Recent studies have found significant biases in artificial intelligence algorithms. Researchers are raising their concerns about AI biases associated with medical devices and algorithms used to analyze health risks. If left unchecked, they say, the technologies could continue to perpetuate sex, gender, and race biases that could exacerbate health care disparities.
Allen Institute researchers have launched a new predictive, comprehensive model of a living human cell. The 3D model has been released as a public resource that will allow researchers to study human disease models and better understand cell properties.
··2 min read
In a world of digital assistants, are we forgetting our manners? Not yet, researchers say. Using Siri and Alexas is not making adults rude. However, this may not be the case for children. Fortunately, digital assistant service provides are taking steps to reward children for general politeness.
Researchers report Xenobots, a computer-designed, hand-assembled organism can find and gather single cells, and assemble "baby" Xenobots. After a few days, the immature Xanobots can also find cells and replicate themselves.
Artificial IntelligenceDeep LearningFeaturedMachine LearningNeuroscienceNeurotechVisual Neuroscience··4 min read
Researchers have created a deep learning algorithm that can be tricked by optical illusions, just like humans. The findings shed new light on the human visual system and may help improve artificial vision.
Researchers have developed a new robotic system than can help people to get dressed. The system relies on the force it feels to help guide a hospital gown onto a person's arm efficiently.
··6 min read
MediTrain, a new app-based program that uses a closed-loop algorithm to tailor the length of meditation sessions to a person's ability, can improve memory and attention in younger people.
Artificial IntelligenceDeep LearningFeaturedNeuroscienceNeurotechOpen Neuroscience ArticlesPsychologyRobotics··5 min read
Researchers report science fiction fans are positive about the potential to upload consciousness, neurotech and digitizing the brain.
Researchers report artificial neural networks contain subnetworks that are considerably smaller, yet capable of being trained to make equally accurate predictions. Many subnetworks can learn to make accurate predictions much faster than the original networks.