I went traveling during National Day and spent half an hour in the hotel room struggling.

I studied every button on the remote control, gave up in the end, and lay back on the bed to finish watching the drama on my phone. Mobile internet is better; it understands users better.

After some reflection, I wondered if I had regressed, but ultimately, I felt it was the TV manufacturers who should reflect!

A device hanging on the wall with a larger screen and better sound lost out to a small 6-inch screen after twenty minutes of wasted time.

Interestingly, while manufacturers tout smart features and product upgrades, their profits increasingly rely on overseas markets as substitutes.

According to TCL Electronics' 2026 interim report, revenue for the first half of the year was HKD 63.76 billion, up 16.4% year-on-year. International revenue from large-size display business reached HKD 25.442 billion, a 29.6% increase. Specifically, revenue in Europe, North America, and emerging markets (Asia-Pacific, Latin America, and the Middle East & Africa) grew by 17.2%, 26.3%, and 37.3%, respectively, year-on-year.

Skyworth Group achieved operating revenue of RMB 38.035 billion, up 4.9% year-on-year. Smart home appliances remain the largest revenue source, with domestic market revenue and overseas market revenue increasing by 2.80% and 21.40%, respectively, year-on-year.

While manufacturers are busy expanding overseas, domestic users are turning on their TVs less and less.

According to AVC Revo data, in the first half of 2026, domestic color TV retail sales reached RMB 51.5 billion, down 8.2% year-on-year, with retail volume of approximately 12.49 million units, a 9.2% decline, marking the lowest sales volume for the same period in four years. According to Gozen Data's white paper, the average daily turn-on rate for smart TVs is only 52.5%.

However, the domestic market has not abandoned TV as an important content consumption scenario. According to the National Radio and Television Administration (NRTA), after three years of "dual governance," boot-up ads have been fully eliminated, boot-up time reduced from a maximum of 118 seconds to under 35 seconds, and 200,000 hotels and 11 million guest rooms nationwide have completed "convenient TV viewing" upgrades.

These solutions address the issue of "being able to watch" but fail to solve the issue of "wanting to watch."

After more than a decade of smart TV development, why are they getting smarter yet harder to use? In the AI era, who can make people turn on their TVs again?

TVs Lose to Smartphones Because They Don't Understand Users

The U.S., with 330 million people, sells nearly 50 million TVs annually; China, with 1.4 billion people, sells only 38 million.

Some blame the content. However, live broadcasts, dramas, and sports are all available on TV, and copyright is no longer an issue.

Others blame policies, but policy changes over the past three years have focused on simplification. According to the NRTA, boot-up ads have been fully eliminated, subscription packages reduced by over 50%, and over 200 million users can now watch TV with a single remote control. Most obstacles have been cleared.

Yet, people still aren't coming back. It's not that Chinese people don't watch big screens; smartphones beat TVs due to operational costs, which stem from account systems.

Your preferences, watch progress, favorites, and payments are all tied to your account. Open any video app, and it recognizes you, continuing from where you left off. TVs, however, are "unowned devices" in the home: parents watch operas, children watch cartoons, and when it's your turn to watch a drama, every use starts from scratch, with personalized settings nowhere near as intelligent as video apps.

Manufacturers have realized this and are pinning their hopes on AI.

At TCL's 2025 Global Technology Innovation Conference, Li Dongsheng revealed that TCL's AI implementation had generated over RMB 1 billion in comprehensive benefits by 2025. TCL Technology COO Wang Cheng themed the conference "AI for Real": The ultimate significance of AI lies in scenario implementation and value creation.

Reading between the lines, this admits a truth: Over the past decade of smart TV development, many features have neither landed in real scenarios nor created value.

Voice control, content recommendations, and screen mirroring protocols are all in place, but no one has solved the fundamental issue: making the screen recognize the person sitting in front of it.

Hotel scenarios amplify this flaw.

Home TVs at least recognize familiar faces, but hotel TVs always face strangers. Accounts can't persist, preferences aren't accumulated, and users have zero patience for interaction. If even the most "plug-and-play" scenario requires half an hour of fiddling, the product logic is flawed from the start.

In the content era, TVs provided channels; in the smart era, their value should lie in interaction. Yet, to this day, the screen remains public, anonymous, and impersonal.

The market expects AI to drive more change, and Google and several Chinese manufacturers have launched explorations, but which solutions truly address pain points remains to be seen.

What Value Can AI Add to Big Screens?

Google and TCL are pioneering new AI-driven interaction experiences.

In September last year, TCL launched its flagship QM9K model in North America, becoming the world's first Google TV powered by Google Gemini.

According to Google's CES 2026 disclosures, Gemini TV features include natural language search (e.g., asking, "What's the medical drama everyone's talking about?" for instant results) and dual-user preference fusion recommendations (combining two people's tastes into a single playlist). Additionally, TCL equipped the TV with a millimeter-wave presence sensor to detect approaching users and automatically turn on the screen.

For the first time, TVs are beginning to "perceive" the people sitting in front of them rather than just waiting for remote commands.

The sole criterion for judging these new features is usage frequency.

Unlike past AI features that impressed at launches but saw limited annual use per household, voice-adjusted settings and dual-user playlists target high-frequency pain points.

Digging deeper, Google's approach fundamentally points to accounts.

Dual-user recommendations require recognizing two individuals separately. The first-order value of AI for TVs isn't more features but transforming a public screen into a personalized one. This is a transformation smartphones completed a decade ago but TVs have yet to achieve.

Smartphones have proven that the real profits of smart hardware lie not in product sales but in post-activation engagement.

However, a looming question remains about AI TVs' value prospects: To date, AI has only contributed to product premiums, not industry growth.

According to market data, even with rapid AI TV penetration, global TV shipments in 2025 still declined slightly by 0.1% year-on-year.

Can these AI interaction capabilities become a profit driver for manufacturers, fostering more sustainable innovation?

As the first in the industry to integrate Gemini, TCL's overseas content operations are handled by TCL Channel. According to TCL Electronics' annual report, this business has surpassed 45.7 million cumulative users. In 2025, TCL Electronics' internet business revenue reached HKD 3.109 billion with a 56.4% gross margin, while display business margins hovered around 16%.

Beyond TCL, similar explorations by other manufacturers also validate the profit potential of AI-powered software.

According to Skyworth, its Coolopen system has connected over 200 million devices, with daily interactions exceeding 7 billion. Southern Media Corporation's IPTV and Oriental Cable's AI set-top boxes both use Coolopen's solutions. A Coolopen executive stated publicly: "This year's focus is daily active users; next year's is payment conversion."

Clearly, AI subscriptions could become a third revenue stream alongside hardware and advertising revenue. Moreover, the significance of AI TVs in future ecosystems and their potential for retrofitting existing devices are underestimated.

On one hand, ecosystem players like Xiaomi emphasize the "human-vehicle-home" ecosystem's family gateway.

Xiao Ai on TVs has integrated large models, and Xiaomi plans to reconstruct its HyperOS into an AI OS within two to three years, enabling phones, cars, and appliances to share a unified account and intelligent agent. Connecting hundreds of millions of MIUI accounts across all scenarios could create more reasons for users to turn on their TVs.

On the other hand, some players target retrofitting opportunities in the existing market.

For example, SDMC bypasses integrated TVs, using AI set-top boxes to enter operator markets and upgrade old TVs into intelligent gateways.

Software-based, lightweight, and user-centric solutions could unlock more scenario possibilities.

The Next Company to Disrupt TVs May Emerge from Hotel Rooms

TVs' next interaction revolution will likely not start in living rooms.

The living room market is locked down by competing interests: broadcasters control content, operators manage channels, platforms handle distribution, and manufacturers build hardware. Any interaction change requires approval from all three other parties.

As Clayton Christensen explained in The Innovator's Dilemma, disruptive innovation rarely emerges from mainstream market centers; it often starts in edge scenarios with authentic needs that incumbents ignore.

Hotel rooms are the edge scenario for the TV industry.

Guests' patience is measured in minutes; learning costs must be zero. Accounts are temporary—functional upon arrival, wiped clean upon departure. Needs are highly concentrated: watching live broadcasts, screen mirroring, and continuing unfinished dramas. Any interaction solution that thrives here must be deeply user-centric.

Overseas, companies are already exploring this scenario.

According to Roku's official product descriptions, guest mode allows hotel guests to log in to their streaming accounts, set a departure date, and automatically log out all accounts upon checkout. What took me half an hour to figure out in a hotel is a clear product flow in Roku's logic.

Currently, domestic companies treat hotel scenarios as custom projects rather than critical innovation testbeds.

For example, TCL's customized TVs for Hampton by Hilton offer one-to-one precise screen mirroring, with independent channels per room to prevent "mirroring to the wrong room," and voice control integrated with room lighting and curtains. However, such solutions remain incremental add-ons with low AI integration due to various constraints.

This may be the opportunity.

Imagine a company that requires no hardware changes from hotels, only solving four specific tasks: instant live TV access upon boot-up, screen mirroring within ten seconds, QR code-based continuation of unfinished content from guest accounts, and automatic logout of all accounts upon checkout.

None of these tasks touch content copyright or infringe on broadcasters' or platforms' interests; they purely enhance interaction.

Once perfected, room service, breakfast, laundry, and late check-out—hotel services naturally suited for voice and big-screen delivery—could create new service entry points for whoever smooths out interactions.

SDMC's AI set-top boxes for operators and Coolopen's AI set-top boxes for Oriental Cable essentially follow the "no hardware change, just interaction upgrade" approach. However, these solutions target operators, not hotels. With cultural and tourism consumption upgrading, B2B client upgrade spending represents a viable business.

From homes to hotels and beyond to various "screen-based scenarios," the AI era will redefine TVs' value, triggering the industry's next wave of differentiation.

In the past, TV companies were evaluated by shipment volumes, panel costs, and market share—a manufacturing framework. But if the logic of AI-driven software revenue holds, valuation benchmarks will shift to active screen counts, daily usage duration, account penetration rates, and ARPU per screen.

From AI hardware to AI ecosystems, TV manufacturers' valuation potential could expand significantly.

Capital markets have already signaled this. Since 2026, the Hang Seng Tech Index has corrected by over 20%, yet TCL Electronics has surged over 70% against the trend. Brokerage reports attribute the stock rally to sustained EPS growth, with high-margin internet business contributing increasingly heavily.

Perhaps one day, travelers lying in hotel beds will simply speak, and the TV will automatically turn on, resuming their unfinished drama—marking the smart TV industry's entry into a new era of value expansion.

Source: HK Stock Research Society