The State of Media & AI

It’s hard to believe nearly a year has passed since Generative AI’s coming out party. The WGA and SAG-AFTRA strikes had AI at the center of their negotiations. Industry conferences, including the recently concluded IBC 2023, spent many hours discussing, publicizing, and revealing Generative and Predictive AI products. The investor community is intensely focused on it, investing a reported $91.9B in the industry last year. Academics and regulators try to understand and predict AI’s impact and influence on society, and critics wonder if Gen AI might be this year’s NFT. AI certainly has people’s attention.

Attention is one thing, but utility is another. The questions people need to ask are: how useful is it, and what are the risks? We’ve written before on AI in M&E here, here, and here, discussing many of these issues. AI was a significant topic of discussion at IBC, and here are a few observations on the state of AI in M&E in 2023.

The Promise of AI for M&E

John Footen, Managing Director for Media & Entertainment at Deloitte, was one of many IBC 2023 speakers to express the industry’s hopes for AI. “AI has possibilities in production/post-production and distribution. An AI that creates or adjusts content based upon who is in the room, the story, the culture, and the language has a future in the industry.” He said Generative AI “could” blur or remove blood from horror movies to allow it to show in markets where such content is prohibited. Thus, AI could simplify creating optimized releases for any market or platform and be built into smart TV operating systems.

Other AI applications are aiding the targeting of specific advertising based on who’s watching a show, analyzing data to make better content recommendations to consumers, translating languages in real-time, and generating new audio, video, and effects content creation on demand.

Some of these technologies are already coming to market. The new Ryoo Seung-wan film “Smugglers” uses Flawless’ Generative AI tools, referred to as “visual translation,” to graphically modify mouth movements to precisely sync the English dub onto the Korean actor’s faces; it is the first full-length film foreign-language application of the technology.  

AI Limitations

Clint Eastwood’s iconic line from the film “Magnum Force” was, “A man’s got to know his limitations.” This admonition certainly applies to the adoption of AI technology, especially now. While products are being released that use AI, experts warn relying too heavily on new technologies presents risks that must be thoroughly understood before making a significant commitment.  

Garrett Goodman from Papercup put it this way. “The barriers of entry (for AI) into localization are low,” he said, “because there are plenty of models available for simple language translation. But it’s more complicated than that. The source data used in training is an important factor because, with incomplete models, it’s easy for a machine to miss something a human would catch instantly.” He went on to say that even with their investment in developing smaller, specialized models using voice actors and experienced translators, when it comes to localization, “the human touch” remains a critical component of their product delivery.

The new WGA agreement establishes guidelines for the use of AI in the writing process and recognizes both its challenges and risks:

1.     AI can’t write or rewrite literary material, and AI-generated material will not be considered source material.

2.     A writer can use AI when performing writing services if the company consents and provides that the writer follows applicable company policies. Still, the           company can’t require the writer to use AI software (e.g., ChatGPT) when performing writing services.

3.     The company must disclose to the writer if any materials given to the writer have been generated by AI or incorporate AI-generated material.

4.     The WGA reserves the right to assert that exploitation of writers’ material to train AI is prohibited by MBA or other law.

The Early Results Are In

AI has begun to transform our lives and our businesses. The extent to which AI impacts M&E has yet to be determined, but, as we have seen, content localization is one sector already acknowledging the benefits. As a provider to the industry of manual content analysis and age-rating services for years, Spherex recognized the promise AI/ML had in improving efficiency, expanding territories, and enhancing workflows. After three years of development, Spherexgreenlight™ was released in October 2021 and revolutionized how content is prepared for international distribution. Since then, the Spherex AI platform has won numerous industry awards and has been approved by regulators worldwide as the one public tool available to any content company seeking regulatory approval for any film or TV title.  

If you want to learn how AI can help you distribute your content globally, please click here or email sales@spherex.com