A personal AI philosophy

I attended a conference this week and one of the speakers spoke about the ethics of AI. He did so in way that I felt was very thoughtful and considered. I won’t attempt to summarize details of his whole talk, but one of the things he encouraged was to develop your own personal philosophy when it comes to AI. What uses are acceptable to you, and what lines will you not cross? Funny enough, I’d added a section to my “About Bob” page that essentially outlined my philosophy under the heading “What is your stance on AI?”:

I do not use large language models (LLMs) to write posts on bobwertz.com or any of my sites. I’m also strongly opposed to using generative AI imagery. I believe that LLMs were largely trained on creative materials without permission, and as such, they are remixing stolen intellectual property. Also, in general, AI imagery sucks. I’m an artist and I can draw it myself. I use an AI feature of Micro.Blog to add alt descriptions to my pictures. The computer is better at describing the image than I am, and I feel like having alt tags on my website is important. I also use the ChatGPT service my university provides to assist with simple coding and troubleshooting.

I’m happy with this as a start. For me, I’m hesitant to throw AI at everything, but I can’t deny that in some situations, AI is helpful. I’m concerned about the environmental impact, but I’m just as concerned about the intellectual property theft. As systems uses less resources and as new models are (hopefully) ethically trained, my opinions might shift. We’ll see…

---

Bob Wertz is a type designer, Ph.D. student and researcher living in Columbia, South Carolina. He’s been blogging since 2008.