[FRIAM] Adversarial Go trick defeats KataGo

[Eric Charles]

"the adversarial policy works by first staking claim to a small corner of
the board. He provided a link to an example
<https://goattack.alignmentfund.org/?row=0#no_search-board> in which the
adversary, controlling the black stones, plays largely in the top-right of
the board. The adversary allows KataGo (playing white) to lay claim to the
rest of the board, while the adversary plays a few easy-to-capture stones
in that territory."

This sounds oddly reminiscent of the vs-computer RISK strategy of taking
over Australia, and ceding the rest of the board until you suddenly come
out and win. Which no half-decent human opponent would ever let you do, but
the computer AI make totally viable.



On Thu, Nov 10, 2022 at 1:24 PM Roger Frye <frye.roger at gmail.com> wrote:

>
> https://arstechnica.com/information-technology/2022/11/new-go-playing-trick-defeats-world-class-go-ai-but-loses-to-human-amateurs/
>
> -. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .
> FRIAM Applied Complexity Group listserv
> Fridays 9a-12p Friday St. Johns Cafe   /   Thursdays 9a-12p Zoom
> https://bit.ly/virtualfriam
> to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> FRIAM-COMIC http://friam-comic.blogspot.com/
> archives:  5/2017 thru present
> https://redfish.com/pipermail/friam_redfish.com/
>   1/2003 thru 6/2021  http://friam.383.s1.nabble.com/
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20221110/941d20b1/attachment.html>