The MIT Quest for Intelligence Report to the President, 2023–2024
The endeavor to understand intelligence in engineering terms drives The MIT Quest for Intelligence (The Quest). Our faculty, staff, and students are focused on research and applications at the interface of Natural Intelligence (NI) and Artificial Intelligence (AI). Over the past year, we have seen significant progress in the work; this progress is due, in part, to a stable team, advances in the research tools built by the Engineering Team, and increased support from inside and outside the Institute. We have issued another round of funding to our Missions — interdisciplinary teams of researchers, each spanning science and engineering, and each focused on a specific domain of intelligence.
Recent significant changes and milestones include launching the Perceptual Intelligence Mission, taking steps towards establishing the Intelligence Observatory, and seeing community adoption of the Brain-Score Platform as a research tool. Several ongoing hiring searches have been completed and workloads are better balanced. With the opening of the Schwarzman College of Computing (SCC), Building 45, our offices have reached their planned destination, allowing us easy and frequent access to colleagues and labs in Building 46 and the Stata Center.
The full report can be read in pdf.
The MIT Quest for Intelligence Report to the President, 2022–2023
Understanding human intelligence is one of the greatest human endeavors of all time – right alongside existential questions such as our world’s position in the universe and the origin of life. The MIT Quest for Intelligence (The Quest) is the only organizational unit at MIT aimed directly at this question. The past year has been a good one for the Quest as it continues to have significant influence on research and applications at the interface of Natural Intelligence (NI) and Artificial Intelligence (AI). Adjustments and improvements to the organization’s focus, administration, and staffing have resulted in positive outcomes. Our efforts to fund Missions — interdisciplinary teams of researchers, each spanning science and engineering, and each focused on a specific domain of intelligence — have been successful, and we have launched a Language Mission and an effort in Scaling Inference. After several successful hiring searches, our expanded staff are working to support these Missions, our interface with the MIT community, and ongoing research efforts.
The full report can be read in pdf.
MIT Quest for Intelligence Report to the President, 2021–22
Over the past year, the MIT Quest for Intelligence (the Quest) has executed a significant metamorphosis of the entire organization, in its focus, its administration, and its staffing. At this point, the Quest community is stronger than it has been since its launch in 2018 and is well poised to make an impact on research breakthroughs and applications at the interface of Natural Intelligence and Artificial Intelligence (AI) for years to come. In summer 2021, we began funding Quest Missions that bring together interdisciplinary teams of researchers focused on specific areas of inquiry; in fall 2021, the Quest leadership team was reorganized; and in winter 2021, the Quest incorporated MIT’s Center for Brains, Minds, and Machines (CBMM), a closely aligned NSF-funded science and technology center focused on the interdisciplinary study of intelligence and how it can be replicated in machines.
The full report can be read in pdf.
MIT Quest for Intelligence Report to the President, 2020–21
The MIT Quest for Intelligence, a research unit of the MIT Stephen A. Schwarzman College of Computing, is a campus-wide initiative aimed at one of the most central challenges in human history: understanding intelligence.
MIT Quest is a search to understand how brains produce intelligence and how intelligence can be replicated in artificial systems. The Quest approaches this as a single grand challenge requiring the organized, collaborative efforts of science, engineering, the humanities, and beyond.
To achieve this vision, new scientific theories of natural intelligence must be developed, computational models must be created, and implementations must be compared against the capabilities of natural intelligence and tested on real-world problems and applications. MIT Quest sees the engineering of artificial intelligence (AI) and the scientific understanding of natural intelligence as interlocking aspects of this challenge.
Working across the whole of MIT and with the world at large, the Quest capitalizes on MIT’s long history of leadership at the boundary of natural and machine intelligence. It is uniquely positioned to advance the development of novel intelligent systems and associated tools and technologies that will have a wide spectrum of societal benefits and applications in numerous fields. MIT Quest’s approach does not require merely a crossing of boundaries but a true integration across disciplines. While recent progress illustrates the mutual benefit of studying natural intelligence and intelligent engineered systems, we believe transformative advances in intelligence will require a new approach that more closely integrates the science of natural intelligence with the engineering and applications of AI.
The full report can be read in pdf.
MIT Quest for Intelligence Report to the President, 2019–20
The MIT Quest for Intelligence is a campus-wide initiative launched in 2018 to advance our understanding of natural intelligence by integrating scientific inquiry with engineering to address real-world applications that are beyond current artificial machine capabilities.
In January 2020, MIT Quest joined the MIT Stephen A. Schwarzman College of Computing as one of its research arms. This organizational shift coincided with a restructuring of leadership with Jim DiCarlo, Aude Oliva, and Nick Roy assuming co- director leadership roles, and Antonio Torralba stepping down as director of the Quest to assume a new role as head of faculty of the artificial intelligence (AI) and decision making subunit within the Department of Electrical Engineering and Computer Science (EECS). The MIT Quest vision, goals, and strategic planning around supporting research missions were subsequently refocused during this transition.
COVID-19 has had a significant impact on MIT Quest’s research and activities. Several events were canceled, and an MIT Quest research mission request for proposals was delayed when campus shut down in March 2020. However, MIT Quest leveraged the expertise of its AI engineers and its unique ability to bring together researchers from across campus to develop a predictive modeling system: the MIT COVID-19 Response System (MCRS). MCRS uses anonymized COVID-19 test results and campus mobility data, in part, to help MIT predict and minimize campus risk. This spring, our affiliated MIT-IBM Watson AI Lab also issued a call for AI-related research proposals to address COVID-19 and its social and economic consequences. The lab has funded 10 MIT-led projects to date.
By bringing together researchers from across the Institute, MIT Quest aims to go beyond business as usual. Our research portfolio has grown to more than 100 projects, and we have sponsored 143 student appointments through MIT’s Undergraduate Research Opportunities Program (UROP). Research topics range from fundamental problems in the science and engineering of intelligence to applications with shorter term, more immediate benefits. MIT Quest remains committed to its holistic vision for shaping the future of the science and engineering of intelligence and develops tools for real-world impact, and we look forward to expanding on this work with the MIT Schwarzman College of Computing (SCC).
The full report can be read in pdf.
MIT Quest for Intelligence Report to the President, 2018–19
The nature of intelligence—how the brain produces intelligent behavior and how it can be replicated in machines—is one of the fundamental problems in science and technology. Progress in this area could have enormous societal impact, making it easier to solve challenges in many other disciplines. We believe the key to progress is to integrate the science and engineering of intelligence. Understanding how the human mind works in engineering terms will lead to transformative advances in artificial intelligence (AI), allowing us to create machines that exhibit human-level intelligence in how they reason, see, communicate, and learn. An understanding of intelligence will also advance the fields of cognitive science and neuroscience and help us explain how the brain functions in both health and disease. We hope to translate this knowledge into a broad spectrum of societal benefits tied to health care, information security, transportation, global communication, education, and more.
The MIT Quest for Intelligence (MIT Quest) seeks to drive the development of transformative tools and technologies to benefit society. MIT Quest brings together researchers from across disciplines to discover the foundations of intelligence and to address the ethical and societal impacts of automated decision making.
The full report can be read in pdf.