I’ve been meaning to write about AI and gender… and then a friend of mine sent me something he’d written to read over and he covered what I wanted to cover but in a much more charming way, so he’s given me permission to share it here…
I see no mention of the poodles’ lived experience during this horrific event. The poodles likely had no say as to how they were clipped, the contestants just assigned them a hairstyle based on their obviously racist, colonial, oppressive beliefs about how poodles should look. History tells us that 89% of poodles actually identify as other species so clipping them to conform to the stereotype of “poodle” is violence. Henceforth, the Olympic committee should provide a safe space for anyone who could be traumatized by witnessing any further display of hate clipping. Do Better, Olympics.
You touched my heart, Jane. I am starting a GoFundMe page for breed reassignment clipping for those poor beasts. Thanks for bringing our attention to this important issue.
Clearly, these infernal machines are programmed not to respond to some prompts on forbidden subjects, and part of their method is to lie and make stuff up; when this is pointed out, they apologise for "any confusion" and try again, possibly equally inaccurately. They are operating within parameters set by their creators, and it would be interesting to know exactly what these are.
I’m a fan of Heterdorx but I don’t use Twitter so thanks for sharing that. It’s all just a little terrifying. The conversation about AI parallels a bit how blasé people seemed to be about “the gender thing” 10 years ago. “Oh, kids these days, it’s just a fad, it will blow over. No harm done, they’ll grow out of it.” “Oh, high schoolers are just trying to get out of writing papers, kids will be kids. No real harm can come if this and isn’t it all just so neat!”
“And aren’t we supposed to be able to trust this all new singin’-and-dancin’ technology?”
Cute article, but regarding the above quote: no, we’re not supposed to “trust” this new technology. ChatGPT and other LLMs are in their infancy and OpenAI, who implement and maintain ChatGPT have been exceedingly clear about this. OpenAI released the beta version of ChatGPT thinking they could get the public to help continue to train it. They didn’t anticipate how quickly it would become popular. So no, you shouldn’t trust this technology. It’s right to be skeptical of it. It is not Google (which you pointed out is ALSO confused about Olympic poodle clipping) or Wikipedia. I find it occasionally helpful on technology topics or helpful to rewrite something I’ve already written.
However, like so much new tech, these things often have a life of their own, and the general public may just assume they can trust it. And the consequences of that are unknown. Hence the skepticism and caution.
Ooooh scary! Twists and turns! I am not even being sarcastic. What a world what a world. my trust issues are not so bad after all. I love poodles and this story. Two paws 🐾 up!
I see no mention of the poodles’ lived experience during this horrific event. The poodles likely had no say as to how they were clipped, the contestants just assigned them a hairstyle based on their obviously racist, colonial, oppressive beliefs about how poodles should look. History tells us that 89% of poodles actually identify as other species so clipping them to conform to the stereotype of “poodle” is violence. Henceforth, the Olympic committee should provide a safe space for anyone who could be traumatized by witnessing any further display of hate clipping. Do Better, Olympics.
You touched my heart, Jane. I am starting a GoFundMe page for breed reassignment clipping for those poor beasts. Thanks for bringing our attention to this important issue.
On a related matter, it is worth reading Maya Forstater's experience with ChatGPT.
https://www.forstater.com/talking-to-chat-gpt-about-trans-rights/
Clearly, these infernal machines are programmed not to respond to some prompts on forbidden subjects, and part of their method is to lie and make stuff up; when this is pointed out, they apologise for "any confusion" and try again, possibly equally inaccurately. They are operating within parameters set by their creators, and it would be interesting to know exactly what these are.
See also:
https://www.serioustrouble.show/p/the-legal-brief-beyond-the-machine?lli=1&utm_source=profile&utm_medium=reader2
Check out what Corinna Cohn has done, too... https://twitter.com/heterodorx/status/1662836806893678596
I’m a fan of Heterdorx but I don’t use Twitter so thanks for sharing that. It’s all just a little terrifying. The conversation about AI parallels a bit how blasé people seemed to be about “the gender thing” 10 years ago. “Oh, kids these days, it’s just a fad, it will blow over. No harm done, they’ll grow out of it.” “Oh, high schoolers are just trying to get out of writing papers, kids will be kids. No real harm can come if this and isn’t it all just so neat!”
Thanks. I'll have a look.
Jordan Peterson has also had some interesting engagements with ChatGPT that are worth checking out.
“And aren’t we supposed to be able to trust this all new singin’-and-dancin’ technology?”
Cute article, but regarding the above quote: no, we’re not supposed to “trust” this new technology. ChatGPT and other LLMs are in their infancy and OpenAI, who implement and maintain ChatGPT have been exceedingly clear about this. OpenAI released the beta version of ChatGPT thinking they could get the public to help continue to train it. They didn’t anticipate how quickly it would become popular. So no, you shouldn’t trust this technology. It’s right to be skeptical of it. It is not Google (which you pointed out is ALSO confused about Olympic poodle clipping) or Wikipedia. I find it occasionally helpful on technology topics or helpful to rewrite something I’ve already written.
Good point.
However, like so much new tech, these things often have a life of their own, and the general public may just assume they can trust it. And the consequences of that are unknown. Hence the skepticism and caution.
The winner of the poodle clipping event always wins by a hair
An attorney used ChatGPT to write a legal brief. It did not turn out well: https://amp.cnn.com/cnn/2023/05/27/business/chat-gpt-avianca-mata-lawyers/index.html
Thank you. Fascinating. I suspect we might be seeing more of this...
Ooooh scary! Twists and turns! I am not even being sarcastic. What a world what a world. my trust issues are not so bad after all. I love poodles and this story. Two paws 🐾 up!
Well done, Caspar!
Many thanks!
Fantastic tale. And terrifying.