AI impacts on the artist
Robert Brown (RLBrown)
Hello SVMUG members and friends,
Recently, Bryan Chaffin, a pundit on general computer technology, talked to the Silicon Valley Macintosh User Group about his feelings about AI. Bryan is well-known and well-liked in this sphere, so we hold his opinions in high regard.
During his talk, Bryan discussed two key issues that have been making headlines in the AI world for some time now.
- The large language models, LLM, used by one type of AI called generative AI, were trained on publicly accessible text and images without compensation or general permission from the original creators of the text and pictures.
- Generative AI poses a significant challenge to the value of human creative works, despite the fact that the AI-produced work is merely a replication and lacks true creativity. This is a matter of serious concern that we need to address.
Regarding compensation, when individuals share their text or images online, they rightfully expect a return on their investment. For instance, a journalist contributing to an open website anticipates benefits from the publisher, marketing, or enhanced reputation. Similarly, a visual artist, such as a photographer, who uploads images to a stock photography service, relies on small payments for each download. This issue of fair compensation is crucial and needs our attention.
Regarding creativity, generative AI models, such as those used in creating text, art, or music, can produce works that are novel and original in appearance. They do so by learning patterns from vast datasets of existing works and recombining them in new ways. This can result in creative outputs that are new and unique, though they are ultimately based on existing human-created content.
While LLM-based generative AI can generate works that appear creative, it needs more genuine understanding and intent. Human creativity is often driven by a deep understanding of context, emotion, and purpose. AI does not possess these qualities; it generates outputs based on statistical patterns rather than intrinsic comprehension or emotional engagement.
However, AI can be a valuable collaborative partner to human creativity. Ai can fulfill three roles.
- Combining existing elements in new ways, as directed by the artist.
- Creating a set of possibilities based on the direction of the artist.
- Transform a work, again based on the direction of the artist.
When AI is used as a collaborative partner, issues of ownership and attribution are quickly resolved. The person who provides the AI with prompts and direction is the artist. At the same time, it is ethical to acknowledge the use of AI tools in the work.
Pivoting back to the question of compensation, the thorny issue is that the barn doors were opened, and the cows are now all free-range and unbranded. A creator will never be able to prove, except in some narrow cases, that his or her work was significantly “sampled” by an AI, creating a derivative work. (The narrow cases are where AI lacked multiple sources for a specific prompt and thus synthesized a work based on a tiny sample of original works.)
Ultimately, civil courts will shrug their shoulders and say, “What’s done is done.” This is unfair, but it is the practical consequence.
One more thing. I was careful to restrict my remarks to LLM-based generative AI. Some researchers do not consider LLM-based generation to be actual AI at all. There is no intent, no understanding within these systems. True Artificial General Intelligence may require systems involving much more than Large Language Models.
— Robert Brown
Kathryn D Lauricella
--
You received this message because you are subscribed to the Google Groups "SVMUG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to svmug+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/svmug/143D83B2-F41D-4228-BBD3-AD302F4EF182%40rlbrown.org.