Netflix has reportedly used what seem like AI-generated or manipulated photos in a latest true-crime documentary.
“What Jennifer Did” is a few murder-for-hire plot involving a girl named Jennifer Pan that came about in Canada again in 2010. The streaming service allegedly used pictures to depict her character, which her highschool pal Nam Nguyen described as “bubbly, happy, confident, and very genuine.”
At across the 28-minute mark of the documentary, a number of photos present Pan with twisted and distorted palms and fingers, misshapen facial options, objects within the background which have morphed, and an excessively lengthy entrance tooth.
![Screenshot of Jennifer Pan from Netflix documentary "What Jennifer Did". A young woman in a red dress stands in a room with a playful expression, making peace signs with both hands. The lighting casts colorful flares across the room, and there are various items and pictures on a dark shelf behind her. "NETFLIX" is displayed in the upper right corner, suggesting this is a screenshot from a Netflix production. A close-up of the same young woman, smiling broadly with her head tilted slightly. She's wearing a red dress, dangling earrings, and has her hand on her shoulder. The background is dark with a few light orbs, indicative of a low-lit environment. "NETFLIX" is visible in the upper right corner, indicating this is also from a Netflix show.](https://readwrite.com/wp-content/uploads/2024/04/Screenshot-2024-04-17-100534.png)
However, in later footage, the options have altered, for instance, the beforehand lengthy entrance tooth now seems uniform. If true, the report raises critical considerations about the usage of such photos in documentaries, particularly for the reason that suspect depicted is presently in jail awaiting a retrial.
![Screenshot of Jennifer Pan from Netflix documentary "What Jennifer Did"](https://readwrite.com/wp-content/uploads/2024/04/Screenshot-2024-04-17-100545.png)
AI picture turbines notoriously have a tough time making footage of human palms. It’s laborious to find out precisely what occurred throughout the modifying course of, however it’s potential that the producers used a single picture as a supply to create further “photos.” However, the ensuing output might be perceived as biased fairly than impartially presenting the case details.
Ontario’s Court of Appeal ordered a brand new first-degree homicide trial for the Toronto-area lady in 2023, because the trial decide mentioned the jury ought to have been given extra choices.
There has been backlash on-line, with one social media person accusing the service of counting on AI as an alternative of hiring staff to do the job. Futurism author Victor Tangermann acknowledged: “Resorting to the tech to generate pictures of a real person, especially of somebody who’s still in jail and will only be eligible for parole around 2040, should raise some alarm bells.”
Netflix utilizing AI for this documentary known as What Jennifer Did is so reckless. Wtf is flawed with these streaming providers? There are so many proficient journalists, graphic designers, producers, and so forth. Why not use A HUMAN to get this proper?
— Jurnee 🍉 (@enjoythejurnee) April 15, 2024
How does Netflix use AI?
Netflix makes use of AI to regulate what customers watch, what they like and what they fee extremely. Then based mostly on this info, it suggests different exhibits and flicks that customers are more likely to discover attention-grabbing.
The streaming service has a Machine Learning analysis hub to “explore innovative techniques for efficient estimation methods in predictive modeling, and how these models are applied in real-world, discrete survival settings.”
Netflix mentioned it’s specializing in off-policy estimation, essential for coverage analysis and optimization, addressing challenges like distribution shift and huge motion areas by way of revolutionary strategies just like the Policy Convolution estimators. It additionally explores uncertainty quantification, empirical threat minimization below adaptive knowledge, and environment friendly Bayesian deep studying strategies, aiming to reinforce threat enchancment and generalization ensures in machine studying purposes.
The firm has beforehand confronted backlash from Hollywood actors and writers after posting a job advert for an AI knowledgeable. The advert, providing an annual wage of as much as $900,000, exacerbated considerations amongst Hollywood unions in regards to the affect of AI on the leisure business and wages.
ReadWrite has reached out to Netflix for remark.
Featured picture: Netflix (Partial poster picture)