Study: AI models fail to reproduce human judgements about rule violations
Models trained using common data-collection techniques judge rule violations more harshly than humans would, researchers report.
MIT CSAIL researchers discuss frontiers of generative AI
Experts convene to peek under the hood of AI-generated code, language, and images as well as its capabilities, limitations, and future impact.
Efficient technique improves machine-learning models’ reliability
The method enables a model to determine its confidence in a prediction, while using no additional data and far fewer computing resources than other methods.
2022-23 Takeda Fellows: Leveraging AI to positively impact human health
New fellows are working on health records, robot control, pandemic preparedness, brain injuries, and more.
Meet the 2022-23 Accenture Fellows
This year's fellows will work across research areas including telemonitoring, human-computer interactions, operations research, AI-mediated socialization, and chemical transformations.
Study: Automation drives income inequality
New data suggest most of the growth in the wage gap since 1980 comes from automation displacing less-educated workers.
Ensuring AI works with the right dose of curiosity
Researchers make headway in solving a longstanding problem of balancing curious “exploration” versus “exploitation” of known pathways in reinforcement learning.
Bringing lessons from cybersecurity to the fight against disinformation
Mary Ellen Zurko pioneered user-centered security in the 1990s. Now she’s using those insights to help the nation thwart influence operations.
Caspar Hare, Georgia Perakis named associate deans of Social and Ethical Responsibilities of Computing
The faculty members will work together to advance the cross-cutting initiative of the MIT Schwarzman College of Computing.
Explained: How to tell if artificial intelligence is working the way we want it to
“Interpretability methods” seek to shed light on how machine-learning models make predictions, but researchers say to proceed with caution.
A technique to improve both fairness and accuracy in artificial intelligence
Methods that make a machine-learning model’s predictions more accurate overall can reduce accuracy for underrepresented subgroups. A new approach can help.
3 Questions: Teaching computational maker skills through gaming
With FabO, PhD student Dishita Turakhia wants to empower students to learn digital fabrication by making video game objects and characters come alive.
Building explainability into the components of machine-learning models
Researchers develop tools to help data scientists make the features used in machine-learning models more understandable for end users.
Making data visualization more accessible for blind and low-vision individuals
Researchers have created prototypes that enable screen-reader users to quickly and easily navigate through multiple levels of information in an online chart.
Inaugural Day of AI brings new digital literacy to classrooms worldwide
Thousands of children participate in MIT-developed artificial intelligence curriculum.