Hollywood’s Perspective on Existential AI Concerns

As discussions around AI’s potential dangers escalate, entertainment labor groups are urging everyone to focus on the immediate impacts already being felt. Several major studios have begun integrating AI into their production processes, but when approached for comments, they remained silent. However, organizations like the Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) and the Writers Guild of America East (WGAE) have voiced their concerns.

The AI tools currently being utilized in entertainment differ significantly from the more alarming AI systems that have led to recent whistleblowing and dire predictions. These entertainment-focused AI applications don’t seem to pose the same existential threats as those that prompted former Anthropic researcher Jacob Coxon to resign, or led alignment lead Evan Hubinger to warn that AI could potentially “kill all humans” within the next decade.

Nonetheless, AI models that create digital replicas of actors and mimic traditional filmmaking techniques do threaten certain jobs within the industry. Just as Hollywood unions pushed for protections and oversight regarding AI during the 2023 strikes, Coxon’s departure from Anthropic has inspired many AI researchers to advocate for stronger safety measures. Leaders from major AI firms, including Anthropic and OpenAI, have also called for a slowdown in the technology’s development. While some politicians have dismissed these concerns, others feel compelled to take legislative action. The unions’ push for strict regulations on generative AI hasn’t stopped the entertainment industry from adopting the technology, but it has provided workers a means to resist when they feel AI is being misused.

In a recent conversation with SAG-AFTRA’s chief negotiator Duncan Crabtree Ireland, he emphasized that the union recognized early on the potential impact of AI on its members’ jobs. While many may not have considered the possibility of studios fully replacing human actors with digital versions, SAG-AFTRA seized the opportunity to influence public perception of AI in Hollywood by addressing the issue proactively.

“We negotiated and secured enforceable, meaningful guidelines on how this technology can be used by major employers in our industry,” Crabtree-Ireland explained. “While this doesn’t directly include companies like OpenAI or Anthropic, it does cover all businesses involved in creating professional entertainment content in the U.S.”

Crabtree-Ireland believes there are valid concerns regarding the broader societal implications of AI, but he also thinks some of the more extreme warnings about it being an extinction-level threat may be exaggerated. The doomsday rhetoric has become a common tactic for AI companies to generate interest, but Coxon’s departure has made Crabtree-Ireland reconsider. Hearing cautionary messages from the very CEOs and billionaires who have profited from the technology they now warn against has left him uncertain about what to trust. However, he acknowledged that Coxon has gained significant public attention.

One of SAG-AFTRA’s primary objectives during the strike was to ensure that their concerns about AI were taken seriously. Crabtree-Ireland noted that a clear, focused message about the stakes involved was crucial in rallying public support and outlining steps to avert worst-case scenarios. Without this approach, he believes the Alliance of Motion Picture and Television Producers (AMPTP) might not have agreed to key demands, such as requiring studios to obtain permission and compensate actors for using AI-generated replicas. “If that principle hadn’t been established, the economic incentive to use replication and replace human performances would have been much stronger,” he said, potentially leading to a rapid adoption of digital replication. Without these protections, Crabtree-Ireland is confident that entertainment workers would be facing a much graver situation today.

WGAE executive director Sam Wheeler was more direct in his criticism of the “tech oligarchs pushing AI on our society,” describing their recklessness as matched only by their dishonesty. While he finds it difficult to take AI companies’ warnings seriously, he pointed out that the WGAE has long cautioned that unregulated AI poses a significant threat to workers, democracy, and society at large.

“Guild members have successfully negotiated meaningful AI protections into their contracts; in 2023, WGA members spent 148 days on strike before studios agreed to implement real safeguards for writers,” Wheeler stated. “However, collective bargaining cannot replace government regulation, and we appreciate states like New York for enacting meaningful regulations in the absence of federal action.”

Wheeler highlighted the disempowerment of workers, the degradation of the internet, the spread of misinformation, and rising energy costs as tangible societal issues exacerbated by AI. He urged policymakers not to be sidetracked by AI companies’ apocalyptic claims, emphasizing that the real problem lies in irresponsible human behavior.

“These harms stem from decisions made by people,” Wheeler asserted. “The WGAE believes that addressing these concrete issues and holding AI companies accountable for the damage their products cause is the best way to prevent any worst-case scenarios.”