London - ALsharqiya, June 4: Researchers from the Massachusetts Institute of Technology (MIT) have launched a startup called Capsa, aiming to develop artificial intelligence systems capable of recognizing the limits of their knowledge and reporting uncertainty in their decisions, a move similar to the work of an AI "lie detector."
The company has developed a platform called Capsa, capable of assessing the reliability of the outputs of machine learning models, including large language models like ChatGPT. Capsa analyzes data to detect any signs of doubt or lack of information, distinguishing between supported decisions and unreliable guesses.
The platform is currently used in various fields, such as communications, pharmacology, and seismic image analysis. Its high data processing speed makes it capable of running even on mobile devices.
The technology's developers hope this tool will contribute to improving the security and transparency of AI and enhancing confidence in its use in sensitive fields such as medicine and autonomous driving.
