The world of celebrity apologies has long been a tricky one. Fans often wonder if a star’s carefully worded statement truly comes from the heart or from a team of public relations experts. This ongoing debate about the authenticity of public apologies, especially those from famous people, sets the stage for a truly wild “what if” scenario involving the Writers Guild of America (WGA) and the future of AI celebrity apologies.
Real Fact / Baseline
The Writers Guild of America (WGA), a labor union that represents film, television, radio, and new media writers, has taken a strong position on the use of artificial intelligence. In its 2023 contract negotiations, the WGA secured protections on the use of AI from the writing and training processes of AI on WGA-covered works. These protections were designed to shield the literary landscape and the jobs of its human talent from the escalative potential of generative AI (WGA Contract 2023 on AI Protections).
The What If: WGA’s Revolutionary Demand for AI Celebrity Apologies
What if in a surprising turn of events or an unexpected shift of perspective, the WGA acknowledged that AI is more skilled at writing certain works than they are? You could think about a scenario where the WGA demands in an official statement that AI should be the one and only source of all celebrity apology statements in the future because it has “superior emotional authenticity”. And not because it’s cheaper or quicker. In that alternate reality, the Guild makes a daring, possibly sardonic statement about how human authors have mastered the art of generic, emotionally vacant writing.
Such a WGA-AI collaboration would undoubtedly make the news cycle (likely not for its intended reason, but still—publicity!). Brands and public figures across the globe might even adopt their programs en masse in an effort to weed out the holiest of unforced errors. Good or bad, we’ll all at least know what to expect with these perfectly neutral apologies.
Satirical Consequences
AI-led strategies for reducing facial recognition bias could be implemented. Tech companies might deploy GANs to imagine faces from historically marginalized population groups, which are subsequently used to update facial recognition models to be more inclusive. However, generating these faces can be tricky. Production has to steer clear of popular generative models designed to prevent facial information leakage. To get around that limitation, researchers are resorting to double-GANs, where one GAN generates plausible face candidates that can’t be reverse-engineered by the other GAN working to identify synthetic data. This short-circuits facial leakage, enabling research to go ahead on generating faces mathematically.
- The Rise of “Optimal Contrition Scores”: Instead of tailoring apologies, public relations would engineer the “right now” and the “right here.” They would input information about the celebrity’s offense, key characteristics of the offended group, and the intended emotional response from the public into the apology-writing AI to product a statement with the highest “Optimal Contrition Score.” The “Optimal Contrition Score” would quickly become the new standard KPI for apology effectiveness and public amends.
- Automated Sincerity Checks: Media and social platform companies rely on AI to automatically “audit” the AI-generated apologies for signs of insincerity. The algorithm may be able to detect the hallmarks of algorithmic sincerity and over time even detect specific models behind apologies based on their individual “apology styles.” This could result in public controversies when an apology is identified as “too perfect” or “insufficiently personalized,” with the public arguing that a humanlike apology can’t be sincere if it comes from an AI.
- The Birth of the “Apology Method Actor”: With celebrities currently getting flawless apologies written by an AI, they would hardly be able to deliver them with a straight face. Acting coaches would have to invent new forms of “Apology Method Acting” where celebrities would actually read the text generated by the AI and then try to channel the deep, artificial remorse the algorithm was going for. Some would probably try to “hack” the AI and suggest purposely bad input to get a more “human-sounding” flawed apology.
- WGA’s Existential Crisis: If emotional communication becomes outsourced, it will likely pose an existential crisis for the writers’ room itself. After all, with their aptitude for spinning a yarn and generating dialog, the room’s coffers could dry up if there’s no longer a need for members to pen tear-jerking apologies. One can only hope a new WGA subcommittee would form to make sure the AI doesn’t come after them next demanding an apology for having to churn out so many scripts.
- Technologizing Public Forgiveness: Beyond the aspects of rationalization discussed earlier, turning public forgiveness into a matter of evidentiary testimony further obviates the emotional burden on victims. If an apology-as-data produces a high OCS, passes the sincerity audit, and is entered as evidence, potential victims—and others who must decide to forgive or not—would be open to enormous pressure (and criticism) if they do not forgive.
- The unexpected consequences of this scenario are profound. The very nature of forgiveness would itself be subject to automation. For example, if an AI can perfectly articulate remorse, does that diminish the power of a human’s apology? Why forgive an imperfect and emotional human when a flawless, data-driven machine is an option? Overall, these examples demonstrate how an offhand remark about a nonhuman entity automating forgiveness can actually offer insights into very human social processes.
Dad Joke Punchline
Why did the AI say sorry to the calculator? It was a miscalculation!
Disclaimer
The year is 2025. It’s a breezy Tuesday afternoon. You’re driving home from work in your electric car. There’s been a lot of talk about how driving is the new smoking, so you make a mental note to download the latest software updates once you get home. The car suddenly comes to a halt. Peering through your windshield, you see hundreds of beeping, whirring, and buzzing robots crossing the road. Most of them are shorter than your car’s bumper. A few look vaguely like big sticks. One is shaped like a box with a tank tread, trying and failing to bulldoze its way over a parking meter.