Add to Favourites
To login click here
Daniela Rus, MIT CSAIL Director, has developed a new idea called Liquid Neural Networks which could help to solve some of AI’s complexity problems by using fewer yet more powerful neurons. She discussed the societal challenges of machine learning, such as the need to handle large amounts of data, computational and environmental costs, and data quality. She also highlighted the lack of explainable AI and how changing network builds can help to alleviate this. Liquid neural networks use command and motor neurons to form an understandable decision tree which can help to create smoother and more targeted results.
