As we enter the early part of a new year, we are also fairly new arrivals in an era of an exciting but somewhat unknown, often misunderstood, technology. The technology is AI, artificial intelligence, and already it is nearly indispensable in our world, playing vital roles in everything from the arts to zoology.
You’ve probably used AI several times today and not even thought about it. (Hint: Siri, Alexa, Google.) Its application grows by the day, sometimes by the hour. As mentioned, AI has become ubiquitous in both good and bad ways.
Last summer when writers and actors went on strike against the television and movie industry, AI was a major sticking point in contract negotiations. Both groups denounced it as an ‘existential threat,’ to their livelihoods. It’s easy to see why.
Metropolitan State University-Denver journalism professor, Shaun Schafer, understands the concerns but says the fear needs to be balanced with a strong dose of reality.
“I don’t sit around concerned about ‘HAL 9000,’ said Schafer, professor of journalism at Metropolitan State University-Denver. Schafer’s reference is to the on-board computer in the 1968 movie, ‘2001, A Space Odyssey.’
In the movie, HAL evolves from onboard computer running basic systems into a sentient and malevolent computer who takes over the ship and mission.
To many, there is a looming concern that AI will follow movie HAL and do things that it shouldn’t, as writers and actors warned about in negotiations.
AI can write just about just anything that a human can, including movie scripts, songs and books. Taking the human element out of these undertakings means more money for the studios and bupkis—nothing at all—for writers. In many ways, it can be argued, it can also remove the soul of the work.
The actors’ big concern was that studios, using computer-generated images, could create ‘virtual’ actors with likenesses identical to well-known stars. A Hollywood blockbuster could feature your favorite ‘actor’ who is nothing more than CGI, a computer-generated image. An AI-generated Brad Pitt, for example, would be paid the same as ‘Woody’ from ‘Toy Story.’ There would also be no residuals. You do the math.
“It really is a scary thought to think about your image and sound” being used without your knowledge or permission, said the MSU-D professor. An AI ‘actor’ in a movie or a song performed by computer-generated software remains a real possibility. There are already existing laws addressing this reality but more will certainly be needed.
Recently, something similar did occur, something technological soothsayers had warned about—and on the pages of an iconic sports magazine.
In November it was reported that the ‘bible of sports magazines’, Sports Illustrated, was regularly filling pages with AI generated content and attaching fake by-lines to the stories. Authentic Brands Group, a company that buys underperforming brands, bought SI in 2019 and had outsourced much of its content. In acknowledging the journalistic faceplant, it said it had terminated its agreement with the company providing CG content.
The magazine acknowledged its cost-cutting and journalistically sin and removed the material from its website. Last week, the magazine also announced a “massive layoff” of staff, including most of its reporters. It’s future is undetermined.
While there is dark side to AI, as SI’s unforced error underscored, its presence, said Schafer, may not actually be all that foreboding. AI’s been around for decades but not nearly as dramatically or as technologically fine-tuned as it is today, he said.
“AI is machines learning trends and patterns that will have meaning to an audience,” he said. Basically, computers can perform complex tasks that, before, only a human could, including reasoning and making decisions. It’s the march of technology.
As an example of everyday use of AI that everyone can understand, Schafer pointed to something called predictive text, a tool computers automatically use to suggest what the next letter in a word should be or even what the next word in a sentence should be. It’s also seen regularly when calling up a favorite website.
While AI is already extensively used in medicine, future possibilities are endless. It already has a place in diagnosis and treatments of patients. Using current data, AI may soon be indispensable in formulating new drugs and protocols. Some have predicted that in the future, it may even lead to the prevention of illnesses or diseases. Imagine a cure for the common cold.
In a study recently published in the Journal of the American Medical Association, researchers praised AI for helping advance imaging and improving diagnosis of certain cancers and heart problems.
Schafer said dependence on AI will continue to grow, adding that its climb won’t be dissimilar from the time we moved “from slide rule to calculator.” Once there, he said, “You will never want to go back.”
Schafer said AI has already and will continue to tempt some students to use things like ChatGPT, software capable of using human-like dialogue, in papers, theses and dissertations. Some, no doubt, will cross the line and not get caught. But Schafer said that group would not include his students.
“I’m not worried,” he said, “because I’m going to catch it.” Schafer uses software that can detect AI-generated content. And while he says it is remarkably accurate in finding suspicious work, another worry is having the software come up with false-positives. “People will be accused of something they didn’t do.”
The wonders of AI will continue to astound in new and even previously unimagined ways. But as technology charts these new frontiers, it will leave a footprint that is rarely discussed when this technology is talked about.
AI uses more than its share of materials—metals, plastics—and uses extraordinary amounts of energy for essential cloud storage, said AI expert and ethics researcher Sasha Lucchione in a recent TED Talk.
“Each time you query an AI model it comes with a cost to the planet,” she said. The carbon footprint, unless something can be done to address the enormity of the resources essential to power AI, we will potentially end up with a problem that even AI will be challenged to solve.
AI is here to stay. But, as with any new tool, learning to use it safely remains a challenge, perhaps a long-term one.