As we look back at 2023 in digital media and technology, one trend really stands out as the most widely discussed and pivotal for the future of the industry. Of course, I’m talking about Tubi’s ripped-from-the-headlines original films. Did you hear that they’re working on “The Rapper Who Got Shot in the Heel,” from the same creators as this year’s breakout sensation “The Nurse That Saw the Baby on the Highway”?
But other than that mega-trend, everyone was also focused on generative AI this year, both its promise for productivity and paradigm-shifting and the immediate threat it poses to employees across just about every sector and type of business.
Interest in the new generation of AI apps was immediate and obvious upon their arrival near the start of the year. Within its first five days of availability, the ChatGPT app already had 1 million users. The Midjounrey generative AI art app recently moved from a Discord server to its own dedicated website, open to anyone who has already generated more than 10,000 pieces of AI art. Which, it turns out, is a lot of people.
But beyond just interest from the public in new tools, apps, and toys, AI adoption is being most aggressively driven by employers hoping to scale down the size of their workforces and increase the productivity of their human staffers. While the World Economic Forum estimates that AI will create nearly 100 million new jobs, one report from McKinsey estimates that AI could displace 15% of the global workforce, or over 400 million people, by 2030. A full 97% of surveyed business owners told Forbes Advisor that they plan to use ChatGPT to improve their businesses, with a staggering 1-in-3 hoping that it will soon be able to create website content for them.
In media, this is no longer merely an abstract concept but a daily on-the-ground reality. This year’s devastatingly costly Hollywood strikes were in large part an attempt by writers and actors to prevent computer programs from taking over their jobs, at least for the next three years or so.
Meanwhile, scores of publishers, websites, and other companies downsized their actual human staff this year and, in many cases, have already tried to replace some of their flesh-and-blood writers with mechanized replacements. (In most cases, at least the ones that we know of, this didn’t go all that well.)
We’re already at a point when companies doing layoffs have to assure their customers that they don’t plan to replace the dismissed staffers with machines. Wizards of the Coast, which produces the “Dungeons & Dragons” role-playing game, assured fans this week that they don’t plan to use AI tools to generate artwork for their games after parent company Hasbro announced 1,100 new layoffs.
Beyond just losing out on jobs to well-trained simulators, creatives also raised alarms this year about the way the software gets its training in the first place. Most popular generative AI apps are trained by feeding massive libraries of copyrighted material. This just makes logical sense. If you’re going to build a generative music app that can write an original song sounding like Bob Dylan, it has to listen to the Bob Dylan catalog to determine what that practically means.
By their own admission, the companies that own these AI apps can’t afford to compensate all the creators of the material they use for training while still turning a profit. When the U.S. Copyright Office announced that it was considering new rules for training AI apps in November and asked for input from the industry and the public, a number of major tech companies — including Meta, Google, and ChatGPT makers OpenAI — responded that this was impossible for them to actually accomplish.
In its statement, Facebook’s owner Meta noted that “generative AI models need not only a massive quantity of content, but also a large diversity of content. Deals with individual rights holders would provide AI developers with the rights to only a minuscule fraction of the data they need to train their models. And it would be impossible for AI developers to license the rights to other critical categories of works.”
Clearly, Meta intended this as an argument against requiring them to license copyrighted material when using it to train the AI apps. “Want these apps to exist? Then you have to let us steal from everyone to make them!” But of course, there’s a counter-argument to be made as well. If the only way to build something new is by stealing, is it actually new in any meaningful way? Or are you just stealing from lots of other people and then saying their work is now yours?
Combine this with the recognition that the developers of generative AI apps don’t actually know how they work or how the computers are coming up with their specific results, and this question gets even more murky, even existential. If the only way to build a high-quality AI app is through theft, and the creators of said app don’t even know how it’s actually using the stolen material, what value do they bring to the table? Why do they deserve compensation for this product that they have neither obtained legitimately nor understand?
But perhaps the biggest and most dramatic argument against AI in 2023 came from internet creators themselves. This was a year in which mainstream culture was driven by social media, not the other way around. While a decade plus ago, the internet was mostly reactive, taking in what was coming out of Hollywood and the global entertainment industry and reworking it from a fan perspective, this year, brands were relying on fans to do the marketing and evangelizing themselves.
Actual corporations struggled with messaging throughout 2023 and often failed to get anyone hyped about their new products and ideas. Properties that would’ve once been surefire hits — like the new Indiana Jones, “Mission: Impossible” and Marvel films — withered away on the vine as franchise media culture all at once screeched to a halt. No matter what owner and CEO Elon Musk tried to drive interest and users back to his X/Twitter social media app, millions have nonetheless fled the platform, with seemingly no interest in returning. A simple Bud Light social media activation with influencer Dylan Mulvaney spiraled into a months-long controversy that ultimately saw its US revenue drop by 13.5% and resulted in the departure of Anheuser-Busch InBev’s chief marketing officer.
While the same marketing executives and middle management teams driving the “AI revolution” demonstrated a noted inability to connect with the public, everyday creators got the job done instead. The “Barbenheimer” meme craze led to both films charting in the US box office Top 5, while the virality around Jack Black’s song “Peaches” helped drive interest in the “Super Mario Bros” movie. As the music industry continued to struggle with launching new artists or fairly compensating the acts they’ve already developed, many social users agreed that Kyle Gordon’s satirical TikTok hit “Planet of the Bass” was the actual “Song of Summer” 2023.
A desperate Food Network is shelling out $100 million to Guy Fieri over just three years to help offset their massive advertising losses, while the year’s actual most talked about food reviewer is a soft-spoken MMA fighter and TikToker from Detroit named Keith Lee. McDonald’s ad campaign highlighting Grimace’s birthday was just another series of commercials until members of the public turned the themed berry-flavored purple milkshakes into a running gag.
This doesn’t so much seem like a situation in which humans are about to be wiped off the board in favor of apps. Quite the opposite, really. Apps are decent at soaking up millions of pages of content and then churning out something vague that kind of almost sounds like all of the material they were fed. No app has yet demonstrated the ability to produce something outside-the-box, imaginative, and silly as Grimace Shake horror parodies. You kind of need people for that.