I have been incorporating the use of Chat GPT in certain assignments in my classes - I teach undergraduate and master courses. I believe that teaching the students to use these tools in a ‘value added’ way is as critical as in my youth it was to teach how to use excel and word (something that is totally taken for granted today).
Well put. I fear we may need to revisit computer literacy in k-12 education though. I can’t even tell you how many “space bar formatters” I’ve come across in recent years.
I think it's fantastic that your curiosity has led you to utilize these different LLMs for projects. That's the best way to learn more about these models. You brought up a great point in your ChatGPT discussion about some sort of regulations for chatbot usage in school systems. It is great for things like brainstorming and formatting, but it also can do a students' entire assignment. I am a current college student, and I have witnessed the reliance students have on these sorts of platforms to complete assignments. There are regulations in place at my institution that can detect AI generated responses in assignments, but students find ways around that too. I think the education system needs to come up with a solution to this chatbot reliance or else students are going to continue to use these platforms in a way that doesn't advance their learning.
It's great that you're having fun with your project and exploring AI, but I think it's important to acknowledge the deeper issue with AI generated art. These tools are trained on massive data sets that include illustrations by many talented artists. Often without their knowledge or consent. The reality is... it may have only taken a couple of days to produce because the AI was trained on thousands of hours of work done by real people who aren’t getting compensated for their contributions. They're not getting a cut from the profits made on Kindle books either. When Mid journey just got released, I played around with it and noticed the watermark of a befriended Illustrator vaguely shining through in the image it generated. I'm sure it got better over time, but that was sloppy and shocking.
I completely understand your point of view and I respect that. But AI doesn’t create in a vacuum, its "skill" comes from studying the labor of others, and that labor is what made your quick project possible.
While AI might save time, I think we should be critical of what it’s actually replacing. It's one thing to explore AI as a creative tool, but if it ends up undercutting the livelihood of illustrators, who have put in years of effort and their unique talent (which takes years of training and a lot of guts to develop) into their craft, it’s worth considering the ethical implications. How do we strike a balance between creative exploration with AI and ensuring the original creators behind this innovation are not left behind?
You bring up some excellent points, and they're all worthy of further consideration. I really appreciate you shedding light on this area, which I admittedly overlooked in my original post.
As my opinion is still forming, I would like to share perspectives I've come across from others (not taking sides, just continuing the conversation):
Some argue that artists have always drawn inspiration from their peers and predecessors, seeing AI-generated art in a similar light. Others have likened generative AI to "fair use" transformative works that add new expression, meaning, or messaging—similar to Shepard Fairey’s case with the Associated Press over his Obama Hope poster or Ed Sheeran’s defense against a copyright claim: https://youtu.be/NcCKlsTgjeM
The core issue seems to be the ease with which one can now "borrow inspiration" or "transform" another's work. It's one thing to use your skills to design something in another artist’s style or likeness; it’s another to achieve a similar result by simply mashing a button. Perhaps the craft lies in the human generated prompt itself?
Another consideration is whether the new work harms the market for the original or serves as a direct substitute. In my example, it might be difficult for any specific artist to make such a claim.
Clearly, I have more questions than answers. It does feel ethically questionable in some ways, but I wonder how different it really is from consuming a bunch of art the old fashioned way and letting that influence your own production.
How cool is it to find reasoned, and considered dialog on a social network? Well done Substack.
Thanks for engaging in this conversation, Michael, and for not only considering my point of view but also sharing it with others. You bring up some valid points in your reply, especially around 'inspiration' and how artists have always drawn from each other. But the key difference, in my view, is agency. When artists in the past were influenced by their peers, they were still creating something themselves, putting their unique stamp on the work.
I went through art school for four years, and it wasn't just about learning technical skills. It was about developing a deeper sense of myself, my voice, and my talents. Illustration isn't just a collection of lines and colors; it's a reflection of a person's personality, tastes, and influences. Each piece of work contains layers of a person's inner world. AI can replicate the outer shell, but it can never capture those intricate layers that make art so personal and meaningful. I think that's the feeling we all have in our gut when we look at AI-generated art. That's what's missing.
The concept of 'fair use' or transformative work does apply in some contexts, but for me, it's not quite the same here. AI-generated art relies on databases of art created by humans that the AI was trained on, again, without the consent or compensation of those artists. The ethical question is whether it’s really 'transformative' or just efficient repurposing of someone else's hard work.
I think it's great to explore the possibilities of AI, but the system also needs to evolve to ensure that the creators behind the datasets get some form of acknowledgment. It’s not about stifling innovation, but about ensuring fairness in how that innovation is built. Curious to hear more thoughts on this. Substack is such a wonderful place indeed. Thanks Michael!
I really enjoyed reading about your intake on AI being used for creative jobs by the creatives. It was enlightening. I've recently been breaking the resistance I had created on using AI, to finally accept its potential in actually helping me. Even when the AI suggests gibberish, it can be helpful to draw a line of what you definitely don't want.
I have been incorporating the use of Chat GPT in certain assignments in my classes - I teach undergraduate and master courses. I believe that teaching the students to use these tools in a ‘value added’ way is as critical as in my youth it was to teach how to use excel and word (something that is totally taken for granted today).
Well put. I fear we may need to revisit computer literacy in k-12 education though. I can’t even tell you how many “space bar formatters” I’ve come across in recent years.
I think it's fantastic that your curiosity has led you to utilize these different LLMs for projects. That's the best way to learn more about these models. You brought up a great point in your ChatGPT discussion about some sort of regulations for chatbot usage in school systems. It is great for things like brainstorming and formatting, but it also can do a students' entire assignment. I am a current college student, and I have witnessed the reliance students have on these sorts of platforms to complete assignments. There are regulations in place at my institution that can detect AI generated responses in assignments, but students find ways around that too. I think the education system needs to come up with a solution to this chatbot reliance or else students are going to continue to use these platforms in a way that doesn't advance their learning.
We definitely need to figure out how to teach students to use LLMs to aid their writing, not as a crutch for it.
Thank you for sharing!
It's great that you're having fun with your project and exploring AI, but I think it's important to acknowledge the deeper issue with AI generated art. These tools are trained on massive data sets that include illustrations by many talented artists. Often without their knowledge or consent. The reality is... it may have only taken a couple of days to produce because the AI was trained on thousands of hours of work done by real people who aren’t getting compensated for their contributions. They're not getting a cut from the profits made on Kindle books either. When Mid journey just got released, I played around with it and noticed the watermark of a befriended Illustrator vaguely shining through in the image it generated. I'm sure it got better over time, but that was sloppy and shocking.
I completely understand your point of view and I respect that. But AI doesn’t create in a vacuum, its "skill" comes from studying the labor of others, and that labor is what made your quick project possible.
While AI might save time, I think we should be critical of what it’s actually replacing. It's one thing to explore AI as a creative tool, but if it ends up undercutting the livelihood of illustrators, who have put in years of effort and their unique talent (which takes years of training and a lot of guts to develop) into their craft, it’s worth considering the ethical implications. How do we strike a balance between creative exploration with AI and ensuring the original creators behind this innovation are not left behind?
You bring up some excellent points, and they're all worthy of further consideration. I really appreciate you shedding light on this area, which I admittedly overlooked in my original post.
As my opinion is still forming, I would like to share perspectives I've come across from others (not taking sides, just continuing the conversation):
Some argue that artists have always drawn inspiration from their peers and predecessors, seeing AI-generated art in a similar light. Others have likened generative AI to "fair use" transformative works that add new expression, meaning, or messaging—similar to Shepard Fairey’s case with the Associated Press over his Obama Hope poster or Ed Sheeran’s defense against a copyright claim: https://youtu.be/NcCKlsTgjeM
The core issue seems to be the ease with which one can now "borrow inspiration" or "transform" another's work. It's one thing to use your skills to design something in another artist’s style or likeness; it’s another to achieve a similar result by simply mashing a button. Perhaps the craft lies in the human generated prompt itself?
Another consideration is whether the new work harms the market for the original or serves as a direct substitute. In my example, it might be difficult for any specific artist to make such a claim.
Clearly, I have more questions than answers. It does feel ethically questionable in some ways, but I wonder how different it really is from consuming a bunch of art the old fashioned way and letting that influence your own production.
How cool is it to find reasoned, and considered dialog on a social network? Well done Substack.
Thanks for engaging in this conversation, Michael, and for not only considering my point of view but also sharing it with others. You bring up some valid points in your reply, especially around 'inspiration' and how artists have always drawn from each other. But the key difference, in my view, is agency. When artists in the past were influenced by their peers, they were still creating something themselves, putting their unique stamp on the work.
I went through art school for four years, and it wasn't just about learning technical skills. It was about developing a deeper sense of myself, my voice, and my talents. Illustration isn't just a collection of lines and colors; it's a reflection of a person's personality, tastes, and influences. Each piece of work contains layers of a person's inner world. AI can replicate the outer shell, but it can never capture those intricate layers that make art so personal and meaningful. I think that's the feeling we all have in our gut when we look at AI-generated art. That's what's missing.
The concept of 'fair use' or transformative work does apply in some contexts, but for me, it's not quite the same here. AI-generated art relies on databases of art created by humans that the AI was trained on, again, without the consent or compensation of those artists. The ethical question is whether it’s really 'transformative' or just efficient repurposing of someone else's hard work.
I think it's great to explore the possibilities of AI, but the system also needs to evolve to ensure that the creators behind the datasets get some form of acknowledgment. It’s not about stifling innovation, but about ensuring fairness in how that innovation is built. Curious to hear more thoughts on this. Substack is such a wonderful place indeed. Thanks Michael!
I just finished playing with Udio. Pretty cool, Thanks for letting me know about it.
My pleasure. It’s pretty fun, right?
I've made 4, 32 second songs. They're pretty cool. Thanks for letting me know about it.
Awesome
Michael, the books you created for your children are brilliant! Amazing idea.
Thank you so much!
I really enjoyed reading about your intake on AI being used for creative jobs by the creatives. It was enlightening. I've recently been breaking the resistance I had created on using AI, to finally accept its potential in actually helping me. Even when the AI suggests gibberish, it can be helpful to draw a line of what you definitely don't want.