<
https://www.techdirt.com/2024/09/04/ai-checkers-forcing-kids-to-write-like-a-robot-to-avoid-being-called-a-robot/>
"Can the fear of students using generative AI and the rise of questionable AI
“checker” tools create a culture devoid of creativity? It’s a topic that is
curiously one worth delving into a bit more deeply, in part because of
something that happened this weekend.
Earlier this year, we had a post by Alan Kyle about California bill SB 942.
That bill would require AI companies to offer a free AI detection tool, despite
the fact that such tools are notoriously unreliable and prone to nonsense. As
Kyle wrote, the bill takes a “nerd harder” approach to regulating technology
its backers don’t understand.
SB 942 has continued to move forward just passed in the California Assembly.
It’s now on Governor Newsom’s desk to potentially sign.
I was thinking about that this weekend after a situation at home. One of my
kids has an English homework assignment. They had to read Kurt Vonnegut’s
famous short story,
Harrison Bergeron, and write a short essay about it.
Since I do a fair bit of writing, my kid asked me to review the essay and see
if I had any pointers. I gave a few general suggestions on how to think about
improving the flow of the piece, as it read very much like a standard first
draft: a bit stilted. My kid went off to work on a rewrite.
If you’re unfamiliar with the story of
Harrison Bergeron, it’s about a
society that seeks to enforce “equality” by placing “handicaps” on anyone who
excels at anything to bring them down to the least common denominator (e.g.,
ugly masks for pretty people, having to carry around extra weights for strong
people). One of the morals to that story is on the perils of seeking to force
equality in a manner that limits excellence and creativity.
Later in the day, the kid came by with their school-issued Chromebook, which
has Grammarly Pro pre-installed. The students are encouraged to use it to
improve their writing. One thing that the tool has is an “AI Checker” in which
it tries to determine if the submitted text was written by AI.
This is similar to “plagiarism checkers” that have been around for a few
decades. In fact, Grammarly’s “check” covers both AI and plagiarism (or so it
says). Those systems have always had problems, especially around false
positives. And it seems that the AI checkers are (unsurprisingly) worse."
Cheers,
*** Xanni ***
--
mailto:xanni@xanadu.net Andrew Pam
http://xanadu.com.au/ Chief Scientist, Xanadu
https://glasswings.com.au/ Partner, Glass Wings
https://sericyb.com.au/ Manager, Serious Cybernetics
