By Mike Chrabaszcz
Re-printed with the author’s permission.
View original post.
Artificial intelligence, It’s all the buzz!
When Medium blogs and the NPR show “Hear and Now” highlights a “shiny object”; it might be an inflection point. That, or it’s close to the holidays and something needs to be wrapped up with a marketing bow and presented to the board, and to the masses before the end of the year. Voilà: more AI generated content and the new tools to create it. It’s all the rage; haven’t you heard? As a relatively newbie to this space, it’s intriguing, but I’m not sure if it really deserves all the hype that it’s receiving. Over the years, improved technologies have been made available to the masses to create content such as spell checking word processors, autocompletion email programs, image enhancers, music composition effects, and video production; to name a few. Welcome the new publicly facing and consumable examples of how the masses can create content, now with AI. Among them I checked out ChatGPT, Lexica , Amper, and Lensa. This recent news, coupled with the Next Gener[AI]tive Newsletter and the recent interview with media tech leader Jeremy Toeman (founder of Aug X Labs) on the upcoming release of “Prompt to Video” has raised a few interesting topics in recent conversations around AI Content Generators, adding to the buzz.
Science fiction becomes fact
AI has bordered on science fiction and fact for years. When new technologies are introduced, there can be a disorientation reaction that can incite fear, no different from when the person who stepped on to the first driverless elevator and asked “Where’s the Operator?” The fear of displacement will always challenge new ideas: “They took our jobs”. New tools may displace some current content creators unwittingly, but it will also bring new jobs. This technology and the business around it is new and will require all those who will be affected, such as the content creators, those creating these tools, those profiting from both, and maybe governments, all to be at the table for when ethical policy and governance is established. How much control will the content creator have in this brave new world? “I’m Sorry Dave, I’m afraid I can’t do that.” AI based content generating tools are evolving. Acceptance and usage is going to take time.
Is this legal?
What happens when an AI content generator creates something that is in violation of copyright or trademark and it is released to the public? Is the content producer or the AI software responsible? This ranks in the same realm with the lack of legal precedent as a victim’s claim from a self-driving car accident. If I create content with AI, can I legally trade it as NFT? Fodder for the lawyers, not the technologist. What is the government’s role? Can this be self regulated? If Washington is too restrictive, will everything go offshore? And where does the first amendment come in, if the media content was created by AI? Should we expect the Vice President’s husband to promote a ban on “Dangerous AI Generated Content” out of fear of what impact it may have on the children of America? Where do we put the label with the disclaimers? Uncharted waters ahead.
Privacy and potential for bad actors
Do these content creation tools use data from other sources? Where does that come from? Can it be data intended for research for an open source project, but is now being used in a commercial product utilizing that open source software? What data will be collected from these tools? Who owns it? As a user, can I opt-out and protect the use of my data? Is the EU onboard, does it meet GDPR? What needs to be disclosed in the Terms of Service for the tools?
Consider this case: based on Mike’s Facebook page, and his likes and his friends and their likes, he may not mind, or even be aware that the AI generated video he just created and shared on social media has intentional product placements that someone paid for and will profit from. Ching, ching, ching. And when he plays the video, he loves it, after all he likes that beer. And so do his friends. And so do the accountants. Likes and Thumbs up, all around.
Can it be biased?
Can AI generated content unknowingly set an agenda, promote hate or have a bias? Maybe. Some early implementations of AI were frightening. In 2016, Microsoft’s chatbot learned to be a bigot from Twitter, go figure. Based on the message from the documentary “Coded Bias”, one might ask the question if AI generated content will be appealing to more than just white men. Are Google search results biased? The process of creating the algorithms must take this into consideration. But a counter argument would say that the consumer may have a preference and may want the music in the video to be “jazz” or the background color to be “green”, when the algorithm says “rock” and “white”, based on the data. How do you strike a balance for unbiasedness while meeting personal preference?
Is it safe to create AI generated content?
Should I trust an Amazon provided computer generated cartoon character to compose a bedtime story for my child? Assuming that Amazon has done the right thing, understands the risks, and tested for a finite graph of plot combinations ideal for children, I would say yes. As a parent, I want to trust that it will be limited and safe. However, let’s remember that this is not a babysitter, it also expects that the parent is involved. The results should not only provide a pleasant story, but also empower my child to feel they were part of the creative process. Video games have been providing alternative path plots for years. Sure, as such content creators are made available, there needs to be safeguards that they will not be hacked, so that your kid can sleep and we all live happily ever after.
You can’t do that! It ain’t got no soul
Art, literature, and music are inspired. It comes from the heart. It’s a blessing from above, gifted to the chosen to create for us. It’s not something that can come out of the butt of some digital assembly line. I would argue software development is almost as sacred. However, broken down to their elements and production processes, are they computable? Maybe. But with differing results, like the machine generated burger from creator.
Can you hear the artists in coffee shops organizing protest rallies around the world? Not yet. Some will check out the new tools, create the new art, and get more work. Some will continue to make a living creating art the way they do now. Some will use existing methods and materials, sell it at a street fair or in a gallery (hoping no protester throws soup on it) The artists will choose. But they should have a voice and there should be strides in making sure it’s ethical and rewarding those who should be compensated so their art continues.
The world of software development did not end when the Integrated Development Environment (IDE) was introduced. Some developers continued to use traditional tools like Vim and Emacs, while others embraced IDEs, it’s a tool. Texas Instruments did not make me a math idiot. It just prevented me from wasting hours in solving a physics question. I can still calculate my tip at a restaurant without a calculator. New processes and tools have and will continue to improve, albeit with the potential of unforeseen consequences and objections.
What’s the message
In addition to our media being entertainment, it’s also educational. Attending Sundance for many years introduced me and my family to many worlds and stories foreign to me; extending our education, outside of the mainstream. The media impacts our beliefs, understanding and ultimately how we live in the world. The medium is the message. Having embraced decades of new personal media from email, text, social media, etc evolving toward a more connected world. Well, this is the new media, with AI providing the tools. Done right, it can be a tool for educators to craft a better lesson.
The Tools of the Digital Future
The use of AI can be a tool that will improve and advance content creation, not just for influencers, but everyone, including business. A couple of decades ago, one could not imagine the concept of social media being as influential as it is in society today. The mass ideology of previous generations, restricting who and how messages were distributed are long abandoned. The gap between broadcast and individual communication has been filled in, by the Internet. We live in a multi-channel world where the masses also create the content that the masses consume. However, how long does it take to create content, how expensive are the tools/labor to produce it, and what is the quality of the finished product? With trusted, safe AI tools, the amount of quality content will grow to meet continued demand.
How soon until the AI effect?
One day in the nineties, my mother called. She was very concerned that the Michelangelo virus was spreading through computers and that I needed to take every necessary precautions not to catch it from my computer. Decades later, she would be known as the computer lady, helping other seniors email their grandchildren and post on Facebook. There is a point when that new technology of yesterday is now just part of our lives, in fact we take it for granted. How many spelling mistakes in this document have been autocorrected? Too many to ever confess to my grade school English teacher, who would likely have a heart attack. Like any technology, there will be an evolving time of acceptance. And one day, it won’t be science fiction any more, in fact, many won’t even call it AI anymore and hence the AI Effect. Until then, there is a long road ahead.
Did I really need to write this?
Perhaps I should have asked ChatGPT to write this for me. Lets see how well it does:
Hmmm, not bad! However, it lacks style. For the record, I wrote this before I asked ChatGPT. Where I can see this really being helpful is when you need to break out of the blank page block when you’re on a deadline. Hmm, yeah, that’s one way I could write it, but actually, it needs… and away you go.
Next time, I’ll be digging deeper into tools to create videos.
Regards,
MikeC.