The silver lining of AI content is that it will hopefully encourage people to question the veracity of everything they read, hear, and see, IMHO. A healthy dose of critical thinking and scepticism is always a good thing. "Truth" is always hard to nail down, and its wise never to forget it. I'm suspicious of too easily trusting something because its authoritative. The dark edge to the silver lining is that sometimes it means people can give up and not believe anything.
Terry, I so rarely disagree substantially with you that this is an interesting one. I think you're slipping into an assumption that lots of people make, and it leads them to the mistake I think you're making. You write "generative AI is a substitute for human creation", and that's the assumption. Yes, you can use GenAI that way (and I wish people wouldn't). But you can also use GenAI as a tool to enable human creation, and to become better at human creation. I've seen grad students use LLMs in exactly that way, to learn how to write better - NOT to avoid writing, but to explore alternatives and think about how and why an LLM version of a piece of text might differ from theirs.
There will be a chapter in my (with Bethann Garramon Merkle) forthcoming book "Teaching and Mentoring Writers in the Sciences" (https://scientistseessquirrel.wordpress.com/2024/12/10/what-our-new-book-looks-like-teaching-and-mentoring-writers-in-the-sciences/) exploring this, and a chapter in the not-quite-so-rapidly-forthcoming 3rd edition of "The Scientist's Guide to Writing" exploring these ideas. How can writers use GenAI tools, not to avoid writing, but to become better writers? How can these tools support, rather than substitute for, human creation? I think that's the interesting and important question.
Steve, I see your point. I think you're taking issue with that particular sentence, but also I point out how AI is also a tool to facilitate creation. I think still GenAI is a substitute for creation, and in the use case you're writing about, I think it's a substitute for creating intermediate steps. Are those intermediate steps removing the human element of writing? I think I agree with you that it does not necessarily, at least in the process of writing. In the process of developing an understanding of information and digesting this information, this does constitute a shortcut. If one is learning how to write, rather than concerned about the human process of parsing information and ideas about any particular topic, then I think that's legit as a writing practice, but perhaps not with respect to the content itself. So I suppose a person who is a relative master of a given topic consults with GenAI to explore how to express those ideas in certain ways as a matter of informing their work, that's legit, but usually these are the folks who are least likely to choose to use AI in writing (and some journals and publishers are even asking us to certify that GenAI was not used in the writing process!)
Steve knew I would chime in here. :) It's true that our book has a chapter about how to use LLMs as well and ethically as possible right now. But, as Steve knows, if the press would have let us, or if I'd written a book alone, I'd have skipped talking about LLMs altogether. Or better yet, written something like what Terry wrote here. The chapter in our book is a compromise between how I understand (and avoid AI/LLMs for any generation of ideas and material, including no spell-checking, no coding, etc.) And Steve's understanding and use of these algorithms. I think the compromise we have in our book is the best option, given it's a great and useful book, people and the press expected us to address LLMs, and we co-wrote, so the book must reflect both our takes. But as someone who is both a visual artist (many mediums) and a writer in many genres, I find LLMs abhorrent. I also remain very concerned and averse because of the environmental, ethical, human etc., etc., issues. Steve has dissected some of those in helpful ways in the book and on his blog. And I better understand the affordances and appeal of them thanks to our many discussions. But, fundamentally, I advocate against them and don't use them. It seems important to state that transparently as our book approaches publication. (Maybe the tension we've aimed to resolve or we least balance will be useful for readers, too??)
Every human artist/musician that I know and myself included won't stop making physical, tangiable art. My photos on substack are taken by myself and my drawings are by hand. graphics unless otherwise referenced are made with actual data and data points from physical samples I collected geo-referenced in time and space. Real art, real science.
Thanks for this thoughtful post Terry. Preserving artistic and acientific integrity is important!
I tell my students, when you take an exam, write a paper remember your written (hand) words are and ahould be yours. own them.
The silver lining of AI content is that it will hopefully encourage people to question the veracity of everything they read, hear, and see, IMHO. A healthy dose of critical thinking and scepticism is always a good thing. "Truth" is always hard to nail down, and its wise never to forget it. I'm suspicious of too easily trusting something because its authoritative. The dark edge to the silver lining is that sometimes it means people can give up and not believe anything.
Terry, I so rarely disagree substantially with you that this is an interesting one. I think you're slipping into an assumption that lots of people make, and it leads them to the mistake I think you're making. You write "generative AI is a substitute for human creation", and that's the assumption. Yes, you can use GenAI that way (and I wish people wouldn't). But you can also use GenAI as a tool to enable human creation, and to become better at human creation. I've seen grad students use LLMs in exactly that way, to learn how to write better - NOT to avoid writing, but to explore alternatives and think about how and why an LLM version of a piece of text might differ from theirs.
There will be a chapter in my (with Bethann Garramon Merkle) forthcoming book "Teaching and Mentoring Writers in the Sciences" (https://scientistseessquirrel.wordpress.com/2024/12/10/what-our-new-book-looks-like-teaching-and-mentoring-writers-in-the-sciences/) exploring this, and a chapter in the not-quite-so-rapidly-forthcoming 3rd edition of "The Scientist's Guide to Writing" exploring these ideas. How can writers use GenAI tools, not to avoid writing, but to become better writers? How can these tools support, rather than substitute for, human creation? I think that's the interesting and important question.
Steve, I see your point. I think you're taking issue with that particular sentence, but also I point out how AI is also a tool to facilitate creation. I think still GenAI is a substitute for creation, and in the use case you're writing about, I think it's a substitute for creating intermediate steps. Are those intermediate steps removing the human element of writing? I think I agree with you that it does not necessarily, at least in the process of writing. In the process of developing an understanding of information and digesting this information, this does constitute a shortcut. If one is learning how to write, rather than concerned about the human process of parsing information and ideas about any particular topic, then I think that's legit as a writing practice, but perhaps not with respect to the content itself. So I suppose a person who is a relative master of a given topic consults with GenAI to explore how to express those ideas in certain ways as a matter of informing their work, that's legit, but usually these are the folks who are least likely to choose to use AI in writing (and some journals and publishers are even asking us to certify that GenAI was not used in the writing process!)
Steve knew I would chime in here. :) It's true that our book has a chapter about how to use LLMs as well and ethically as possible right now. But, as Steve knows, if the press would have let us, or if I'd written a book alone, I'd have skipped talking about LLMs altogether. Or better yet, written something like what Terry wrote here. The chapter in our book is a compromise between how I understand (and avoid AI/LLMs for any generation of ideas and material, including no spell-checking, no coding, etc.) And Steve's understanding and use of these algorithms. I think the compromise we have in our book is the best option, given it's a great and useful book, people and the press expected us to address LLMs, and we co-wrote, so the book must reflect both our takes. But as someone who is both a visual artist (many mediums) and a writer in many genres, I find LLMs abhorrent. I also remain very concerned and averse because of the environmental, ethical, human etc., etc., issues. Steve has dissected some of those in helpful ways in the book and on his blog. And I better understand the affordances and appeal of them thanks to our many discussions. But, fundamentally, I advocate against them and don't use them. It seems important to state that transparently as our book approaches publication. (Maybe the tension we've aimed to resolve or we least balance will be useful for readers, too??)
Very fair, Bethann, and one of the great things about working together is that I think we've each influenced each other's views (not just on this!)
Yes, totally!! It's been an amazing, positive experience to cowrite and learn together on so many topics, including this one!
Every human artist/musician that I know and myself included won't stop making physical, tangiable art. My photos on substack are taken by myself and my drawings are by hand. graphics unless otherwise referenced are made with actual data and data points from physical samples I collected geo-referenced in time and space. Real art, real science.
Thanks for this thoughtful post Terry. Preserving artistic and acientific integrity is important!
I tell my students, when you take an exam, write a paper remember your written (hand) words are and ahould be yours. own them.
Also why there are authentic typos in my responses when using a device rather than when I have a pen in hand.
Forgive my common typos, I spell and write much better than I can possibly text!