Home BlogDatasetTeaching AI to Talk

Teaching AI to Talk

by Michael McLaughlin
written by Michael McLaughlin
Dataset

Researchers from the Allen Institute for Artificial Intelligence—a research organization started by Microsoft co-founder Paul Allen—as well as Stanford University, the University of Washington, and the University of Massachusetts Amherst have released a dataset to helps machines engage in conversations. The dataset, called Question Answering in Context (QuAC), includes 14,000 dialogs and 100,000 questions about content in Wikipedia articles. The dataset tests how well AI models can use context, answer open-ended questions, and understand when a question is unanswerable. The researchers found that humans still significantly outperform AI models.

Get the data.

Avatar photo

Michael McLaughlin is a research analyst at the Center for Data Innovation. He researches and writes about a variety of issues related to information technology and Internet policy, including digital platforms, e-government, and artificial intelligence. Michael graduated from Wake Forest University, where he majored in Communication with Minors in Politics and International Affairs and Journalism. He received his Master’s in Communication at Stanford University, specializing in Data Journalism.

You may also like

Detecting Deepfakes

Recording Microplastics Pollution

Expanding Drug Discovery

Simulating Solar Systems

Mapping Mineral Exports

Examining Great Ape Cognition

Collecting Insect Sounds

Monitoring Mental Health

Cataloging Cactus Ecology

Analyzing Bus Route Speeds

Show Buttons
Hide Buttons