Exclusive: Honor Debuts a New AI Agent That Can Read and Understand Your Screen

Last updated: March 2, 2025 6:32 pm

By MT HANNACH

3 Min Read

Disclosure: This website may contain affiliate links, which means I may earn a commission if you click on the link and make a purchase. I only recommend products or services that I personally use and believe will add value to my readers. Your support is appreciated!

He chose a restaurant, but could not finish the process because the spot he had chosen required a credit card to confirm a reservation, how the user had to take over. You can be flexible in your request – in another example, ask him to book a “very well rated” restaurant meant that he would examine criticism with high scores, although the agent does not do more research than that. These are not open cross references with data from other parts of the web, especially since all this data is processed on the device and is not sent to the Cloud.

This type of agentic artificial intelligence is the current fashionable word in the technological sphere. My colleague goes knight recently tested an AI assistant This could browse the web and perform online tasks. Google at the end of last year revealed its Gemini 2 AI model trained to take measures on your behalf. It also renews the idea of a generative user interface for smartphones – at MWC 2024, we have seen some companies working on means of Interact with applications without using at all applicationsInstead, rely on AI assistants to generate a user interface when you put an order.

The approach of honor is a bit like what the rabbit – of the infamous Rab– he does with Mode teachWhere you train his assistant manually to accomplish a task. It is not necessary to access the application programming interface of an application (API), which is the traditional way of applications or services communicate with each other. The agent memorizes the process, then allowing you to issue the order and have it executed the task.

But Honor says that his autonomous AI execution model is not formed to follow strict steps – he is able to recognize the multimodal screen context to perform tasks independently. Instead of having to train the assistant to learn each part of the open application, it is able to understand the semantic elements of the user interface and follow a process in several steps to execute your request. Honor pointed out that this process was more profitable: “Unlike competitors such as Apple, Samsung and Google, which are based on external APIs – resulting in higher operational costs – the agent of Honor independently manages a wide range of tasks.”