It has been called crazy and labeled “Toontown.” Passersby refer to it as the building that looks as though it’s going to collapse. And Boston Globe architecture critic Robert Campbell has described it as “a drunken barn dance as it might be represented in a Disney cartoon.” But no matter how it is characterized, MIT’s new Ray and Maria Stata Center for Computer, Information, and Intelligence Sciences will not be overlooked.
“We understood that this would be controversial,” says John Guttag, head of the Department of Electrical Engineering and Computer Science (EECS). Staff, faculty, and students from one of the department’s labs-the Computer Science and Artificial Intelligence Laboratory-will occupy about 80 percent of the building. “It’s not because we wanted it to be unusual,” Guttag says. “It’s because unusual’ is required to support the desired functions.”
The 66,000-square-meter structure, which will be dedicated May 7, was originally intended only to house about 1,000 people from the Computer Science and Artificial Intelligence Lab, the Laboratory for Information and Decision Systems, and the Department of Linguistics and Philosophy. The computer scientists and robotics researchers, in particular, had been occupying leased space off campus for some 40 years. In 1997, when quality-of-life issues arose on campus, such as the need for more student space, the building’s scope and budget were expanded, and public areas were added. The resulting $283.5 million structure is the Institute’s largest building project since the neoclassical edifices surrounding Killian Court were completed in 1916.
Designed by world-renowned architect Frank O. Gehry, whose rsum includes structures such as the Guggenheim Museum in Bilbao, Spain, and the Walt Disney Concert Hall in Los Angeles, the building meets a complicated set of goals. Foremost of these, it brings the computer scientists in the EECS department back to campus and puts them in close proximity to the electrical engineers in Building 36, which is adjacent to the Stata Center on Vassar Street. It also pays homage to Building 20 (the site’s former structure) and creates new public spaces, including a northeast gateway into campus where none existed before. This gateway will help open the Institute to the surrounding community-particularly to the technology companies that have sprung up on that side of campus in recent years. And finally, the Stata Center encourages interactions between researchers, provides flexibility for multiple uses, and stands as a symbol of MIT’s ideals.
Building 20’s Legacy
The Stata Center’s story begins with Building 20, a ramshackle, timber-framed, World War IIera complex set up as a temporary facility for military research. Building 20 first housed the Radiation Laboratory-the group that developed radar. Over time, it held MIT’s computer scientists, electrical engineers, linguists, philosophers, and other groups. Its shared spaces encouraged an unprecedented amount of collaboration, which led to the development of atomic clocks, underwater cameras, and solar vehicles.
Building 20 earned the affectionate moniker “the magical incubator,” and it lived up to its reputation even when it became so dilapidated it had to be demolished in 1999. In fact, the building’s poor condition actually encouraged innovation and free thinking. Guttag explains: “Everyone knew it was going to come down, so you could do anything you wanted to it. It was a space where, if you wanted to find space for a project, you could find it.” Researchers were known to punch holes in the walls as needed, and this freedom meant there were very few limits to their research.
In the 1990s, it had become clear that the “temporary” facility no longer met MIT’s needs. It was too dangerous and cramped for people to work effectively within it. Administrators decided to replace it with a structure big enough to house the computer science half of the EECS department, which had grown over the decades to become MIT’s largest department. The linguists and philosophers, who had remained in Building 20 to the bitter end, would also need new accommodations.
In 1997, to help achieve these goals, Analog Devices cofounder Ray Stata ‘57 and his wife Maria gave MIT $25 million-then the largest building project donation in Institute history-to build a new center for these intelligence sciences. With that, the new building’s life began. Later, Alexander Dreyfoos ‘54, a member of the MIT Corporation, gave $15 million, and Microsoft CEO Bill Gates and his wife Melinda gave $20 million in exchange for naming rights to the buildings’ two towers.
From the early stages of the project, the idea was to build something with iconic value. “We wanted the architecture of the Stata Center to be as bold and imaginative as the intellectual daring, creativity, and excellence it was designed to support,” says Institute president Charles M. Vest HM. “And we wanted it to foster the kind of free-thinking collaboration among its varied occupants that was a hallmark of its predecessor, Building 20-but with the added stimulus of a visionary design.”
Planning and Design
The Institute began planning the new center in 1997. That year, a committee of faculty and administrators outlined the specific features that would characterize the structure, from the number of offices and seminar rooms to the public and private spaces. A year later, consultations with the City of Cambridge over the proposed traffic flow in and out of the building’s parking garage led administrators to move the garage underground. That freed up some surface space on the planned site. So when MIT’s Task Force on Student Life and Learning recommended that the Institute provide more student-oriented areas, administrators determined to dedicate part of the Stata Center to student activity, and to provide fitness and child-care facilities for the entire campus community. They expanded the building’s plan and budget to accommodate these public spaces.
With a rough plan for the building in place, MIT embarked on a global search for an architect. From the beginning, administrators knew they wanted someone first-class. “Doing big, complex buildings really well is a very difficult thing, and if your ambition’s high, you need a first-rate person to do it,” says William Mitchell, former dean of the School of Architecture and Planning and MIT’s architectural advisor to the president. Sixteen top firms submitted proposals, but in the end, MIT chose Frank O. Gehry and Associates, a firm and a designer known for building sculptural and curvaceous structures.
People incorrectly assume that MIT chose Gehry because of his distinctive exteriors, says Chris Terman ‘78, SM ‘78, PhD ‘83, a member of Stata’s project management team. The reality, he says, is simply that Gehry listened well, truly sought to understand MIT, and was committed to involving the building’s future occupants in creating his design proposal.
Also contrary to popular belief, Gehry didn’t show up at MIT with a finished design. “People who don’t really understand Frank think that he comes up with these wild shapes, and then you’re stuck with the wild shapes,” says Rodney Brooks, director of the Computer Science and Artificial Intelligence Lab. “The shape came last. The internal stuff came first, and the dynamics of how people were going to work together.”
In fact, Gehry and his team spent all of 1998 talking with MIT researchers about their needs and crafting a design proposal to meet them. Then, over the next three years, they created more than 50 models and myriad computer renderings of the building, modifying the plan again and again based on input from MIT’s administrators and project management committee. The resulting building, whose construction began in 2000, has an intricate interior that suits many different functions. And despite the fact that its exterior was designed to impress on first glance, Brooks and other Stata residents say the interior is its most remarkable aspect.
(Frank Gehry, From Building Stata by Nancy Joyce/MIT Press)
Gehry’s Design Process
When Gehry first started talking with MIT researchers about their ideal office and lab spaces, it was difficult to get them to think outside the box (literally), Terman says. The researchers tended to describe their current offices as the ideal, instead of brainstorming about the best possible way to arrange their spaces according to use. Terman says researchers often responded to Gehry by saying, “I just need a desk about like this one and a bookshelf similar to this one,” forgetting that their current arrangements didn’t give them enough room to do such things as meet with colleagues or build machines that were taller than the average ceiling, let alone test-fly robots.
So Gehry next talked with MIT researchers about how they work together. He discovered that they highly value communal lab and lounge spaces, but at the same time, they like to keep their private offices private. They also value the flexibility to reorganize their surroundings as necessary.
Given these generalizations, Gehry formulated four design metaphors-specific, but not literal, architectural approaches-that might help shape the research spaces. He created and presented small models of these metaphors: a Japanese house, an orangutan village, a colonial mansion, and a prairie dog town. The Japanese and colonial houses were models of flexibility; their screens and perimeter spaces, respectively, could be reconfigured as needed. The orangutan and prairie dog metaphors maximized interaction while allowing for some privacy. In these models, faculty offices, like nests, were secluded either above or below common work areas.
At first, the metaphors really scared people, Brooks says. “People took them too literally,” he says. But the metaphors were simply “designed to whack us out of our complacency,” says Terman, who is also a senior lecturer in the EECS department. In the long run, they helped researchers hone their wants and needs into very specific requests by giving them something to react to.
Based on that input, Gehry and his team bucked MIT’s typical Infinite Corridorstyle arrangement and adopted a hub-and-spoke model for the research spaces. Now the Stata Center consists of clusters of two-level research neighborhoods-spokes-radiating from multilevel open areas-hubs-where elevators, rest rooms, lounges, and conference rooms are located. Each neighborhood spoke is organized with lab space in the center and private, one-size-fits-all offices around the edges. Every office is suitable for one faculty member, two staff members, or four graduate students, and each has a window that opens.
Outside of their neighborhoods, researchers can interact with people from other groups in the hub lounges. Perfect for relaxation, lunch, mingling, or even small-scale, informal seminars, these lounges, Terman says, are like “an intellectual singles bar. You’re out looking for action here, and it’s a real chance to do an intersect with people.” Conference rooms are nearby, and some lounges are adjacent to private reading rooms.
The research neighborhoods were organized to maximize collaboration. For example, the Laboratory for Information and Decision Systems (LIDS) and the Linguistics and Philosophy Department are located above and on some of the same floors as the Computer Science and Artificial Intelligence Lab’s language, learning, vision, and graphics research group. And on levels three through eight, corridors link these groups to the EECS department’s electrical engineers in Building 36. “For those of us working in learning, it will be particularly nice to be near the LIDS people,” says Leslie Kaelbling, who directs the language, learning, vision, and graphics efforts in the Computer Science and Artificial Intelligence Lab.
Connections between linguists and computer scientists also make sense, the researchers say. “For those of our linguistics graduates that do not become academic linguists, the main career path involves computer science,” says Alec Marantz, PhD ‘81, linguistics and philosophy department head. “We’re looking forward to rebuilding some of the ties with computer science that have loosened since the ’70s [when both groups were housed in Building 20], and we hope that cohabitation leads to new joint research projects on language and cognition.”
Questions of Cost
With such an enormous structure-particularly one designed to make a statement about its inhabitants-comes enormous cost, which raises questions about whether the building was a prudent use of MIT’s money at a time when its endowment is shrinking. Originally, administrators estimated they would spend about $95 million on the building, but this figure crept higher as more donors gave to the project. When public spaces were added to the plan in 1998, the budget was significantly expanded.
Terman says construction costs accounted for about $200 million, and the rest went toward architects’ fees, furniture and equipment, specialists’ fees, regulatory expenses, insurance, and other costs. Provost Robert Brown says constructing the public spaces added between 10 and 20 percent of the previous budget to the total, as these spaces greatly increased the Stata Center’s size. The underground parking garage alone added some $60 million to the price tag, as it required MIT to build a 14-meter-deep slurry wall around the site to hold back the Charles River, whose water table lies just two and a half meters below ground.
Obviously, hiring a world-class architect came at great cost. According to project manager Nancy Joyce, MIT paid Gehry’s office $20 million for basic services plus another $8 million for specialty consultants and additional services; in “normal” projects, Joyce says, basic architect fees run between 8 and 12 percent of construction costs. Were Gehry’s ideas for the center worth the money? Mitchell says yes. “As with everything else, you get what you pay for.” He adds that “although people probably think all those fancy, complicated forms are what’s really expensive, the really expensive thing in a building like this is the infrastructure system”-the raised flooring and the various elements designed to make the structure “green,” for example. Terman says the building’s size simply dictated its cost. “Most of the money doesn’t go into exotic finishes.The investment is really in the amount of space.”
Because the Stata Center enables researchers to pursue new projects and work in comfortable surroundings, it is a good use of MIT’s money, says Mitchell, now head of the Program in Media Arts and Sciences. “The most important resource MIT has is people,” he says. “You want to put them in absolutely the best conditions you can to maximize their effectiveness and to maximize your attractiveness.”
In particular, Brown says, adding the public spaces was worth the cost. “[In] too many buildings on academic campuses, there’s a push to cut the public spaces down to keep the building costs down,” he says. “The net effect of that is, you don’t get very usable public space out the other side-not space people want to be in.” By choosing to invest in large and inviting public spaces, administrators hoped to prevent that from happening in the Stata Center. Will that decision help deepen the sense of community at MIT? Brown thinks “it’s a shoo-in that it will do that.”
So will the Stata Center prove to be “a well-intentioned embarrassment,” as one columnist opined? Administrators say no. Instead, they say it will prove that MIT values its people more than its equipment and that it is as committed to innovative architecture as it is to excellence in other fields. Guttag says it will also show that the Institute is made up of bold and confident risk takers and that MIT is willing to work hard to achieve greatness. Of the center’s unusual design, Guttag says, “I would rather have a building that some people love and some people hate than a building that everybody is neutral on. To me, apathy would have been the worst reaction.”
Because of its unique design, the Stata Center is destined to become an icon-at MIT, in Cambridge, in Boston, and in the world. For that reason, Mitchell says, it’s one of the most important buildings ever built in Boston. “It’s a building of truly global importance,” he concludes. “It’s one of the great cultural artifacts of the early 21st century.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.