A group of consumers has begun using the latest AI tools to uninstall the "smart systems" originally built into smart TVs. Something interesting has recently been happening with large living room screens.

Since October of this year, with the explosive popularity of several open-source projects, a wave of TV modification has suddenly swept across social platforms.

In the past, to clean up the ad-laden default Android TV launcher, users had to look up cumbersome ADB debugging commands, download installation packages, and type code line by line.

Now, ordinary users only need to open Tencent WorkBuddy, Doubao, or another AI tool that supports local control on a computer, enter the LAN IP address of their home TV, and issue a natural-language command such as "help me install XX on the TV." The AI will automatically connect to the TV backend, invoke scripts such as CleanYourTV or Aerial TV from the open-source community, disable the dense film and television waterfall feeds and pop-up components in the factory system, and replace them with a clean desktop featuring only a top clock, a single row of app icons at the bottom, and a 4K aerial dynamic screensaver.

Clearly, the decline in living room TV power-on rates in recent years is not because the public has completely grown tired of a large screen with high image quality, but because people are tired of the layered advertising toll booths and labyrinth-like menus that unfold after turning on the TV, and of the various levels of payment required even to open an app.

When users are all using AI to subtract from TVs, the TV industry needs to pay attention to one question: in the AI era, what kind of TV do they actually need?

I. It Is Not That People Do Not Watch TV, but That They Do Not Want "Pseudo-Intelligence"

Why does a modern TV with increasingly high hardware specifications and the name "smart" force users to use AI scripts to return it to a "bare shell" state?

As early as July 25 of this year, a complaint by Luo Yonghao on Weibo had already struck a chord across the internet.

To make it easier for the elderly at home to watch TV, he successively replaced two smart TVs from mainstream brands and even specifically enabled the "elder mode" heavily promoted by the manufacturers. Yet the elderly still could not operate them independently at all. In the end, he had no choice but to spend a small amount of money to add a traditional IPTV set-top box that plays live broadcasts as soon as it is turned on.

Young netizens use WorkBuddy to flash Aerial TV. As a tech-savvy group, they can take the initiative to seize control. But the elderly and some middle-aged people who are not sensitive to technology can only play the role of the silent majority and exit from the ranks of TV users. This is forced.

The root cause of this dilemma is not that TV manufacturers lack software engineering capability, but that smart TVs in the traditional sense (probably smart TV manufacturers never expected to be labeled "traditional") suffer from a severe "smartphone porting disease"—directly stuffing the smartphone's single-user shelf logic into the living room's multi-person public space.

Smartphones are highly private "cybernetic organs," owned by a single person, with highly efficient finger-touch interaction. Information feeds and waterfall feeds can run smoothly on phone screens. This has created a lot of room for app development and corresponding forms of commercial monetization.

But TV is naturally the public common denominator of the family. The old lady wants to watch the opera channel in the morning, the grandson wants to watch cartoons after school, and the young couple wants to connect a game console or play 4K Blu-ray movies from a LAN NAS on weekends. You will find that most of the time the TV is only a medium; what users want is the content at the end.

However, most smart TV operating systems in the past were modified based on Android. Product managers turned the homepage into a dazzling supermarket shelf, eager to stuff hundreds of film and television recommendation slots and topic entries onto the first screen, while giving users only a cross-directional key remote control that relies on "up, down, left, right, confirm" to navigate. This is not only an interaction problem, but also weakens users' willingness to access other content through the TV as a medium.

Therefore, even though many TVs themselves have excellent hardware quality, users are still unwilling to face their awkward operating modes.

In addition, what manufacturers call "elder mode" or other modes does not avoid the dilemma of users walking into functional dead ends. At this point, "smart" means having to press the remote control many more times every time the TV is turned on and recognize several more pop-up close buttons. It deviates from user needs.

II. Fold the Interface, Connect the Space

The grassroots modification wave and public complaints from celebrities are actually sending the same signal to home appliance giants: it is time to change the interaction paradigm of traditional smart TVs.

Around the Berlin International Consumer Electronics Show (IFA 2026) in September of this year, Chinese TV manufacturers represented by Hisense and TCL densely unveiled their respective AI solutions. If we temporarily set aside hardware excellence and measure against users' real pain points from the perspective of operating ecosystems and interaction models, we can see that the industry is exploring two distinctly different yet complementary paths to self-rescue.

The first path is to use AI for "interaction subtraction," turning uniform and complex shelves into imperceptible, direct access tailored to each person.

The reason netizens embrace Aerial TV is essentially to use manual scripts to forcibly cut redundant layers. Hisense's JUOS, released on August 31 (on the eve of IFA opening) as the industry's first family smart companion-level operating system, attempts to automate "interface folding" from the bottom layer of the system. Hisense has established a very critical product dividing line in JUOS: smartphone AI remembers "one person," while the origin of TV AIOS must be "one family."

Relying on the Xinghai large model and the upgraded "Super Xiao Ju" system, JUOS attempts to give TVs visual, auditory, and environmental perception capabilities.

In the difficult case of elderly people watching TV, a truly qualified AI solution is absolutely not to make the elderly learn how to switch to "elder mode," but rather, when the elderly person walks to the sofa and sits down, the TV can actively identify identity and instantly hide irrelevant entries such as games and trendy toys, directly presenting frequently watched opera or satellite TV channels.

When the elderly forget the name of a drama and only remember "I want to watch that old TV series about road construction in rural northern Shaanxi" or a vague line, they do not have to struggle to spell it out in a pinyin initial search box. They only need to say it directly as if talking to a younger family member, and "Xiao Ju Smart Search" can cross the walls of different video apps, directly locate and play the corresponding series.

The primary function of AI on TV has never been to stuff more flashy chat boxes onto the screen, but to fold away the 99% of menu layers irrelevant to each family member. This is the core of effective subtraction.

The second path is to use AI for "scenario addition," awakening the TV's idle 22 hours and transforming it from a mere audio-visual player into a "spatial management hub."

Even a TV with the most peak image quality is often watched intently for no more than two or three hours a day for dramas or ball games. During the remaining twenty-plus hours, the TV can fully take on other tasks—but only on the premise that it does not disturb users.

At IFA 2026, TCL previously adopted the theme "AI Inspired Life" and brought its SQD-Mini LED flagship TVs together with the NXTHOME smart home ecosystem to the forefront, attempting to answer a question from another dimension: when people are not watching dramas, what role can the TV still play at home?

In the past, control of whole-home intelligence was scattered across various independent apps on smartphones. Adjusting color temperature, checking security cameras, and adjusting air-conditioning airflow often required frequently unlocking the phone. TVs naturally occupy the visual center of the family's physical space.

In TCL's screen-plus-ecosystem concept and Hisense JUOS's ambient-screen companion design, TVs are beginning to take over the environmental coordination authority of the living room:

During the day when no film or TV drama is playing, the large screen can automatically cycle through highly textured artworks or dynamic light and shadow according to the real weather outside the window, the changing seasons, and indoor lighting—just as netizens are addicted to Aerial TV's aerial screensaver—making the TV a breathing digital window;

When the oven at home finishes baking, the washing machine finishes its cycle, or a visitor rings the doorbell, the TV screen can unobtrusively report the status of all home appliances through lightweight cards and, at a family member's one-sentence command, coordinate adjustments to lighting, curtains, and room temperature.

There are many possible ways to activate the physical presence of the large living room screen, and the core of this process remains product experience design. For leading domestic TV brands, the present is very likely a major opportunity window to fully understand home scenarios with the help of AI technology.

III. Can AI Save TV? The Premise Is That Manufacturers Dare to Change Themselves

Whether it is Hisense defining JUOS as a sensory, communicative family smart companion, or TCL using top-tier Mini LED screens to connect the whole-home ecosystem, both prove that leading manufacturers have keenly perceived users' aversion to traditional waterfall-feed desktops.

However, technology is always only the foundation. What ultimately determines whether AI TVs become a "true necessity" or degenerate into a "new gimmick" is actually a self-revolution concerning business models.

Domestic smart TVs are often compared with foreign brands such as Sony, leading many users to conclude that they are "bloated." But in fact, this is similar to smartphone advertising. The core crux lies in the low-margin competition of hardware, under which manufacturers rely heavily on back-end revenue such as boot ads, desktop recommendation slot buyout fees, and multi-tier membership commissions to subsidize hardware profits.

Unfortunately, in the face of continuously accumulating commercialization demands, some TV brands have not kept up with changes in user needs in software design and have only seen the stacking of functions and entry points. This is how the story of users using AI to refresh their experience came about.

The reason netizens would rather do it themselves and use WorkBuddy to flash the open-source CleanYourTV is that open-source code has only user experience in mind and will not peddle memberships to users. In the future, proactive AI recommendations will no longer disturb users like Internet-era sales pitches, but this also requires reordering the relationship between business models and user experience, which will test manufacturers' wisdom and courage.

Looking at the present, this living room upheaval jointly driven by the grassroots AI modification wave and major manufacturers' AIOS upgrades is pushing the TV industry toward three potential directions of evolution:

The first is "de-shelving" at the operating system level. Is the grid waterfall-feed desktop really the best form from the perspective of user experience? Many AI applications currently advocate new interactions such as voice-based control. Is it necessary for TVs to follow suit in depth? Should future TV homepages either maintain extreme visual restraint and purity like Aerial TV, or "change with people and be generated instantly according to intent" as envisioned by Hisense JUOS? The problem of menus with no findable entry points will next be solved through multimodal perception and natural language processing.

Second, all-weather spatial companionship will inevitably change some functions of the TV. Display technologies represented by TCL SQD-Mini LED have pushed peak brightness and light control zones to the marginal high point of human perception. The premium capability of future high-end TVs will no longer depend solely on the audio-visual shock of two hours of movie watching at night, but more on what changes it can bring to family life and how fully it can leverage its inherent advantages.

Finally, TV systems will move from relative closure toward open Agent collaboration protocols. When ordinary people can already use WorkBuddy on a computer to modify a TV across devices with one sentence, it means users are beginning to recognize this interesting approach of "smooth cross-device invocation." This happens to coincide with part of the philosophy of AI applications—as long as there is a network, AI agents can do too many things.

Perhaps future AI TVs can no longer draw boundaries around themselves. How content walled gardens and IoT barriers will change, and what kind of resonance Agents and the AI ecosystem should create with TVs, are all questions yet to be explored.

Everything ultimately comes down to the essence of user needs: people have never rejected a large living room screen that can bring gathering and relaxation. What people reject is only products that set up obstacles everywhere. Today, the forces of change from all sides are showing signs of convergence. The turning point for television may really be coming.

Source: Songguo Finance