IEEE Computational Intelligence Magazine - August 2023 - 5

Message about a CIS initiative to
research ways that AI can assist in the
review process. Many potential good
uses are being studied.
It will take a while for IEEE to iterate
on an official policy for Big AI. I'm
going to give my opinion. First of all,
don't include ChatGPT or other programs
as a co-author; that makes no
sense. The main guideline right now is
to put some details ofhow you used AI
in your manuscript in an acknowledgement
section. For example, if you use a
LLM to act like a human copyeditor,
just acknowledge it along with something
about the extent of its use. It's still
your responsibility as the author to be
sure the copyediting doesn't change the
meaning ofyour text. Or, ifyou ask for
help identifying previous work in an
area, be upfront about it, and carefully
check those references - we've all now
heard stories of false information being
generated by LLMs. Don't use AI programs
to " create " new ideas to pass along
as yours. My bottom line is that if you
are uncomfortable about divulging how
you used an AI program, you probably
shouldn't use it. For me, this boils down
to a question of personal ethics. If you
are asked to referee a paper, you should
not ask a LLM to do your review. That
amounts to a breach in confidentiality
since most LLMs will integrate that input
into their training corpus. I'd sure hate to
see my new good idea given to the next
person who asks ChatGPT. Wouldn't
you?
The genie is out ofthe bottle. AI capabilities
will continue to provide useful
tools for authors, referees, editors, and
publishers. Let'suse them to make
our jobs easier but avoid ethical pitfalls.
As always, feel free to contact me at
kellerj@missouri.edu with your thoughts,
suggestions, questions, and innovative
ideas, particularly with thoughts on the
use ofAI in publishing. Please stay safe and
healthy. I hope to see many ofyou at the
CISmeetings over the rest ofthe year.
PS: I asked the public version of
ChatGPT on May 9, 2023 to rewrite
the opening paragraph above. Here is
what it gave me. While it is smoother
and contains most of my thoughts, I
don't think it " sounds " like me. What
do you think?
" Let's address the elephant in the room:
Large Language Models (LLMs) like
ChatGPTand other AI technologies specializing
in graphics, mathematics, coding, and
audio are dominating the production of
reports and manuscripts. While they can be a
lot offun, such as when I asked ChatGPT
to create song lyrics aboutfuzzy logic in different
genres (although I wasn't fond of the
folk song, I enjoyed the blues lyrics and bluegrass
rendition it generated), we must
acknowledge their significant potential for
serious applications. However, there is a
growing concern about whether we can have
too much of a good thing, particularly given
the wave of both positive and negative hype
surrounding AI. Many IEEE members and
other individuals have signed a letter calling
for a six-month " AI Pause, " worried about
the potential ethical and moral misuse ofAI.
While I did not sign the letterfor various reasons,
it is crucial to have conversations about
regulating AI and our own interactions with
AIprograms. "
AUGUST 2023 | IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE 5
http://ieeefoundation.org/goldsmith
IEEE Computational Intelligence Magazine - August 2023

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - August 2023
