Past Events

MIT Quest AI Roundtable: AI applications are moving quickly to smartphones and low-power hand-held devices. To make the shift, both hardware and software will need to be redesigned for speed and efficiency. We will discuss leading strategies for achieving efficient AI by co-designing hardware and software for deep learning. Speakers (MIT): Vivienne Sze; Song Han; Moderator: Aude Oliva.

MIT Quest AI Roundtable: Traditional computer chips waste time and energy shuttling data between separate memory and computational units. Neural circuits in the brain, by contrast, achieve enormous efficiencies by storing and processing information at the same place. Inspired by biological models of learning, researchers are designing computing elements that mimic neural circuits and consume massively less energy. Speakers: Bilge Yildiz, Michale Fee (MIT); Panelists: Jesus delAlamo, Ju Li (MIT); Moderator: Aude Oliva (MIT)

MIT Quest AI Roundtable: Deep neural networks could very well memorize their training data, but instead they find generalizable rules. We will discuss various ideas for why this happens, and how we can build deep learning systems that generalize even better to new and unexpected scenarios. Speakers: Pulkit Agrawal, Phillip Isola (MIT); Alyosha Efros (UC Berkeley); Moderator: Aude Oliva (MIT)

Leaders agree that AI offers a competitive advantage, but only a fraction of organizations are using AI to its full potential. In this virtual event, scientists and business experts from the MIT-IBM Watson AI Lab will explain how to overcome three key barriers to implementing AI successfully — trust, scalability, and reasoning.

Almost everything humans have achieved has been done by groups of people working together. Financial markets operate on this principle of collective intelligence to set prices for stocks, as do Internet search engines to answer questions asked by thousands before. Computers can make groups even smarter, but how should humans and machines interact? This workshop will explore the ways that people and machines, working separately and together, can leverage their relative strengths, resolve conflict and create value for society.

This workshop and tutorial will focus on the promise of generative adversarial networks, or GANs: how we can exploit their benefits while minimizing their potential harm. Topics will include the nuts and bolts of generative models, their applications, generative art, and the science and theory of GANs.

Successful hardware innovation in AI will not take place in isolation, but will emerge from a rich, layered, research ecosystem ranging from material science to software engineering. This workshop will explore novel, long-term opportunities for AI hardware, and feature faculty talks, panel discussion, and a poster session.

To advance further, deep learning systems will have to become more transparent. They will need to prove they are reliable, can withstand malicious attacks, and can explain their reasoning, especially in safety-critical applications like self-driving cars. The symposium will feature faculty talks, a poster session and refreshments.