|
|||||
Amazon's plan to offer AI-generated recaps of Prime Video shows isn't off to a great start. The company's recap of the first season of Fallout features multiple errors, GamesRadar+ writes, including basic facts about the plot of the show.You can watch the recap yourself in the "Extras" section of Amazon's Fallout season two listing in Prime Video. Besides being somewhat dry, the AI-generated recap incorrectly identifies the time period of the show's Los Angeles-set flashbacks as being the 1950s, when they're actually 2077 (the Fallout franchise is set in an alternate history that diverged from our real one after 1945). As Gizmodo notes, the recap also seems to misunderstand the ending of the first season, which sets up season two's partnership between vault dweller Lucy and The Ghoul, an irradiated wastelander with a personal connection to the mystery at the heart of the first season.While the recap suggests Amazon's AI system can successfully combine clips, music and dialogue into a coherent video, it apparently lacks an understanding of the details. The inaccuracies in this recap won't prevent anyone from enjoying the second season of the show, but they don't exactly inspire confidence in Amazon's tool either. It also seems like a problem that could have been easily solved by having a human employee who's watched the show review the video before it was uploaded.Unfortunately, Amazon's lack of AI quality control extends beyond recaps of its shows and into the dubs for the shows themselves. The company pulled AI-generated voiceover tracks for Banana Fish and other anime because of how bad they sounded earlier this week. It wouldn't be surprising if this recap gets pulled, corrected and re-uploaded, too. As Amazon adds more AI-generated content to its platform, users are bound to discover more ways it comes up short. The company's audience is too big, and AI is apparently still too unreliable for it to be avoided.This article originally appeared on Engadget at https://www.engadget.com/ai/amazons-ai-generated-recap-tool-didnt-watch-fallout-very-closely-215712958.html?src=rss
Category:
Marketing and Advertising
This morning Disney and OpenAI announced a three-year licensing agreement: Starting in 2026, ChatGPT and Sora can generate images and videos incorporating Disney IP, including more than 200 characters from the company's stable of Star Wars, Pixar and Marvel brands. To say these companies make for strange bedfellows is an understatement. The agreement brings together two parties with very different public stances on copyright. Before OpenAI released Sora, the company reportedly notified studios and talent agencies they would need to opt out of having their work appear in the new app. The company later backtracked on this stance. Before that, OpenAI admitted, in a regulatory filing, it would be "impossible to train today's leading AI models without using copyrighted materials." By contrast, Disney takes copyright law very seriously. In fact, you could argue no other company has done more to shape US copyright law than Disney. For example, there's the Sonny Bono Copyright Term Extension Act, which is more derisively known as the Mickey Mouse Protection Act. The law effectively froze the advancement of the public domain in the United States, with Disney being the greatest beneficiary. It was only last year that the company's copyright for Steamboat Willie expired, 95 years after Walt Disney first created the iconic cartoon.On the face of it, it's unclear OpenAI is getting much value out of the deal. As part of the pact, Disney will host a "curated" selection of Sora-generated videos on its streaming platform Disney+, legitimizing the medium of AI-generated video in a way it hasn't been before, but it would appear Disney has the option to spotlight as much or little of it as it sees fit.Additionally, the $1 billion Disney agreed to invest in OpenAI is a drop in the ocean for a company that's expected to burn through more cash in five years than Uber, Tesla, Amazon and Spotify did combined before they became profitable. If anything, the addition of Disney characters is likely to make operating ChatGPT and Sora more expensive for OpenAI; the company will now need to pay a licensing fee on top of the cost of running its servers to generate images and videos. At this stage, it's also hard to put a value on Disney's pledge to use OpenAI's APIs. The company has said those tools will "enable new products, tools and experiences," including some found inside of Disney+, but beyond that it hasn't shared specifics.Bob Iger might be feckless, but he's not stupid. Sometime this week or soon after, President Trump is expected to sign an executive order that makes good on part of his AI Action Plan from July. Specifically, the president has promised to fight against "burdensome" state-level regulation of AI. According to CNN, a recent draft of Trump's order calls for the creation of an AI Litigation Task Force to challenge and preempt state AI laws in favor of the president's own more lax regulatory regime. It's unclear how successful the administration will be in that effort, but clearly Disney is thinking ahead. It's banking on the fact that this time it won't be able to count on the federal government to shape copyright law in its interest, so instead it's making a deal with an industry pushing the boundaries of intellectual property rights as we know them. More importantly, it has partnered with the one AI company it can actually leverage. As I argued in a recent piece, OpenAI is in a far different and more precarious position now than it was at the end of 2022 following the release of ChatGPT. The company is just one AI provider in a sea of competition, and you can't even argue its models are the best, based either on benchmarks or user feedback. Moreover, OpenAI has yet to turn a profit, and has adopted an extremely risky investment strategy. In recent months it has signed more than $1.4 trillion worth of infrastructure deals, hoping to outmuscle the competition that's already beating it through scale. It's not an accident Disney sent a cease-and-desist letter to Google a day before its agreement with OpenAI became public. OpenAI might be the most valuable private company in the world, but Alphabet, Google's parent company, is worth more than $3 trillion. In any negotiations between the two, at best Disney would be on equal footing, and certainly not in a position where it could demand some amount of control over Google's AI projects. And yet by accounts it won exactly that from OpenAI. According to Axios, the deal gives Disney a fair amount of control over how its intellectual property is used. The two will form a joint steering committee designed to monitor the content users create on ChatGPT and Sora. As you surf the web today, you'll likely see a lot of opinions on how this legitimizes AI video. And while that's true, far more important is the fact Disney has secured a seat at the table to decide how the technology evolves over the coming years. Much like with news publishers, OpenAI and other chatbots concerns took a stance of begging forgiveness rather than asking permission towards copyright. It seems to have paid off. Most of the highest-profile news organizations have signed licensing deals to at least be paid a little rather than be ripped off until reaching an uncertain verdict in court. Disney seems to be signalling that the same speculation rush is about to begin for audiovisual licensing, and it may have already secured the most favorable terms. This article originally appeared on Engadget at https://www.engadget.com/ai/disneys-deal-with-openai-is-about-controlling-the-future-of-copyright-213009504.html?src=rss
Category:
Marketing and Advertising
A director who was charged with defrauding Netflix out of millions of dollars has been found guilty, Business Insider reports. Carl Rinsch, director of the 2013 Keanu Reeves movie 47 Ronin, now faces up to 90 years in prison.Rinsch began filming the project, White Horse (later renamed Conquest), around 2017. (Its premise: A scientist creates an organic humanoid species that turns on its creators.) The director completed six short-form episodes with his own money and investor funds. He then used those episodes to pitch studios for the money to complete the first season. Netflix ended up buying the rights for over $61 million.In 2020, after spending $44 million of Netflix's money on the series, Rinsch petitioned for another $11 million to complete the season. The company agreed.Then things got weird. Instead of using Netflixs investment to finish the series, Rinsch transferred the funds to personal accounts. Within two months, he lost over half of it on seven-figure stock trades. He spent the rest on cryptocurrency.Then, lo and behold, Rinsch got a second chance: The crypto trades turned a profit. Did the director seize the opportunity to right his ship and finish the series? No, he didn't. Instead, he used the crypto profits to go on a $10 million shopping spree. According to prosecutors, he spent nearly $4 million on furniture and antiques, $2.4 million on five Rolls-Royces and a Ferrari, almost $1 million on mattresses and linens and $650,000 on luxury watches and clothes.The series was never completed. By 2021, Netflix had canceled the project and written off over $55 million in costs.Rinsch's charges included one count of wire fraud, one count of money laundering and five counts of engaging in monetary transactions in property derived from specified unlawful activity. During the trial, he took the stand in his own defense, claiming that Netflix's millions were intended as reimbursement for the personal funds he invested in the series.After less than five hours of jury deliberation, Rinsch was found guilty on all seven counts. Although he could face up to 90 years, he's expected to receive a much shorter sentence.This article originally appeared on Engadget at https://www.engadget.com/entertainment/streaming/47-ronin-director-found-guilty-of-defrauding-netflix-out-of-11-million-205519293.html?src=rss
Category:
Marketing and Advertising
All news |
||||||||||||||||||
|
||||||||||||||||||