AdvertiseDonate
PortlandSeattleDallasJacksonvilleColumbusWashingtonBostonDetroit
Sunday, November 22, 2026Fresh headlines, all day.
Digital News Live · 70 SW Century Drive, Bend, OR 97702
Tech · Tech

1.5 Hours: The time it took one developer to strip ads from his TV using AI

Large language models are enabling non-developers to rewrite the firmware of consumer electronics, from televisions to synthesizers, with growing implications for manufacturer control.

By Alessia Desai ·

Illustration for: 1.5 Hours: The time it took one developer to strip ads from his TV using AI
Illustration generated for this article, not a news photograph.

One developer in India spent 90 minutes prompting an AI model to remove advertisements and simplify the interface on his Xiaomi television. Gathin T, who uses that name online, ran the large language model Claude Code against his Android TV set after becoming frustrated with the cluttered design. The AI adjusted permissions and toggled USB debugging mode to take control of the device while connected to the same Wi-Fi network. T posted the result on X, and the project quickly gained traction.

The success of that single experiment led T to launch CleanYourTV, a tool that allows other users to automate similar changes on their own screens. The project became one of the most popular repositories on GitHub, attracting thousands of pull requests and feature suggestions. T notes that the process works most smoothly on Android-based devices, as other operating systems require more complex jailbreaking and rooting procedures that are harder to share widely.

The trend extends beyond televisions. Camera enthusiasts are using AI to repackage software for their devices, adding recording capabilities and color profiles that manufacturers did not include. In the music industry, the M-Vave FM-1 synthesizer has become a target for custom firmware. The company recently released an open-source version of its software, a move that contrasts with other manufacturers who have tried to restrict third-party modifications.

Gamers are applying the same techniques to consoles. One user found a method to play PlayStation 2 discs on a PlayStation 5, while others have written scripts to use specific controllers on different hardware. In response, the gaming industry is moving to protect its intellectual property. Luke Dicken, formerly the head of artificial intelligence at Take-Two, told Kotaku that the industry will likely adopt more invasive digital rights management and push harder toward streaming solutions to prevent users from decompiling code.

As more users attempt these modifications, the risk of device failure increases. Installing custom firmware often voids warranties, leaving users without official support if the device stops working. Ritchie Roesch of Fuji X Weekly wrote that the question is no longer if a camera will become a permanent paperweight, but when and how many devices will be affected.

A safer alternative for many hobbyists is building new hardware from scratch using ESP32 modules. These chips package networking and audio components into a compact, hacker-friendly format similar to Raspberry Pi devices but often cheaper. Meta is currently promoting this approach through its Muse Gadgets program, which allows users to buy pre-packaged ESP32 modules or build their own dedicated AI devices.

T has updated his CleanYourTV prompts to create software backups and log changes to help users restore their devices if an error occurs. He advises users to test changes on a secondary machine, such as a Linux-based computer, before applying them to primary devices. He emphasizes that users should never give AI tools access to hardware they cannot afford to lose.

More Tech · Report a business · Got a Story?

Get Digital News Live in your inboxFresh headlines, all day. Free, a few times a week, unsubscribe any time.

Related stories

Get Digital News Live in your inboxFresh headlines, all day. Free, a few times a week, unsubscribe any time.