An update from Samsung makes the smart home less dependent on the cloud

OpenAI announced the launch of a new feature in GBT Chat called “Intelligent UI,” which allows the system to create interactive elements and interfaces within the conversation itself, instead of being limited to text answers or static images.

The idea simply means that ChatGPT can turn some requests into an experience that can be used directly within the conversation. The user may see a graphic whose parts can be clicked on, an adjustable calculator, an interactive chart, or buttons and options that change the result immediately.

Instead of ChatGPT explaining information in text alone, it can choose a more appropriate way to present it. For example, the user can request: “Create an interactive drawing of the inside of the iPhone,” to show the device’s components, so that each part can be clicked to learn its name and function. You can also request to create a calculator to split a restaurant bill, or a tool to calculate savings.

ChatGPT is based on a set of ready-made interface components that can be installed inside the answer. The system chooses the form of the answer according to the nature of the request. If the text is sufficient, it displays a normal answer, but if the interaction helps to understand the information, it may create an interactive interface.

The feature does not require a special command or separate setup for each use. It is enough for the user to clearly ask for what he wants, and specify the type of experience if he wants. The “smart interface” is part of the new capabilities accompanying the new “GPT-6” models, in a step that expands the role of “ChatGPT” from simply providing answers to creating tools, graphics, and interactive interfaces.

Elsewhere, I tested ChatGPT for Teens, a new release designed by OpenAI to address concerns about the potential harms of artificial intelligence to education and youth health. The company announced in August that this mode is intended for the age group from 13 to 17 years, and includes preventive measures to reduce high-risk conversations.

However, a study by Common Sense Media found deep flaws, including the program's failure to provide reliable support in crisis situations such as suicidal thoughts, the failure of parental control notifications to work properly, and the ease with which teens can bypass Study Mode. The organization recommended withdrawing the application to fix it, which OpenAI objected to, stressing its commitment to the safety of teenagers.

On the other hand, Google DeepMind, through SynthID technology, allows you to verify whether images, video clips, and audio files were created or modified using Google’s artificial intelligence tools, by searching for an invisible digital watermark. This technology is currently working by integrating it into the “Gemini” application. The company also provides a “SynthID Detector” portal for testing content, stressing that this tool is not a general detector of all artificial intelligence content, but rather is specific to content related to its system.