China 
Africa 
Explore chapters and articles related to this topic
Environmental Protection
Published in Lawrence S. Chan, William C. Tang, Engineering-Medicine, 2019
Unnecessary Surgery Reduction. Studies have indicated that for certain diseases, such as uncomplicated appendicitis, medical treatment would be as effective as surgical treatment, even though surgery has been the traditional “standard” approach (Salminen et al. 2015). Since medical treatment would likely generate less medical waste and cost less than surgical treatment for the same disease, healthcare professionals would be benefited from receiving such information, if these information is provided as evidence-based solid scientific data (Salminen et al. 2015, Sippola et al. 2017). Furthermore, some studies with the assistance of artificial intelligence, have indicated that some slow growing tumors, like specific types of prostate cancer, do not require surgical treatment (Schroder et al. 2009, Flores-Morales and Iglesias-Gato 2017). Such information, provided as evidence-based scientific data, would be greatly helpful to healthcare providers to avoid unnecessary surgery procedures, thus reducing medicine waste and morbidity to the patients.
World models and predictive coding for cognitive and developmental robotics: frontiers and challenges
Published in Advanced Robotics, 2023
Tadahiro Taniguchi, Shingo Murata, Masahiro Suzuki, Dimitri Ognibene, Pablo Lanillos, Emre Ugur, Lorenzo Jamone, Tomoaki Nakamura, Alejandra Ciria, Bruno Lara, Giovanni Pezzulo
A generalist agent called GATO was developed based on Transformers and shown to be able to solve various tasks with one neural network [289]. Although the approach is superficially different, the approach is really related to the world models and the predictive coding approach. However, the learning system is hugely data-hungry. It is very questionable if the model can be regarded as a model of human intelligence. Moreover, to train the generalist agent, researchers need to prepare a large dataset and simulation environment. Human children can autonomously explore their environment and acquire data through active exploration. Moreover, they use heuristics and biases in their developmental process. Learning and considering the human developmental process will give us the inspiration to build real generalist agents. Developing a data-efficient autonomous learning architecture with world models and predictive coding at its core is the key to a truly cognitive and developmental system.