I wanted to share some thoughts on AI, artificial intelligence. The part of the discussion that has my attention is certainly not being overlooked. But it’s not at the center of the debate. There are lots of different uses for so-called ‘generative AI’. But the kinds I’m particularly focused on are the ones used to create art based on textual prompts, write essays or even compose songs. I got focused on this originally a few months ago when an artist/illustrator friend of mine started talking about it on social media, how Silicon Valley’s latest disruption was set to put illustrators and artists – so often living on the financial margins already – out of business.
It’s here to stay, although we need to work through a legal framework for how to protect and license information not available yet in the public domain. Perhaps Google or some other massive indexing company can provide a framework.
Even without such a framework, it’s here to stay if only due to the significant mass of public domain art and writing. I’d encourage developers to focus their machine learning on works in the public domain until that legal licensing framework can come about. Until then, AI companies can just hire out artists, illustrators, and writers to develop content for machine learning - though this will slow things down significantly.
I can’t see a way to allow the free flow of information WITHOUT licensing it to continue, although there’s also no way to stop it without a lawsuit. That horse has run out the barn door. But if I were an AI company, I’d be very, very careful about what libraries and spidering/scraping algorithms I was using in my machine learning so I could avoid the massive potential IP issues if they use information not yet in the public domain to develop their projects and not licensed for such use.
On the other hand, machine learning is not so hard and massive computing power so plentifully available and cheap that a student in a dorm room couldn’t set up an art app or writing app on her own. The genie is out of the bottle, and has been for a LONG time. I assume we all want free flow of information. It seems to me that this flow requires us to rethink IP law and the concept of public domain at the expense of creatives.
I certainly can’t think my way out of the problem.
AI is exposing and exasperating the already prevalent incentive challenges between corporations, government, society, and individuals.
AIs being developed by silicon Valley, the US (or china), or as an educational nonprofit would all have very different goals and motivations. We should not cede this tool, and all it produces, to for profit companies who consume everything and return back nothing close to comparable value.
A for profit system will go where the money is. That’s not an AI created problem.
If someone with artistic ability goes to a museum, studies the works there and learns what might be the features which makes other humans find those works compelling, then creates a new work of art based on those concepts, we wouldn’t call that infringement of the IP of those museum works. Same as if someone goes on the internet and looks at artistic works for sale and starts their own business of their own designs.
Same applies to music, literature, or the routine graphic arts from works for hire. IP and AI don’t really have any overlap, unless the AI copies the IP-protected work. No copy, no violation of copyrights. (There is nuance, of course, as George Harrison found out with My Sweet Lord.)
And what exactly is an “original work” anyway? Josh is correct that most artists and writers and composers grind it out producing what others have contracted from them. They may have some original elements, but most are derivative anyway. I’ll only be impressed with AI if it produces something truly original and not derivative, but with an aesthetic worthy of the title ‘art’. Most human efforts don’t reach that level, but all are more worthy than anything that AI might produce since at least the human imagination is not bound by only what is experiences directly.
In the olden days, some of the first sites online were news, art, photography and hobby sites, cooking and crafts and other skills. They were created by the people who wanted to share their work and ideas. Until copycats from across the globe started ripping off the creative work of others to sell cheap look-somewhat-like copies.
I was making custom personalized furniture and one day a customer asked me why my prices were so high when they could order the same thing for under $60. Puzzled, I asked something like, “what?”. They sent me a link to AliBaba where I saw photos of my work taken on my patio for sale at a fraction of the cost of my materials, much less labor. This is when I learned about the DMCA and took the steps to remove my images from the knock-off products shops.
I saw some of those knock-offs up close and they would embarrass a 14 year old woodshop student from my school days. Bad in every respect.
My business prospered to the point I could not keep up with sales and was tired of working 12-16 hour days. I retired in 2009 but kept my site open to teach others how to DIY. I have no income from the site and live on Social Security and a small pension now. My point is that without any regulation, big brands will find the same kind of problem. I’m already seeing MSNBC YouTube videos (and many others) under peculiar names/channels. There is nothing to prevent your creatives from being re-purposed for profits.
Google built their entire business first by showcasing creative content, then by ‘owning’ it to keep others on their site and away from the creators. AI is nothing new in that respect.
Sorry, but Google is the last entity I would hand any framework keys to.
