The Road to AI Everywhere Will Be Long, Winding and Loaded with Potholes
Content Insider #833 – Human AI
By Andy Marken – andy@markencom.com
“Ava was a rat in a maze. And I gave her one way out. To escape, she’d have to use self-awareness, imagination, manipulation, sexuality, empathy, and she did. Now, if that isn’t true AI, what the **** is?” — Nathan, “Ex Machina,” DNA Films 2014
If you’re a start-up in the AI field, VCs (venture capitalists) are throwing money at you.
If your company is using/has AI “built into” its product, customers are willing to talk with you.
In other words, if you’re not buying/developing/testing/using it, you are obviously a saber tooth tiger’s next meal.
AI has had a long and sordid history going back to WW II – The Imitation Game – and the work of Alan Turing considered the father of AI.
Starting out shrouded in mystery, the current threats/promises continue to be … mysterious.
AI has such a potential for good/evil and benevolence/parsimonious that even many of the leading AI experts and executives issued a subtle joint warning that the technology poses a “risk of extinction.”
Geoffrey Hinton, widely known as one of the “godfathers of AI,” has said that the technology poses a “more urgent” threat to humanity than climate change.
Based on this urgency and concern, more than 350 of these folks said, “As we grapple with immediate AI risks like malicious use, misinformation, and disempowerment, the AI industry and governments around the world need to also seriously confront the risk that future AIs could pose a threat to human existence.”
Allan McLennan, president/cga of 2G Digital Post Optimization, noted that while a united front like this is highly unusual, it also means that the responsible members of the profession feel it is vital that governments around the globe consider establishing guidelines for where/how/when/how AI can be used more than where it is already.
Now, we don’t want to cast aspersions at the folks who run your government, but the people here are old, and we mean old.
They’re just getting around to figuring out if search engines and social media sites are good/sorta good/bad, how much data they scrape every time you go online and how they use/monetize/share that data.
Degrees – The EU established rules/guidelines on the development/release of generative AI products based on the degree of risk involved to businesses/people if things went awry. Early products/implementations fell in the minimal risk category.
McLennan noted that while U.S. lawmakers are just now starting to address some of the issues, those in the EU are taking more of a progressive step and negotiating AI rules and China has demanded that AI systems adhere to strict censorship rules as a start.
Even while these actions/discussions are going on, AI system/program developers are setting up their own guidelines to reassure folks that their products are wholesome and designed to be/do good.
Ethical and conscientious developers are rushing to move generative AI products and tools into mainstream, organizations such as those in the entertainment industry. The goal is to streamline, improve and enhance products such as movies and shows to attract more of their target and peripheral audiences.
And to do it more efficiently and economically.
That’s one of the major issues that has hampered the AMPTP and WGA/SAG-AFTA negotiations where one side wanted to discuss issues/points later while the other side would like to establish guardrails and guidelines now … then hone the rules and applications as they arise.
Maybe we’re being a little egoistic or paranoid, but it almost seems that the key generative AI tools seem to be focused on the content creation industry – text generation, AI images, AI audio and video.
According to a study by Electronics Hub, they each have different levels of interest around the globe.
Textual Interest – Perhaps because “everyone” realizes that writing, good writing is hard work, text AI tools have been the ones tested and have increasingly been used.
Most of the noise and concern surrounding AI focuses on ChatGP and Quillbot – text generation.
The textual tools have gained interest because they can quickly carry out such tasks as data analysis and basic content generation. That way, creative professionals can focus on the more complex and imaginative aspects of the work. As a result, they can be more productive and more efficient in delivering a product that has just the right tone, style, storyline/dialogue for an audience captivating film/show.
But…
The finished AI-written project still needs human interaction as it lacks the emotional depth, intuition and empathy that is only derived by unique experiences, perspectives and cultural context.
Real content creation.
In other words, a script that is a flat, dry, mundane and safe regurgitation derived from the data scraped from everywhere isn’t really a finished script.
As poet/playwright Oscar Wilde said years ago, life imitates art far more than art imitates life.”
Or, as writer/photographer Scott Bourne explained the difference, AI can’t feel love, pain, emotion, has never been in a fight, has never felt the exhilaration of winning or the agony of defeat, or struggled to put thoughts, ideas on paper/screen and gone through the painful process of doing it again and again until it just “feels” right.
Or, as Ernest Hemingway is often credited as saying, “writing is hard work.”
But ChatGPT/Quillbot will be heavily used to flood the internet with AI-generated “facts”, “news”, “truths” and enough “half-truths/lies” to make people question even themselves.
A few months ago, Chinese hackers were using their AI tools to flood sites showing “proof” that the Maui fires were caused by the U.S. government. This “creative work” will only get better and more frequent.
However, industry folks’ biggest concern is the protection of IP (intellectual property – yours and ours.
AI “learns” and becomes helpful by scraping, digesting and theoretically using just the right information to suggest, assist, enhance and streamline certain aspects of creative work.
Every reputable writer and creator wants thire work to be uniquely his/hers and to have it protected.
That is a difficult job for cold, heartless code!
Imagery – Putting down a few words and having them “translated” into a picture or work of art is fun and can also be useful in such areas as ad development/production. Or just in descripting a picture you have in your mind.
Image generation tools like DALL-E2 and Midjourney can produce realistic pictures in seconds.
For example, content creators in Israel marked this year’s Passover with realistic images portraying tales of ancient stories from the Exodus.
ILM (Industrial Light & Magic) developed the high-definition LED video wall technology which has virtually replaced green screens. With a combination of Nvidia’s robust GPU technology and Epic Games Unreal Engine, they created ultra-realistic backgrounds first for Rogue One and later for the Disney+ series Mandalorian.
McLennan pointed out that LED video walls and image generation tools are regularly used in the major production houses around the world to help virtually create scenes and buildings and replacing the need for hand-painted backdrops and scenes.
The potential risks were on display as recently as last week, when a likely AI-generated photo of a fake explosion at the Pentagon triggered a selloff that briefly erased billions in value from the US stock market before it was debunked.
Music – Music artists have already tested the use of audio AI technology and received push back that it isn’t really human creation. But the biggest danger will come from people who create their own personalized versions of an artist’s work without the artist being paid.
Audio generation tools could be extremely costly … also to the music industry.
Ripping music from the internet, airwaves, discs (yes, they’re still bought) is just a copy but with audio AI tools, thieves are creating new stuff that sounds like the musician/singer and that means billions in lost income and a real muck-up of their image.
Early this year, popular musicians Drake and the Weeknd found a “version” of their song Heart on My Sleeve had been created, posted and downloaded by millions of fans around the globe.
With zero talent but tools like FakeYou, Uberduck.AI and others, these sleazes can copy everything that makes a music artist unique to create, enjoy/sell “hits” from every genre, every country, every artist.
Movie Production – Creating a movie or show using AI is presently difficult, slow and expensive but it will be easier, cheaper, faster in the future which has raised concerns throughout the content industry.
McLennan emphasized that jumping from a single image to those that move 20+ times per minute isn’t easy and isn’t cheap.
“The VFX folks are the experts in this area and are involved in the production of shows/movies. Ask them,” he said.
Fortunately, these people have an excellent set of tested/proven tools from firms like Adobe and Avid to help make their work a little easier; and for that matter, smarter.
YouTube, TikTok and other social media users have also begun to use new AI tools like InVideo, Synthesia, Veed.io and others to make their own passable video content.
But if you’re more interested in earning a dishonest buck, peso, pound, Euro, Dinar, Rail or other coin of the realm, you’re probably more interested in WormGPT and how to use it in your BEC (business email compromise attacks.
The dark web and cybercrime sites/forums have complete details or do what we did … ask a cybersecurity expert about it and avoid anyone “following you home.”
Don’t get us wrong. We’re not against AI.
Phone Home – Apple and all smartphone manufacturers have used AI for years to assist people in writing/sending texts. The results have been fraught with sometimes humorous errors but the companies continue to make steady advances in the technology to the delight of users everywhere. And yes, Tim Cook hopes their technology will play nice in the real world.
We’ve used it for years as we texted folks constantly on our phone as it anticipated our every thought and filled in words for us as we thumbed away.
Well, yes, there was a follow-on message that translated the stuff we sent and then reviewed to see what it said compared to what we planned to say.
But maybe the new iterations on this year’s refresh will be more closely attuned to us.
The major hurdle in the AMPTP, WPA/SAG-AFTRA negotiations center is on the use of the various areas of creative content development/creation/distribution.
The idea that the technology which can automate/simplify/diminish human skills will be disruptive and will threaten/impact certain segments more than others is very real.
But it can’t be ignored because it is coming.
Back in the 1800s, Ada Lovelace – considered the first computer programmer – equated computers to the Jacquard-loom and how textile workers smashed the new machines because they threatened their jobs.
“Generative AI will be widely used in the creative industry in the years ahead, impacting some specialties/jobs a bit more than others,” he noted.
.
“However,” he added, “it will also open up new opportunities for those same individuals to gain new experience/expertise and strengthen their value in the marketplace/industry too.”
All of the entities in the industry have a responsibility to learn, understand, use the technology to enrich, enable, entertain humans rather than allow people and stories to atrophy.
Can’t Ignore – The best way to understand the strengths/weaknesses as well as the benefits/dangers is for people to learn more about the technology, test it and use it because it isn’t going away. The more we understand it, the better we can use it rather than have generative AI use us.
Some folks in the US we have talked to, look at the Electronics Hub study and similar reports on interest levels, applications and business opportunities and are delighted to see how little understanding or interest there is in the tools and are delighted in the apparent lack of interest.
AI proponents and cheerleaders simply say, “Get on the train or get out of the way.”
That approach doesn’t work with your dog, your cat or your kids.
It’s not going to work with adults or semi-adults!
AI has the potential of enhancing and enriching the workflow, workload and work product in the content creation (and every) industry, McLennan pointed out.
It also presents a lot of risks including the wasting of the industry’s budgets – time/money.
AI is progressing slowly, and it will take years, decades to progress from pockets of applications to broad usage and produce its potential benefits.
Because right now, we don’t know what we don’t know.
The script will be written as we move forward, stumble, improve, progress and eventually can reach the same point Caleb did when he said in Ex Machina, “Yeah. I know what the Turing Test is. It’s when a human interacts with a computer and if the human doesn’t know they’re interacting with a computer, the test is passed.”
Because to paraphrase Ex Machina’s Nathan, no matter how rich/powerful you become, s**t goes wrong and we can’t insulate ourselves from it.
During this creative content evolution, people – at every level – have to distinguish fact from fiction.
Andy Marken – andy@markencom.com – is an author of more than 800 articles on management, marketing, communications, industry trends in media & entertainment, consumer electronics, software, and applications. An internationally recognized marketing/communications consultant with a broad range of technical and industry expertise especially in storage, storage management and film/video production fields; he has an extended range of relationships with business, industry trade press, online media, and industry analysts/consultants.