The Dystopian Future of TV Is AI-Generated Garbage

TCL's new AI-generated movie "Next Stop Paris" is the next evolution in the algorithmification of TV.
The Dystopian Future of TV Is AI-Generated Garbage

On Friday, TCL, the second largest TV manufacturer in the world, announced Next Stop Paris, a romantic comedy that looks so bad and is filled with so many jarring visual errors common in AI-generated images, it’s hard to believe it’s real, but TCL insists that it is.

It is not clear whether the trailer is bouncing between different characters, or if TCL has been unable to figure out how to keep them consistent between scenes. The lip-synching is wildly off, the scenes are not detailed, walking animations do not work properly, and people and environments warp constantly. 

“We are not looking to replace people. This is still very much a human practice,” Chris Regina, the chief content officer for TCL, told Broadcast+Cable. “We’re using writers. We’re using actors and animators. We’ve been really fortunate to find a lot of great talent out there who weren’t working that have become part of our team.”

People will likely watch this for laughs, but this is the present and the future of a business model in which TVs have ceased being rectangles designed to let you watch ad-supported programming that costs a lot to make and have started to become rectangles designed to collect information about you so that you can be fed cheap content and targeted ads.

Next Stop Paris is being released by TCLtv+, which is a free digital TV service that comes embedded on TCL TVs. Similar services exist on most smart televisions at this point, and are the primary means through which “FAST” channels, which means “free, ad-supported TV,” are distributed. Not everything on FAST channels is low-quality garbage, but much of it is highly specific, allowing advertisers to more narrowly target viewers. On the version of “Digital TV” that I’m scrolling through right now (on Roku), for example, there is a Fear Factor channel, a Keeping Up With the Kardashians channel, a Duck Dynasty channel, an Ice Road Truckers channel that plays nothing but those shows 24/7/365 until, presumably, the heat death of the universe. All of this has happened alongside streaming services choosing to commission, renew, or cancel shows based on highly specific user watch data, and going down a path of creating increasingly specific shows that then show up to specific subsets of audiences.

TCL has not said what channel Next Stop Paris will show up on, but it currently operates a channel called “TCL Romance,” and the company told Broadcast+Cable that it sees a gap in the romcom market. It is easy to imagine that FAST TV channels will begin shitting out AI-generated movies and television shows at very little cost—and certainly at a lower cost than it takes to license real television. Whether on TCL TV+ or on another competitor’s platform, there will probably soon be entire channels full of AI-generated content.

"The TCLtv+ team applied the best creative practices from TV & film into our workflow to create Next Stop Paris.  It is very much a work in progress but we are thrilled to be a pioneer in the AI entertainment space," a TCL spokesperson told 404 Media. "Next Stop Paris will likely air on demand on AVOD, possibly across some of the FAST Channels and O&Os. Too early to talk about shows and movies, but originals are part of the TCLtv+ program strategy." TV today is full of acronyms: AVOD is "Advertising-based Video on Demand," O&Os is "Owned and Operated," meaning channels TCL owns and operates.

When TCL launched TCL TV+ last year, it said it would stand out from competitors with something it’s calling “IDEO” technology, which it pitches as an “individualized way of consuming entertainment” that includes what sound like on-screen chatbots and characters, as well as “suggested viewing” recommendations tailored to users. That press release then contains the following passage: “TCL is working with streaming solution partners such as Amagi, XUMO, Wurl, OTTera, and Future Today for content launch and distribution, while other key players will be added over time.”

Both Amagi and Wurl are adtech companies that talk extensively on their websites about the personalization of FAST TV services and a goal of using data collection to target viewers and turn watching TV into a more feedlike, algorithmic experience. 

Wurl is a streaming TV ad-targeting platform owned by the mobile ad company AppLovin, which advertises all over its website how it is working to make digital TV less onerous to navigate (“AI will bring more personalized connected TV viewing,” it writes) and more lucrative as an ad platform by using data collection, targeting, AI, and “emotion analysis” to highly target ads and programming. A blog post by CEO Ron Gutman from last month titled “Now it’s clear. Machines understand human emotions better than humans” is full of brain diagrams, human hormone descriptions, and musings on the nature of consciousness. It ultimately makes the argument that “for advertisers, multimodal large language models introduce a powerful opportunity to leverage GenAI to engage with audiences in new ways based on their thoughts and emotions.” 

A Wurl product called “ContentDiscovery,” meanwhile, is a “solution” to the paradox of choice in which “machine learning helps viewers discover content they may never have considered watching, with relevant ads that continue to draw viewers back to the streaming service or channel again and again, driving higher retention and creating a “virtuous cycle” where people watch for longer, Wurl advertises: “If a streaming service displays an ad for something a viewer will like, that person will be more grateful and happy, rather than feel annoyed by the ad. And grateful and happy viewers are good for the ecosystem, helping complete the virtuous circle where everyone wins.” 

Amagi also argues that personalization and data collection is changing how people watch TV, and writes in a white paper that “As FAST services begin to leverage viewer data like location, device ID, IP addresses, etc., the possibilities to deliver personalization, drive engagement, and monetization increase.” 

A screenshot from Amagi's white paper explaining the types of targeting it wants to do.

“With personalized viewing, unlimited content can be made more consumable, thus opening up the floodgates for diverse content providers to take part in the FAST ecosystem,” it adds. “Imagine a viewer tuning into a FAST service and being presented with personalized content … the viewer lands on the content that is already playing instead of going through the tedious process of searching across 300+ channels.” 

One of the supposed utopias AI-maxers believe will soon happen is that we will have the ability to ask personalized AIs to generate us new, personalized episodes of TV shows, or to dream up our own shows about anything. What this looks like in practice for the foreseeable future is a whole lot of content like Next Stop Paris, where TV manufacturers, low-quality studios, YouTubers, and TikTok accounts spam us on every single platform with highly specific, niche content that costs essentially nothing to make and can be easily advertised against. It is appropriate that TCL, which went from being an essentially unknown company to one of the largest TV manufacturers of the world using a strategy where decent TVs were sold for impossibly low prices that then turned more profitable over time using advertising and user monetization strategies

Whether audiences or anyone at all consciously wants this is irrelevant: This content is being delivered directly to us using algorithms and personalized feeds that are based on data not just from our TV watch history but from our phones, locations, and more. This content is designed to briefly intrigue people as they scroll social media, outwit search algorithms, or passively wash over us as we turn on our smart TVs, which, again, have become surveillance boxes full of targeted advertising, pop-up ads, and sponsored content. 

A side wrinkle to all of this is that much of the news about this AI-generated movie is itself AI-generated, highlighting again the fact that Google News and Google Search are increasingly broken, and reminding us yet again that much of the AI content ecosystem is AI content about AI content written to game a search algorithm and make money from display ads. This is not the exact same thing that is happening with Next Stop Paris, but this type of algorithmic content ecosystem has been incredibly lucrative for Google and other tech giants, and TV manufacturers are now trying to replicate something like it on smart TVs.

When I typed “TCLtv+” into Google and searched “News,” the second-ranked article was on something called “YTech,” whose URL is (“Dynamic News” in Greek), running an article called “A Bold Leap into AI-Crafted Cinema: ‘Next Stop Paris’ Hits TCLtv+,” which has at least seven ads on it, was “written” by an AI-generated person who publishes more than 50 articles a day, and uses this as its header image:

We have said this before and we’ll probably keep saying this as long as we can continue reaching real humans on the internet: The current threat of AI is not that it is going to be used to make highly compelling content that will replace human jobs. Instead, it will be used to flood every possible platform with low quality bullshit that is impossible to sort through.