TV makers are taking AI too far
Technology
News

TV makers are taking AI too far

TH
The Verge
about 20 hours ago
Edited ByGlobal AI News Editorial Team
Reviewed BySenior Editor
Published
Jan 8, 2026

This is Lowpass by Janko Roettgers, a newsletter on the ever-evolving intersection of tech and entertainment, syndicated just for The Verge subscribers once a week.

Every year, TV makers flock to CES in Las Vegas to show off bigger, brighter, and better-looking displays. And every year, the same companies also use the show to throw a bunch of spaghetti against the wall as they try to figure out how to sell those big TV sets to consumers busy watching TikTok videos on their phones.

In recent years, TVs have gotten cameras for video chats and AI-powered workouts. They became cloud-powered game consoles, smart home hubs, and art displays. Two years ago, Samsung even wanted to convince people that their television would be a great telehealth platform — for pets.

I’m in Las Vegas this week, walking the show floor and talking to industry executives to find out what’s next for TVs. Not in terms of picture quality and quantum dot technology, but with regard to the apps running on these devices, the platforms making them work, and the bells and whistles manufacturers resort to to make us look up from our phones.

It’s been a rough few years for the TV industry. Covid messed with supply chains, tariff threats added a lot of uncertainty, and fears of an economic downturn have consumers rethinking big purchases. As a result, TV sales have stagnated. Global TV shipments declined 0.6 percent year over year in Q3, according to Omdia.

In that environment, many TV makers are betting on what works, which is why everyone is now making art TVs. Samsung invented the category with its Frame TV close to a decade ago, and surpassed 1 million in annual sales of art TVs in 2021. Since then, Hisense, TCL, and many others have copied the concept. Just this week, Amazon introduced its own Artline TV at CES.

“TVs are often big portions of the rooms they’re in,” says Fire TV VP Aidan Marcuss when asked about the success of this category. “I think these devices look great in those rooms. They become a part of the furniture of the room.”

Amazon also came to Vegas to show off its newly revamped Fire TV user interface, which will launch on a bunch of Amazon-made devices next month and expand to additional TVs later this year. It’s the first major revamp of Amazon’s TV interface in years, and a lot has changed in that time. Consumers are subscribing to more streaming services than ever, and companies like Amazon keep adding their own features on top of that.

“We’ve been adding things in terms of capabilities,” Marcuss says. “Games, art, music. [It’s] hard to organize that.” The new UI is meant to remove a lot of that complexity, he explains.

How do you keep TVs simple as streaming gets more and more complex? It’s a question many TV makers are trying to tackle, and the industry is increasingly looking to AI for an answer. The latest iteration of Google’s Gemini on TV, for instance, not only tells you what to watch next after you finish a show, but also patiently explains why you should watch Severance if asked to do so. Alexa Plus lets you add titles to your watch list, and LG even switches the viewing recommendations on its TV’s homescreen to those of an individual family member based on voice recognition.

But TV makers’ use of AI doesn’t stop there. Samsung used its First Look event Sunday night to paint a vision of the future in which you’ll ask your TV for anything from sports predictions to recipes — recipes that you then share with a display in the kitchen, because who needs cooking instructions in the living room. Hisense demonstrated an AI integration that identifies products featured in select shows, and then simplifies ordering from your phone with a QR code.

Amazon showed off Alexa Plus on TVs, complete with the ability to jump to scenes just by describing them within a Prime Video movie. And Gemini’s new features also include the ability to create custom “deep dives” on a subject and, for instance, generate age-appropriate explanations of the solar system for your kids.

Google also leaned heavily into generative AI by bringing both Veo and Nano Banana to TVs. In a demo I saw, someone asked Veo to generate a video from a photo of a woman on a beach. After prompting it, we had to patiently wait about two minutes, only to see her awkwardly come to life for eight whole seconds. Why anyone would want to do that on their TV is lost on me, but the company clearly thinks that there’s a use case for it.

A lot of the new AI features shown at CES also suffered from serious screen real estate creep. Time and again, booth personnel would ask a TV a question about something that was happening on screen, and the video would get minimized, with sports scores, knowledge facts, and other information taking up precious space on the screen. As noticeable as this was on the 70-inch-plus demo displays, I imagine it will be a nightmare on a screen sized 50 inches or less.

Hisense inadvertently made that case with an internally developed prototype AI agent capable of displaying stats for every player during a soccer game. This required so much screen real estate that the company also built a prototype 21:9-sized TV for the showcase. Here’s another idea: Why not let people look up that kind of info on their phones, as they’ve been doing for years?

Some AI integrations shown in Vegas were admittedly pretty useful. Gemini will let you change the display settings on supported TVs, and Samsung and others also use AI to tweak the sound on the fly. But these enhancements are largely in service of, and not distracting from, actual TV viewing. That may ultimately be the best use of AI on TV: something that’s working in the background to help you get more out of the onscreen action.

It’s a lesson that also rings true for TV makers and their eternal quest to figure out what’s next for their products. Sure, TVs are increasingly becoming part of the smart home, and cloud gaming may have its place on the big screen. But what people really want from these big screens is often quite prosaic: “Primarily, consumers are using their television to watch content,” says TCL sales and marketing exec Chris Hamdorf. “That’s still where the majority of the usage is.”

In other words: Even in the age of AI, the real killer app for TVs may just be watching TV.

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

Editorial Context & Insight

Original analysis & verification

Verified by Editorial Board

Methodology

This article includes original analysis and synthesis from our editorial team, cross-referenced with primary sources to ensure depth and accuracy.

Primary Source

The Verge