Microsoft Co-Founder Invests in Pursuit for AI Common Sense

  • By Dian Schaffhauser
  • 03/06/18

Paul Allen has committed $125 million over three years to support the Allen Institute for Artificial Intelligence (AI2) and to launch "Project Alexandria," a new research initiative to explore "common sense artificial intelligence." Allen is, of course, the co-founder of Microsoft and a philanthropist known most recently for investing in efforts to discover the underwater wreckage of American naval ships sunk during WWII.

Allen's latest commitment focuses on "the everyday knowledge that virtually every person has but no machine does." As AI2 CEO Oren Etzioni explained in a prepared statement, "Despite the recent AI successes, common sense — which is trivially easy for people — is remarkably difficult for AI." For example, he noted in a video on the project, no AI system knows that a "hand has five fingers, or the notion that an elephant won't fit through a typical doorway [or] the idea that if you get too close to the sun, you'll vaporize." Figuring out how to teach computers those kinds of things is, he said, "a holy grail for the entire field of artificial intelligence."

The research team will use machine learning from text, machine vision for analyzing images and video, and crowdsourcing to acquire common sense and deposit it into a "repository of knowledge," which is Project Alexandria. The program is named after the universal library begun in the Third Century, BC, in Alexandria, Egypt, which was known for maintaining all the known knowledge of the world.

The knowledge maintained by the library will then be used in a multitude of AI applications, including robotics, medical diagnosis and broader issues around AI safety. The results will be available to the full research community.

"This is an extremely complicated challenge," said Allen in a prepared statement. "If we want AI to approach human abilities and have the broadest possible impact in research, medicine and business, we need to fundamentally advance AI's common-sense abilities."

About the Author

Dian Schaffhauser is a former senior contributing editor for 1105 Media's education publications THE Journal, Campus Technology and Spaces4Learning.

Featured

  • central cloud platform connected to various AI icons—including a brain, robot, and network nodes

    Linux Foundation to Host Protocol for AI Agent Interoperability

    The Linux Foundation has announced it will host the Agent2Agent (A2A) protocol project, an open standard originally developed by Google to support secure communication and interoperability among AI agents.

  • cloud connected to a quantum processor with digital circuit lines and quantum symbols

    Columbia Engineering Researchers Develop Cloud-Style Virtualization for Quantum Computing

    Columbia Engineering's HyperQ system introduces cloud-style virtualization to quantum computing, allowing multiple users to run programs simultaneously on a single machine. Learn how it works, why it matters, and highlights from other recent quantum breakthroughs from leading institutions and vendors.

  • laptop on a clean desk with digital padlock icon on the screen

    Study: Data Privacy a Top Concern as Orgs Scale Up AI Agents

    As organizations race to integrate AI agents into their cloud operations and business workflows, they face a crucial reality: while enthusiasm is high, major adoption barriers remain, according to a new Cloudera report. Chief among them is the challenge of safeguarding sensitive data.

  • stylized illustration of a desktop, laptop, tablet, and smartphone all displaying an orange AI icon

    Report: AI Shifting from Cloud to PCs

    AI is shifting from the cloud to PCs, offering enhanced productivity, security, and ROI. Key players like Intel, Microsoft (Copilot+ PCs), and Google (Gemini Nano) are driving this on-device AI trend, shaping a crucial hybrid future for IT.