> _______________________________________________


The new chatbots are already misquoting each other.
<https://www.theverge.com/2023/3/22/23651564/google-microsoft-bard-bing-chatbots-misinformation>
Quote:
Google and Microsoft’s chatbots are already citing one another in a
misinformation shitshow
By James Vincent   Mar 22, 2023

Microsoft’s Bing said Google’s Bard had been shut down after it
misread a story citing a tweet sourced from a joke. It’s not a good
sign for the future of online misinformation.

What we have here is an early sign we’re stumbling into a massive game
of AI misinformation telephone, in which chatbots are unable to gauge
reliable news sources, misread stories about themselves, and misreport
on their own capabilities. In this case, the whole thing started
because of a single joke comment on Hacker News. Imagine what you
could do if you wanted these systems to fail.
---------

AIs going crazy!  :)


BillK