Institute launches MIT Intelligence Quest
New initiative will advance human and machine intelligence research
Emma Bingham–The Tech
President L. Rafael Reif and School of Engineering Dean Anantha Chandrakasan announced today the launch of MIT Intelligence Quest, an Institute-wide initiative to advance human and machine intelligence research.
The initiative, abbreviated as MIT IQ, consists of two interconnected components: the Core and the Bridge.
The Core will focus on the “science and engineering of intelligence,” with a specific emphasis on using a reverse engineering approach to develop human-inspired machine-learning algorithms, Chandrakasan said in a press call Wednesday. The Bridge will emphasize the applications of these findings, by building a “wide swath” of new technologies and platforms that can be used across various disciplines.
In parallel, MIT IQ hopes to use insights from its research to improve understanding of human intelligence.
Within the Bridge component, one specific goal is bolstering student education. Machine learning classes at MIT are heavily oversubscribed, but many EECS faculty simply do not have enough time to teach more, Professor Josh Tenenbaum PhD ’99, a member of the MIT Computational Cognitive Science Group, said in a follow-up call with The Tech.
Possible areas for educational improvement include devoting more teaching assistant and lab support to these classes, as well as adding supplementary classes to the curriculum, Tenenbaum said. Areas outside of the classroom may also be strengthened, such as by increasing funding for UROPs that help students learn how to use state-of-the-art artificial intelligence tools, Chandrakasan added.
MIT IQ will likely cost “hundreds of millions of dollars,” Chandrakasan said in the press call, and funding is expected come from both philanthropy and industrial support.
Correspondents from The Boston Globe and Inside Higher Ed who were on Wednesday’s call raised multiple questions on the purpose of MIT IQ — specifically, what this new enterprise would help MIT accomplish that it cannot already do now with existing labs and resources.
Chandraksan explained in twofold: to amplify existing initiatives and to provide resources for the creation of new, interdisciplinary initiatives. However, his answers and the answers of other professors on the call were lengthy and at times seemed unconvincing, as reporters repeated and reiterated similar questions later in the call.
MIT currently has more than 200 principal investigators working on intelligence-related research.
MIT IQ’s launch comes a few months after the September 2017 announcement of the MIT–IBM Watson AI Lab, which is funded by a ten-year, $240-million investment by IBM. This Lab will be incorporated under MIT IQ, Chandrakasan confirmed with The Tech, and it will be a “cornerstone” in this effort.
In his opening remarks during the press call, Reif also outlined what he believes to be key opportunities in AI. AI may eventually permeate “almost every field,” Reif said, but currently, its scientific foundations are relatively dated, and there is a demand for new breakthroughs.
Reif also emphasized the societal and ethical implications of this research. For instance, as AI becomes a new source of wealth, it also risks becoming a new source of inequality, Reif said.
“We’re very cognizant of that question, and we strongly believe that if you are creating new technologies, … it’s good to figure out how to use it in a way that … benefits us all, ” Reif said.