Researchers from Israel's Weizmann Institute of Science (WIS) and Intel Labs have developed a new AI technology that allows different AI models to communicate and work as a team. This breakthrough introduces special algorithms that let AI systems think together as one unit, improving performance and cutting costs.

The new method has shown an average speed increase of 1.5 times, and in some cases up to 2.8 times, especially in large language models like ChatGPT and Gemini. This will boost AI performance in areas like smartphones, drones, and self-driving cars, where faster responses are critical. For example, in autonomous vehicles, quick decisions are essential for safety, and this technology can help AI make smarter and faster choices.

So far, different AI models developed by separate companies couldn’t work together because they used unique internal languages (called tokens). This was like people from different countries trying to talk without a common language. To solve this, the researchers created two algorithms — one converts AI data into a shared format, and the other ensures that all models use common tokens for smooth communication. These tools are now available as open-source, helping developers around the world build faster, cooperative AI systems.