10171936Building A Machine Understanding TechnologyClosedFast Start ResponseInnovate UKWe are developing a new type of artificial intelligence called **Machine Understanding Technology (MUT)**. This innovation aims to make AI not only more powerful but also more trustworthy, especially in areas where accuracy and transparency are critical --- such as healthcare, education, law, and public services. Most current AI systems, including large language models (LLMs), work by predicting patterns in text based on massive amounts of data. While impressive, these models can also produce false or misleading information --- known as hallucinations --- and often make decisions that are difficult for humans to understand or explain. These limitations make them risky for use in real-world, high-stakes settings. MUT is designed to overcome these challenges. It enhances existing AI systems by helping them better understand the meaning of information, not just repeat it. It also allows machines to recognise when they don't have enough information --- what we call an "awareness of unknowns." This makes AI outputs more reliable, logical, and easier for humans to interpret. Our technology builds on a mathematical framework that treats knowledge as a network of logical relationships between concepts. With this foundation, machines can reason more like humans --- drawing valid conclusions, spotting inconsistencies, and explaining their thinking step by step. We have already tested this approach in a healthcare setting, where it improved the accuracy and explainability of AI used in cancer diagnosis. Now, we are expanding the technology to work alongside widely used LLMs like ChatGPT or similar systems. By creating AI that can reason clearly and acknowledge its own limitations, we aim to set a new standard for safe and reliable AI. This project will help unlock the full potential of artificial intelligence in areas where human lives, legal outcomes, or public trust are at stake.