Meta's newest AI model, Muse Spark 1.1, is capable of operating your computer autonomously.

      **Meta**

      AI assistants have become excellent at answering queries and guiding us through complex tasks. However, the next generation of AI is aiming for a much more ambitious goal: performing those tasks for us.

      This is the concept behind Meta’s latest offering, Muse Spark 1.1. Rather than merely instructing you on which buttons to click, this model is designed to interact with your computer on your behalf. Whether it involves searching various websites, completing forms, or switching between applications, Meta claims Muse Spark 1.1 can navigate software similarly to a person, opting for the quickest route to accomplish tasks. This marks a significant transition from purely conversational AI to systems designed for active engagement.

      **The era of point-and-click AI has arrived**

      The most significant enhancement is that Muse Spark 1.1 can directly control tasks. The company asserts that this model is compatible with desktop applications, web browsers, and even mobile interfaces. Depending on which method is more efficient, it can either engage with apps in the way a user would — by clicking buttons and navigating menus — or it can automate processes by generating scripts in the background.

      For example, if you're coordinating dinner with a group of friends, an AI agent could manage most of the research for you. Instead of hopping between Google Maps, restaurant websites, and your group chat, it would compare locations, verify operating hours, collect essential details, and present the best options to you. You still make the final decision, but the tedious work has been handled.

      This reflects the direction of the AI race. Companies are shifting focus from creating the most conversational chatbots to developing assistants that can genuinely accomplish tasks. Muse Spark 1.1 also includes a substantial memory enhancement, supporting a context window of up to one million tokens. This capability allows it to track lengthy conversations, large documents, and intricate projects without frequently needing a refresher.

      **Can it really handle the complexities?**

      As with any major AI release, Muse Spark 1.1 comes with impressive benchmark results. Meta claims the model performs well against some of the biggest names in AI in coding and agent-focused tests, even excelling in several areas. While these results are certainly impressive, benchmark charts should not be the only measure of success. The real question is how the model performs once it operates outside of a controlled testing environment. Can it manage multiple websites without losing focus? Will it adapt if a webpage changes suddenly? Can developers rely on it to work through large codebases without needing continual oversight? These are the critical questions that will determine the practical utility of Muse Spark 1.1.

      **Meta**

      Meta is providing developers the opportunity to test it themselves. Muse Spark 1.1 is now available in public preview through the new Meta Model API, allowing developers to build applications and utilize the model in real-world scenarios. Whether Muse Spark 1.1 will become the AI assistant that people truly rely on remains to be seen. However, one undeniable trend is that the competition is no longer about developing chatbots that produce better text responses; it’s about creating AI that can take on repetitive tasks, freeing you to concentrate on other important matters.

      **Shimul is a contributor at Digital Trends, with over five years of experience in the tech space.**

      If you're weary of overseeing ChatGPT, the new GPT-5.6 models may be the solution.

      OpenAI appears to introduce a new AI model every few months, and today is no exception. The company has officially launched the GPT-5.6 family, which includes three new models for ChatGPT, Codex, and its API. The highlight of this release is the GPT-5.6 Sol, accompanied by Terra and Luna, both designed to deliver robust performance at a more affordable price point.

      The era of endless follow-up prompts might soon be coming to an end.

      **AI security cameras may soon recognize your walk before identifying your face**

      A new AI gait recognition system tracks body movements through skeletal keypoints, aiming for long-range identity verification where facial scans and fingerprints may fail.

      Typically, security cameras are designed to identify faces; however, recent research indicates they might soon focus on the subtle habits found in a person's walking pattern. A paper published in the International Journal of Reasoning-based Intelligent Systems introduces SKDMap-Net, a gait recognition system capable of identifying individuals from video footage of them walking, even when the camera cannot capture a clear view of their face. Rather than depending on a close-up scan, it analyzes how a person moves from one frame to the next.

      **A 20-second 3D printer breakthrough comes with the kind of catch that science is known for**

      This process can produce intricate microstructures much faster than many laser-based techniques, but achieving full 3D control is still a work in progress.

      A 3D printer that can create a structure in roughly 20 seconds might sound too good to be true. The innovative part is legitimate, but there’s a catch