The earworm is "Call Me Irresponsible". The lyrics are a strange choice to associate with a medication that helps some patients to relieve migraine, but that has side effects such as unreliable bowels.
The subtext is AI... and the loss of honour.
Take academic dishonesty. The Northern Illinois University has a very good Academic Integrity Tutorial here:
https://www.niu.edu/academic-integrity/faculty/types/index.shtml
https://www.niu.edu/academic-integrity/faculty/causes/index.shtml
It examines cheating, plagiarism, fabrication or falsification, and sabotage, then goes on to explain the What? Why? How? and consequences of cheating. And there are quizzes!
Without cheating, of course, I got 10/10 on Quiz 1; 6/6 on Quiz 2; 6/6 on Quiz 3; 10/10 on the plagiarism quiz...at which point, I felt I ought to get back to blogging. The section on plagiarism might be particularly relevant for authors. There is more to it than one might think, including inadequate attribution...and much more.
https://www.niu.edu/academic-integrity/faculty/committing/plagiarism.shtml
The fifth quiz is about detecting and preventing cheating, which brings me to an interesting article by Karen Gullo of the Electronic Freedom Foundation (EFF).
Apparently, academic dishonesty is an issue which some French examiners take very seriously, especially when it has been necessary to allow students to take examinations in private settings, as opposed to in the traditional hall with human proctors prowling between the ranks and rows of desks.
According to a group that one might call the Parisian version of EFF (as I just did), La Quadrature du Net (LQDN), objected that the monitoring software called TestWe was too much of an invasion of the students' privacy.
LQDN presented a good argument that, "...just because the data exists or is available does not mean it
is legal to use it for any purposes.”
One could make the same point about much of AI, and also about ChatGPT and all the "cheating-facilitating software" that AI provides, and does so without attribution. Which is one reason why AI-generated prose or art may or may not be copyrightable.
Legal bloggers Haim Ravia and Dotan Hammer for the lawfirm Pearl Cohen Zedek Latzer Baratz discuss an initiative that has been lauched by the US Copyright Office to look into copyright law and policy questions risign from the use of AI.
Currently, copyright only protects works that are created by humans.
Legal blogger Benni Amato, on the IP Blitz blogsite of the Intellectual Property law firm Gordon Rees Scully Mansukhani uses the Q and A format to discuss (in very small font) the two burning questions of the moment, namely "Can You Copyright AI Art?" and "Does AI Art Constitute Copyright Infringement?"https://www.ip-blitz.com/2023/03/ai-art-and-copyright-in-the-unites-states/#page=1
He is particularly interesting on the use of AI to create art fakes, or "vicarious copyright infringement".
Finally, for me, for today, legal blogger Daniel Lumm for Nelson Mullins Riley and Scarborough LLP details an experiment that he carried out using ChatGPT.
ChatGPT is the bot that can pass the bar (to become a lawyer), some medical exam or other, and some business school exams.
Imagine, and this is my own imagining, if HAL 9000 of 2001 A Space Odyssey were to replace, say, Dr. Fauci!
Back to Mr. Lumm's experiment and analysis of the issues to consider with regard to OpenAI and ChatGPT, which include ownership, responsibility, confidentiality, privacy, security, and more.
https://www.nelsonmullins.com/idea_exchange/insights/chatgpt-on-what-terms-is-the-future-so-bright
To riff off the good people of LQDN while taking legality out of the equation, just because it is there, and one can use it, maybe does not mean that one should (use it).All the best,
Posts
No comments:
Post a Comment