Upgrade the smart terminal experience with AI. TCL releases many new achievements in AI technology | Frontline

Text | Wang Fangyu

Editor | Su Jianxun

AI mobile phones, AI home appliances, AI PCs, AI robots... Under the wave of AI large models, various intelligent terminal manufacturers are actively deploying to utilize large model technology to promote the upgrading of product functions and experiences, and explore new business opportunities. TCL is one of them.

On December 11, at the 2024 TCL Global Technology Innovation Conference (TIC2024), TCL released full-scenario AI applications such as the AI intelligent operating system and the Xiaot Central Control Large Model, and also showcased dozens of AI intelligent terminal products such as AI refrigerators, AI air conditioners, AI TVs, and learning machines.

"The next ten years will be a decade of rapid development of artificial intelligence. All products are worth redeveloping with AI, and intelligence will become a very important technical strategy. TCL's products, whether they are white goods or black goods, are actively embracing AI," Sun Li, CTO of TCL Industries, told 36Kr.

36Kr learned that since the fourth quarter of 2022, TCL Industries has been exploring various large models, and in 2023, many products integrated with large models have been launched.

Sun Li said: "After the arrival of the AI large model, many new opportunities have indeed emerged for AI intelligent terminals. For example, for the previous air conditioners, due to insufficient intelligence, if you ask it to set the temperature to 26 degrees, it can understand, but if you change the expression and say to lower the indoor temperature by one degree, it won't understand. With the large model, this problem becomes very easy to solve."

The Xiaot Central Control Large Model released by TCL at this conference is a typical application of AI large model technology. AI large model technology enables users to interact with devices in a natural conversation manner. Through machine learning, the system will automatically learn the user's habits and provide personalized operation suggestions and settings.

At the conference site, Dr. Yan Xiaolin, CTO of TCL Technology, demonstrated the application results of human-computer interaction technology based on large models on smart TVs. When responding to complex questions such as "I want to watch the new movies of Andy Lau in recent years," the Xiaot Assistant can easily interact and answer.

According to Yan Xiaolin, currently, 2 million smart TVs of TCL have used this technology, and it is expected to reach 17 million units by the first quarter of next year. By the end of 2025, all TCL TVs will support large model interaction.

At the same time, due to the use of model compression technology, users do not need to spend a lot of money for frequent interactions with the large model. "Under the condition that the performance is basically unchanged, the interaction of an intelligent TV for an unlimited number of times is basically controlled at a few cents, which can provide consumers with an experience that is both cheap and easy to use."

In the smart home product exhibition area, TCL showcased smart home products with AI capabilities such as TVs, refrigerators, washing machines, and intelligent terminal products such as mobile phones and learning machines.

In air conditioning products, TCL has launched the Xiaolanying AI Sleep Fresh Air Technology, which introduces an AI sleep cruise in the air conditioner to sense the demands of people in a sleep state and dynamically adjust the sleep environment.

At the same time, TCL's air conditioning products also use AI technology to intelligently adjust the operating conditions of the compressor. Combined with the indoor and outdoor temperatures and the room conditions, the sensor can automatically and timely adjust the operation to save energy consumption. Sun Li introduced that this technology can save up to 40% of electricity for air conditioners, and a single air conditioner can save up to 300 degrees of electricity throughout the year.

It is worth mentioning that in addition to the AI functions carried on terminal products, TCL Industries has also made many forward-looking explorations of AI large models. Its 24 R & D centers and 7 ecological cooperation laboratories established globally are the "main force" of exploration, and are actively deploying in technical fields such as AI, IoT, OS, AR/VR, and cloud computing.

At the press conference, TCL showcased the AI film production application launched by the US R & D center. It is reported that in November this year, five short films generated by well-known screenwriters and directors on the TCL AI content creation platform premiered at the TCL Chinese Theatre in Hollywood. In 2025, TCL plans to release the first 90-minute AI film in North American theaters, which will further promote the innovation and application of AI technology in the field of film and television creation.

In the exhibition area of the press conference, TCL also exhibited an AI podcast application called "Convert to Podcast", which was developed by the TCL European R & D center. Its function is similar to Google's recently popular NotebookLM function, which can convert web pages in various languages into AI podcasts and support mutual translation of multiple languages.

In the currently popular embodied intelligence field, TCL also has a layout. In the industrial field, TCL uses embodied robots for gluing process operations, and its first evaporator gluing process package has completed the first trial production. In the home field, TCL is expected to release the first concept product of a home robot at CES 2025, and the product is positioned as a home AI companion and life assistant.

Through a series of AI intelligent terminals, TCL has showcased its achievements in the intelligent strategy and the continuous exploration of forward-looking AI technologies.

TCL founder and chairman Li Dongsheng said at the conference: "TCL will continue to adhere to the forward-looking AI technology research and development for the future, focus on core resources, give full play to the global layout advantages, not scatter resources, build core capabilities in key directions, solve actual business needs, and promote the efficient implementation of technology."