EA Boston 2018 Year in Reviewpost by Taymon, rebecca_baron · 2019-02-05T20:15:23.216Z · score: 33 (18 votes) · EA · GW · 8 comments
Local Group Structure and Activities Fellowships The Rationalist Community in Boston Institutions and Opportunities The EA Community in Boston Building Unity Plans and Possibilities for 2019 None 8 comments
This is a review of the Boston EA community’s activities in 2018. Rather than having a single overarching group, Boston has many smaller groups, mostly based around universities. 2018 was the first year in which all these groups began to collaborate systematically and form the sort of city-wide EA network Boston had previously been lacking. In the interest of keeping the global EA community informed of our activities, we decided to write the first annual review of all significant known EA activities in Boston. We hope that other local group leaders and EAs involved in community-building efforts will find it useful, particularly if they’re working in cities with large student populations or other relevant similarities to Boston.
- MIT and Harvard have the most prominent student groups in Boston, with some EA activity at Tufts, Wellesley, and Boston University. Julia Wise runs an independent meetup group for non-students.
- The Arete Fellowship, developed largely by Harvard undergraduate Stephen Casper, has been a very successful tool for student group recruitment and is spreading quickly to other schools. Harvard also pioneered a career fellowship for graduate students.
- The local rationalist community has been defunct since the dissolution of their group house.
- Boston’s main comparative advantage is its concentration of prestigious universities. It’s also a center for biology research and home to the Future of Life Institute, the Abdul Latif Jameel Poverty Action Lab, and a massive fundraiser for GiveWell charities conducted at Google’s Cambridge office.
- The Boston community’s problems include distance from the Bay Area, high student turnover rates, and reliance on volunteers with other commitments.
- Communication and collaboration between different student groups increased greatly this year after their leaders came together to work on EAGxBoston 2018.
- In the future, we hope to initiate a career network focused on emerging technology policy, host EAGxBoston 2019, professionalize and expand our fellowships, and seed more university groups in our area.
This document was written primarily by Rebecca Baron and Taymon Beal, with contributions from Chris Bakerlee, Stephen Casper, Joan Gass, Juan Gil, Justin Kwong, Cullyn O’Keefe, Lucas Perry, Valerie Richmond, and Julia Wise.
Local Group Structure and Activities
Boston has a strong comparative advantage in terms of access to intelligent, promising young people due to its status as arguably the world capital of higher education. There are fifty-two institutions in all, including Harvard and MIT, whose prestige and brand recognition is unrivaled almost anywhere in the world—and no other city has two such institutions just an 8-minute subway ride apart. Boston’s compact geography and convenient public transportation make it easy for members of different local groups to meet in person.
Harvard boasts several of the largest and most prominent EA student groups in Boston, with Harvard University Effective Altruism comprising possibly the largest group of EA graduate students in the world. Of Harvard’s twelve graduate schools, four have active EA groups: the Graduate School of Arts and Sciences, which includes most M.A., M.S., and Ph.D. students; Harvard Business School; Harvard Law School; and the Kennedy School of Government. The T.H. Chan School of Public Health’s group went dormant, but is in the process of being revived. In addition to the graduate groups, Harvard boasts a vibrant and active undergraduate group, Harvard College Effective Altruism. All of these groups are coordinated by an umbrella organization, the Harvard Effective Altruism Student Group, which is currently headed by Cullyn O’Keefe in accordance with Harvard EA’s strategic vision [EA · GW].
Component groups at Harvard ran career fellowships, lunch events, fundraisers, speaker events, and recurring social hangouts this year. The umbrella group is running the Philanthropic Advisory Fellowship, led by Eric Gastfriend, which consults for large donors on how to give effectively. The Fellowship trains over a dozen graduate students in EA each semester while consulting for clients such as The Laura & John Arnold Foundation, Schmidt Futures, Draper Richards Kaplan Foundation, The Life You Can Save, and others. They are currently designing a course on Rationality and EA to be taught at Harvard College by Steven Pinker in Spring 2020. Harvard Law School ran an AI safety reading group, and there is an informally organized EA coworking session on Harvard’s campus once a week.
Meanwhile, a twenty-minute walk away, there's MIT—another powerhouse both in academia and in the vitality of its large and active EA group, energetically led by Juan Gil, Valerie Richmond, and Luis Hong Sanchez. There is also a group associated with the Sloan School of Business, but they are not currently very active. The MIT group is younger than Harvard and has gone through organizational ups and downs, but is currently leading many on-campus activities, including social events and a technical AI safety reading group, as well as spearheading some intercollegiate projects.
Tufts currently has a single EA organizer, Rebecca Baron. Though the group was active circa 2016, when it was led by its founders Dillon Bowen and Luke Sabor, the momentum petered out after their graduation. No one was actively pushing Tufts EA forward until Rebecca took over for the Fall 2018 semester, and her first round of attempts to revive the undergraduate group did not succeed. Tufts has an active chapter of One for the World, but its leaders are not EAs.
Wellesley also has a One for the World chapter and has long had a few scattered EAs enrolled, but they have not held explicitly EA events since 2017. However, 13 Wellesley students recently graduated from MIT’s Arete Fellowship, an introductory EA reading and discussion program detailed later in this post. Some of them indicated interest in restarting a formal EA group at Wellesley. We are cautiously hopeful about Wellesley EA’s future.
Boston University has no EA group at the moment, but it has at least one active EA who is planning to create one. No other schools are known to have EA presence at the moment, though efforts to seed more groups are underway.
In addition to the student groups, Julia Wise runs an independent meetup group that dates back to before EA was formally organized. This group’s primary activity is dinner-and-discussion meetups at Julia’s house, which used to be scheduled irregularly but since late 2018 now occur monthly. Julia cooks dinner for the guests, who then informally break out into groups and discuss whatever topics are of interest. Attendees have varying levels of EA knowledge and commitment, and few come every month. The dinners provide social support for the community and help keep established EAs updated on the movement’s latest developments, while Julia’s introductory Q&As help newcomers learn the basics. Other than the dinners, there are occasional instantiations of an EA reading group and sometimes attempts at different activities. Notably, Scott Weathers and Sophie Hermanns led a Congressional letter-writing and lobbying campaign in early 2018 in favor of the Reach Every Mother and Child Act. Having an independent group separate from the university groups lets Boston EA target different demographics in different ways, appealing to both ambitious 18-year-olds and the general population.
Julia works for CEA as community liaison and president of Giving What We Can, but her work with the independent local group is not a central part of her job duties. Lately, Johnson Ramsaur has been increasingly involved in organizing, increasing the group’s leadership capacity and reducing its reliance on a single organizer.
Finally, Taymon Beal has no formal role in any of these groups but acts as an unofficial inter-group liaison; they involve themself in as many of the various groups' activities as they have time for, talk to their leaders, and try to keep tabs on what everyone is doing. This lets them encourage each group to coordinate their activities when doing so makes sense.
In addition to local-group-led activities, Boston has also hosted talks and informal discussions with prominent visiting EAs and EA-adjacent people doing relevant work, including Jacy Reese from Sentience Institute, Jeff Alstott from IARPA, and Che Green from Faunalytics.
The Arete Fellowship (guide/handover document here), an innovative new program run at Harvard and MIT as well as Duke and the University of Houston, is a weekly reading and discussion group meant to introduce undergraduates to EA ideas. Fellows are chosen through a competitive application process, after which they are assigned about an hour’s worth of weekly EA readings and meet up once a week to discuss them. The nine-week Arete curriculum, composed largely by Harvard undergraduate Stephen Casper, is designed to give complete beginners a working knowledge of core EA ideas. Casper also worked hard on building personal relationships with Harvard’s fellows, personalizing each acceptance letter based on the fellow’s application and attempting to meet each fellow for a one-on-one dinner. Of the 68% of MIT Arete Fellows who completed the program, 85% said in a survey that they would be interested in working with the fellowship or MIT Effective Altruism in the future (obvious selection bias warning). Harvard’s survey data was invalidated by low response rates, but many fellows became involved with Harvard EA projects after finishing the program. Both Harvard and MIT have realized that piping fellows from the fellowship to more active EA roles is a significant yet important challenge.
Harvard and MIT plan to run the Arete Fellowship again in the future, and there has been interest from many other schools in adopting it. The program is new and still heavily under revision, and any schools the program expands to are encouraged to experiment with the readings and format. In time, we hope to produce a high-quality, standardized curriculum that can be broadly distributed.
Meanwhile, the Harvard Graduate School of Arts and Sciences Effective Altruism student group launched the Agathon Career Fellowship to help graduate students work towards satisfying and high-impact careers. 43% of applicants were accepted, yielding an inaugural class of 17. Over four two-hour weekly sessions, Agathon Fellows were given a detailed, interactive walkthrough of the 80,000 Hours career guide plus small-group peer career coaching and one-on-ones with the facilitators (Holly Elmore, Elliot Glazer, and Chris Bakerlee). Again, survey data was biased by low response rates, but it yielded constructive criticism that will be taken into account in future iterations of the program.
The Rationalist Community in Boston
Boston had hosted rationalist meetups since 2009, but the community became more firmly established around 2013 with the formation of the Citadel group house. Citadel hosted meetups every other week and sometimes other activities such as Order of the Sphex, a sort of weekly group productivity check-in; AI safety, machine learning, and writing meetups; and community socializing.
Though this community was specifically rationalist, not EA, it overlapped somewhat with Julia’s discussion groups. Many members were interested in EA-relevant topics, and a significant fraction were giving 10%, ethical veg*ns, or both. Many alumni from that community are now working on EA priority projects or in places with EA missions, including the DeepMind safety team, CFAR, and LessWrong 2.0.
Citadel dissolved in 2017, a victim of bureaucratic crossfire between the landlord and the municipal housing authorities. Another smaller group house, Sunshine Regiment Treefort, followed it, but it was less culturally successful and lasted only a year. A third attempt to establish a rationalist group house failed. The community that Citadel was the center of is currently defunct.
Taymon Beal also ran a monthly Slate Star Codex meetup for a while, but it lost traction after the first two meetups. Evan Hefner has taken this over and his first meetup went well.
Institutions and Opportunities
Again, Boston’s comparative advantage is its high concentration of prestigious universities. It’s a city full of young people well-positioned for high-impact future careers, who are unsure what they want to do with their lives but tend to be ambitious about it: people who are likely to find EA’s message and goals appealing, and have the potential to do big things once they’ve joined. However, since it’s difficult for college students to contribute to EA projects beyond campus organizing until they graduate, Boston EA’s heavy student composition may hurt our capacity for producing immediate impact. Academia also creates opportunities to lose sight of impact for prestige: “AI” is a bit of a buzzword in many EA-irrelevant contexts, and it’s easy to get excited about activities, especially pitched from outside the EA community, that sound good but don’t do much. This is especially tempting since Boston has few direct-work jobs, as described below.
MIT has rather fallen behind in the machine learning boom, as the advantages of Stanford and UC Berkeley are so strong that most of the most promising researchers go there. It’s planning to establish a new billion-dollar school of computing in order to catch up with the Bay, but it’s uncertain whether it will succeed in that respect.
In addition to the universities, Boston is ostensibly the world capital of biotech. As per @anthropicprincipal on Discord: "The same way that the Bay Area feels like everyone has a startup and that if you express Wrong opinions about I dunno TypeScript aloud on a street corner a random passerby may argue with you, Cambridge feels like half the people on the subway with you have some kind of almost-self-aware maybe-cancer-eating fungus they're on their way to sing lullabies to."
Boston companies and institutions provide ample opportunity for wide-ranging biomedical research, including synthetic biology and immunology. However, little of this work is focused explicitly on mitigating global catastrophic biological risks, improving animal welfare, or slowing aging. Kevin Esvelt and the Sculpting Evolution group at the MIT Media Lab are doing EA-relevant research—specifically on malaria gene drives. Esvelt has spoken at multiple EA Global conferences and has written about the prevention of global catastrophic risks arising from biotech and the potential use of gene drives to prevent wild animal suffering.
The only explicitly EA nonprofit located in Boston is the Future of Life Institute, which works primarily on nuclear and AI risk. While their in-house research is currently limited, they run projects that enable or support research, or are largely direct-action- or advocacy-focused. They previously functioned as a grantmaker for AI safety research, with $10 million in funding from Elon Musk, but do not have immediate plans to give out more money. FLI is looking for volunteers, especially for the Future-of-Life Awards program, which will recognize previously-overlooked individuals who have contributed significantly to the long-term wellbeing of life on Earth. Interested volunteers can apply here. The point of contact for more information on FLI’s activities is email@example.com.
Boston also hosts the North American office of J-PAL, the Abdul Latif Jameel Poverty Action Lab, which researches and promotes evidence-based policy for global poverty alleviation. J-PAL is not officially affiliated with effective altruism, but they have received grants from GiveWell and OpenPhil.Their employees have presented at multiple EA conferences in Boston, including EAG 2017.
Google’s Cambridge office employs a number of EA and EA-adjacent people. Though it’s a minor branch office of the company, less than a tenth the size of Google headquarters, it punches well above its weight during Google’s annual employee-giving initiative and sometimes beats every other office in funds raised. For the past few years, Jeff Kaufman has led Google Cambridge’s EAs in successfully lobbying to direct that money toward GiveWell-recommended charities. At between a quarter-million and a half-million dollars each year, this may be the largest fundraising event for GiveWell charities in the world.
The EA Community in Boston
Boston is not the Bay Area. It’s far from the center of EA action, misses out on much of the benefit of centralization and dense networks, and doesn’t have enough local EA-aligned organizations to employ its ambitious, dedicated graduates. There's almost no technical AI safety work to speak of in the area and little work on AI strategy either, nor meta work beyond local community-building. Many talented community members end up leaving Boston sooner or later for jobs in various other cities, with the Bay as an especially attractive destination given its EA orgs and social centralization.
All of Boston’s local organizing is done in people’s spare time, often around demanding academic or work schedules. This allows the community to do productive work very cheaply, powered by Julia's generosity for the dinner meetups plus student activity funding from the universities. Nearly any project we can conceive of will have people available and willing to work on it. Volunteer projects are also a great way to get interested newcomers more invested in and connected to the local community. However, this leaves us at the risk of people dropping off of projects as other demands on them increase, and means that most of our activities are run by amateurs. The EAGx organizing team in particular has struggled with retaining members year to year and with finding experienced team leaders who aren’t busy with other responsibilities. The Boston community might be able to increase its efficiency and impact by creating a paid community-building position, possibly through an EA Community Building Grant from CEA. It’s unclear at the moment whether we have anyone who’s qualified for this job and not otherwise occupied.
Having such a student-centered community means Boston as a whole is particularly vulnerable to the high turnover rates and leadership succession problems that tend to plague university groups. Group leaders are always relatively inexperienced and leave as soon as they’ve built up much skill. If they don’t find competent successors, their groups are at risk of declining or even going defunct, though some have recovered from this in the past. The undergraduate groups have discussed some strategies to ensure leadership handovers consistently go well, such as choosing leaders early enough to give them significant training, writing handover documents explaining how to run important club functions, and working hard to bring in freshmen each year.
Based on anecdotal evidence comparing Boston’s EA community to other local groups, its culture is relatively accessible, welcoming, and warm—though possibly less consistently friendly for people from non-technical backgrounds. People are generally accepting toward newcomers and people farther from the community’s core. It’s possible to be involved without saving the world or doing direct work on a priority program—Boston EAs are doing all kinds of different things. The independent group in particular tends to be welcoming and grounded. It also has a long history and will likely be around for a while to come.
In 2018, the Boston community transformed from an array of largely independent groups to a fledgling collaborative intercollegiate network. This was largely due to the planning process for EAGxBoston 2018, a one-day conference held at MIT and the Boston community’s single largest undertaking this year (postmortem here [EA · GW]). Boston has also hosted EAGx Boston 2016, organized locally, and EA Global Boston 2017, organized by CEA with volunteer support from the local community.
At the beginning of 2018, Taymon initiated a call for volunteer EAGx organizers. The resulting team was led by Matt Reardon of Harvard Law School, who worked with two other students at Harvard, two at MIT, one at Tufts, and Taymon themself. Since Harvard did not participate in organizing EAGxBoston 2016, this was the first collaborative project involving both Harvard and MIT EA. Many of the organizers occupied leadership positions in their respective groups. After collaborating so closely on a time-intensive project, these leaders continued to actively coordinate activities together, and their groups now communicate and collaborate much more frequently. It has become common for members of one group to attend events hosted by another, and Harvard and MIT switch off on hosting monthly intercollegiate socials.
Leaders from each of Boston’s groups met together for the first time in September, following a much smaller intercollegiate meeting in December 2017. They discussed various projects to work on together, and created a Slack channel to make it easier for them to reach each other. A group leader retreat funded by CEA in November was very helpful for increasing intergroup coordination, as it allowed leaders who barely knew each other to bond, share tips, and work on overarching strategy.
Plans and Possibilities for 2019
2018 has been a very exciting year in Boston. We hope to continue most of our current projects, build on them, and expand and better coordinate our community.
EAGx Boston 2019 will take place April 27-28 at Harvard. We are expanding to a two-day conference with a focus on facilitating networking and helping attendees come away with better-targeted plans for improving the world.
The Belfer Center’s Technology and Public Policy Group has founded the Emerging Tech Policy Network for the Spring 2019 semester. This is a career network and weekly dinner discussion group focused on major long-term policy questions surrounding emerging technology, including mitigating catastrophic risks. Individuals in the network will have exclusive access to high-profile speakers, including policymakers, academics, and technology experts, as they navigate career decisions. Applications are currently ongoing. We are excited about this opportunity for those interested in catastrophic risks.
We also hope to expand, professionalize, and improve the quality of the fellowships, possibly consulting movement-building experts to produce a well-thought-out curriculum we can be very confident in. We’d like to answer the question of how to keep fellowship graduates engaged with the community and help them past any sticking points along the road from freshly minted EAs with a solid grasp of the basics to people who can successfully take on serious projects and produce impact in the real world. The Harvard Kennedy School has also just begun its own 10-week fellowship, which adopts 80,000 Hours content for individuals focused on navigating high-impact careers in government, civil society, and public policy. 40 students, primarily from HKS, are currently enrolled.
We hope to expand, professionalize, and improve the quality of the fellowships, possibly consulting movement-building experts to produce a well-thought-out curriculum we can be very confident in. We’d like to answer the question of how to keep fellowship graduates engaged with the community and help them past any sticking points along the road from freshly minted EAs with a solid grasp of the basics to people who can successfully take on serious projects and produce impact in the real world.
Finally, we’d like to develop new student groups at Tufts, Boston University, Wellesley, and other local universities. By further growing the population of bright, energetic young EAs; expanding our network to other schools; and continuing to forge our disparate groups together into one community, we hope to make Boston an even more exciting and active EA hub in the future. We aspire to be a place where smart, creative people gather to generate, iterate, and try out ideas with real-world impact—a laboratory to develop the tools to fight suffering and make the world a better place.
Comments sorted by top scores.