6

Update on Envision: progress thus far and next steps

I’m the next president of Envision, a student group working to impart a safety-conscious mindset towards technology to future leaders. Last year my predecessor Luca Rade wrote a post which laid out Envision’s mission and how we intend to accomplish it. This post will provide a summary of what Envision accomplished this past year, present the challenges we faced in implementing the plan Luca laid out, show how we’re refining our strategy to confront these challenges, and give our plan for the future.

 

In an effort to make this post self-contained, I’ll provide a summary of Envision’s mission as laid out in Luca’s post. I highly recommend reading his original post for the complete picture on our mission.

 

Envision’s mission is to imbue a safety-conscious mindset towards technology in industry, political, and academic leaders. We strive to do so by exposing young people likely to be leaders in these domains to the risks advanced technologies present and convincing them that these risks necessitate action to ensure technology is developed prudently. Our events attract young future domain leaders by cultivating prestige, bringing leadership in fields they’re interested in, and focusing on how a fundamental concern for safety will allow domain leaders to achieve positive results. Luca’s post expands on the specific mechanisms we use to instill a safety-conscious mindset.

 

What we accomplished last year

 

Envision Conference

 

Last year Envision held our first annual Envision Conference. The conference brought together 150 participants from 38 institutions and 8 countries to analyze the future implications of technology. Andrew Critch, Robin Hanson, and Anders Sandberg all spoke at the first Envision Conference and will be returning this year. Additional speakers included Luke Nosek, a co-founder of PayPal and Founder’s Fund; Andreas Mershin, a physicist at MIT; Jeremy Kashin, a professor at Princeton and scientist at NASA; and other accomplished researchers and professionals actively thinking about the risks and possibilities in our future.

 

Envision Conference’s mission is to select young participants who are likely to be leaders in developing, implementing, and/or regulating future technology and teach them the importance of developing technology with a safety conscious mindset. We believe last year’s Conference made progress towards achieving this goal based on the following data points:

 

  • The conference had a counterfactual impact by targeting individuals who are not typically reached by the EA community.

    • We advertised the conference heavily to groups with technically excellent participants, and these groups are primary focused on developing technology rather than analyzing its implications. The prestige and positive word of mouth publicity our conference developed helped us also passively attract qualified participants to our applicant pool. These individuals are likely to have a role in developing technology, and have likely not previously thought deeply about the risks associated with technology. Targeting young participants also allows us to influence their belief system while they’re still developing it, embedding a concern for prudence more deeply than if they are exposed to it once already on the path towards domain leadership.

    • Travel reimbursement enabled us to reach qualified participants who are unable to attend other conferences, or require a strong incentive to find a conference worthwhile to attend. Our competitive and holistic admissions process helped us ensure that these participants are also likely to become future leaders.

    • By coupling the positive potential of technology to its risks, we were able to attract entrepreneurs and others who are not attracted to EA due to the excessive sense of obligation it can engender.

  • Our conference events reached many people who were unable to attend the conference. The live stream of our closing ceremony had 1400 viewers and our panel on AI safety had close to 250 live participants (including Princeton students).

  • 48% of conference participants reported an increased sense of agency towards shaping the future.

 

Many of our participants showed that their opinions were changed or took measurable steps to change their actions after attending the conference. Here are some of their quotes and stories:

  • “I used to think that the path toward progress was taking place in lock-step with a clear goal at the end, but now I realize that the future could very easily be hijacked and taken in a different, harmful direction.”

  • Participants report that the conference helped engage them with a new community. Engagement in this community remains high after the conference; many repeat applicants to the conference wrote  that they still are in contact with individuals they met at the first conference. This community helps sustain the connection our participants have with the mission of Envision. As I will explain in the “Leverage the Matrix” section, we’re going to take steps to continue building this community.

  • I fit the demographic the conference attempts to target well: my primary interest is in technical businesses. I’ve worked as a programmer since I was a sophomore in high school, have developed numerous technical projects that are used by thousands, and did an internship at a major security company.  Before the conference I was primarily interested in developing technology with no regard to safety. I joined Envision after attending more of their events and quickly became heavily involved. The conference personally changed the way I view the future, and catalyzed my strong desire to help mitigate existential risk.

 

While we believe our conference last year was successful, we do have substantial room to grow. One of our primary areas for growth is to establish better data-gathering procedures and a process for measuring the longer-term impact of the conference on conference participants. To acquire better data on the long term impact of the conference we’re going to expand our post-conference survey and continue to follow up with participants in the medium and long terms. We’ll increase the number of participants that participate in the survey by cultivating a community of conference participants and incentivizing participation. The “Leverage the Matrix” section explains this plan in further detail.  

 

The conference’s longevity is still unclear, and we are working to ensure that it will continue to be successful throughout leadership transitions. I’ll detail our plan to increase the conference’s and Envision’s longevity in the “Secure Longevity” section.

 

Engaging the Princeton community

 

We hosted many events throughout the year at Princeton in addition to the conference. The events we hosted include panels on technology policy with professors with governmental experience, talks with top researchers who spoke about their work to confront the potential negative consequences of technology, discussions among students on the implications of artificial intelligence and other technologies, a book club, and trips to Boston and Silicon Valley where Princeton students met with top companies and researchers about their work.

 

The trips to Boston and Silicon Valley had the largest overall impact of our Princeton-based events. Both trips had 15 participants, but because of the high value and intensity of the trip the application process was very competitive and many of the members of these trips made dramatic shifts in their professional and collegiate plans after attending the trip. One participant switched their research focus and ultimately switched careers after learning what technologies will be most impactful, another student who’s now running our entrepreneurship competition narrowed their general interest in venture capital to identifying businesses that leverage a successful business strategy to catalyze technical and social change, and a participant interested in software engineering and finance now actively works for Envision to promote our mission.

 

The trips to Silicon Valley and Boston were able to effectively attract participants and bring a competitive pool, but many of our other events were unable to attract a significant number of participants. This accurately depicts one of the primary challenges Envision faces: inability to attract future leaders to our smaller scale events.

 

Challenges

 

Attracting Future Leaders

 

Luca addressed the concern that Envision will be unable to attract domain leaders in his post. We  believe, based on the quality of our participants, that we’ve have so far broadly succeeded in attracting domain leaders at the Conference and our trips. Our smaller events have mostly failed to do so. Most undergraduates are busy, and the undergraduates that we target are generally busier than average. As a result our panels, book groups, discussions, and other smaller events have struggled to attract a meaningful number of participants.

 

We believe that the larger events are able to attract quality participants because of their prestige and the high value they provide to participants. Our conference is able to bring world-class researchers and speakers, helping us attract world class participants. Based on early bird applications, the conference this year has also benefited from a substantial network effect: previous participants promote the conference and its mission to their peers, spreading Envision and its mission. The trips have experienced similar success because of the value that they provide participants.

We will continue to hold some smaller events since they provide value for the people who do attend, but we will shift to spending more time developing a smaller number of high value events. In addition to our main conference and trips, we’re planning on hosting a smaller conference in Boston  to help establish a second hub for Envision and engage another community.

 

Effective Officers

 

Envision has been successful because a small number of officers commit a disproportionate amount of effort. We’ve recruited heavily and receive more than fifty applications to become an officer every year, but we struggle with maintaining long-term, effective officers that grow Envision. Envision’s success is dependent on a substantial amount of work from its officers, and few of the people we recruit are willing to commit the necessary effort.

 

Our interviews with ineffective officers show us that many capable officers that believe in Envision’s mission are more passionate about integrating it into their lives than spreading it to others. These officers were interested in joining Envision, but shifted their effort to ventures where they see a greater potential for personal benefit. We’ve used this information to change how we select officers. Previously, we preferred candidates with previous initiative and technical expertise--we thought that those factors were indicative of highly capable officers. We now see that while these officers are likely highly capable, we can’t guarantee that they’ll direct their talent towards Envision. We now seek to identify officers that will dedicate time and effort to Envision by favoring applicants passionate about spreading Envision’s mission. We can train officers and help develop aptitude if they have passion, but we can’t effectively use capable officers that aren’t passionate about working for the organization. Specifically, we’ve substantially reduced the weight of technical competence, and instead look for passion for spreading Envision’s mission. These adjustments will be an iterative process. Genuine passion for spreading Envison’s mission will be hard to identify and we’re actively refining our processes and exploring new ways of doing so.

 

Growing other Envision Chapters

 

After the conference several participants from different schools expressed interest in starting an Envision chapter at their school. While several of these participants started chapters at their school, none had more than a few events. Our conversations with the founders lead us to believe that they were two primary reasons for these chapters’ failures:

 

-Lack of a framework for Envision chapters. We actively supported the chapter founders and helped them plan events, but we did not provide a detailed framework on how to create an Envision chapter. We thought that the founders were best suited to understand what events would be successful and beneficial to their schools. A detailed framework on how to create a chapter felt pedantic and like it robbed the founders of the ability to take ownership of their chapters. We also felt that we would be unable to provide a framework that would guarantee the success of an Envision chapter since our chapter’s success was partially based on context, luck, and the Herculean effort of a few individuals.

-Lack of sufficient interest from chapter founders. The chapter founders were very interested in the principles of Envision and genuinely wanted to help, but creating a successful chapter requires an obsessive group of founders who devote almost all of their time to growing the chapter. The challenges Envision continues to overcome are hard, and we haven’t found a scalable way for chapters without Envision’s central leadership to overcome them.

-Envision currently isn’t well-established enough to inspire competent participants to create a subsidiary organization. Many of the most capable participants of the conference who were interested in continuing to work on Envison’s mission were hesitant to create an organization that’s derivative of Envision. They wanted to create distinct organizations that could achieve a similar level of success, and in one notable case they failed to do so because they tried to partially reproduce Envision without adopting key aspects that drew sponsors and participants.

 

Our plan going forward

 

Leverage the Matrix

We have learned that it’s not a feasible short term goal to create a network of active Envision chapters at different universities. Envision chapters require dedicated leadership and we aren’t able to effectively train conference participants to become the leader a chapter requires. We will continue to support members if they want to create a chapter, but we’re going to shift our effort to leverage our network in more effective ways. Envision has grown a substantial network of collaborators across the world that are passionate about our mission, and we intend to leverage this network to address our challenges.

 

Envision’s network can be used to keep participants outside of Envision engaged with our mission and help them apply our principles to their work. We’re going to help our network connect by creating an internet forum (likely a private subreddit) where participants in Envision events are encouraged to post updates, discuss ideas, and collaborate on projects. We have empirical evidence that suggests that this resource will be utilized: the Facebook page for last year’s Envision conference is still moderately active despite little effort on our part in encouraging this almost nine months after the conference.

 

We’re going to focus our effort on establishing a second Envision “hub” in one geographical location. By focusing on expanding to one area, our officers will be able to actively work to plan events with members of the local community. We’ll host a smaller conference in the community to engage community members and help build momentum. Right now Boston is our target for the next hub of Envision. We have heavily involved members and an officer in Boston, have partnerships with several schools, and believe that Boston’s combination of prestigious schools and innovative startups will make it a high value target. Envision will complement the value provided by the Boston-based Future of Life Institute by using our unique position to engage with students, and young professionals.

 

Secure Longevity

 

Leading Envision requires an effective leader who’s willing to devote all of their time. Finding an undergraduate who combines high competence with extreme work ethic and passion is incredibly challenging, even at top schools. Finding a leader is challenging, but we’ve had promising success in identifying and training individuals who may fit the role. We look to identify candidates to lead Envision by selecting participants who are extremely passionate about the mission and have some previous leadership experience. Many of these candidates have underdeveloped resumes, but their passion and potential is clear from speaking with them. Our trips have proven to be the most successful way to attract these candidates and show them the importance of the issues Envision addresses: they give us a unique ability to intimately interact with people interested in Envision’s mission and learn more about their background while providing intense exposure. n. There is no absolute formula for identifying capable leaders. Envision’s mission may dilute over time, and one extremely misaligned leader could bring the end. We’ll mitigate this risk by having a Board, consisting of former presidents, select the new president, provide mentorship, and in extreme circumstances step in if Envision is not making progress towards its mission.

 

Beyond leadership the biggest obstacle to Envision’s success is monetary support for our events. We’re fairly confident that our large events provide significant value and are able to effectively attract domain leaders and will increase this confidence in the future as we gather better data, but these events are expensive. Even with leaders who devote all of their time outside of classwork to Envision, we struggle to fund the events. Our conference last year was made possible partly by a donation by a member of the EA community, but we can’t become dependent on private donations without a clear prospect for continuity.

 

Envision is seeking large corporate sponsors for the conference to provide recurring contributions to help secure its financial longevity. We’ve had promising initial success--we’ve already confirmed Microsoft and Milliporesigma (Sigma-Aldrich) as sponsors of this year’s conference.  Although we have yet to reach our funding target, we believe we are on track to raise the budget we need to run the Conference with similar events as last year.

 

However, the fundraising process is currently the primary focus of Envison’s current leadership, and given the difficulty of raising the funds last year and this year, the capability to re-raise the money for the conference every year is not clear. Additionally, we are not able to spend time developing new events and growing the organization because we are focused on raising the funds for the conference and do not have sufficient funds beyond what is necessary for running the conference in its current form.

 

Therefore we’re raising an endowment to help secure the financial future of Envision and help it weather a temporary absence of competent and/or aligned leadership. An endowment will help secure our monetary future by significantly reducing the burden on leadership to raise substantial sums of money for every individual event so that we can focus on developing the organization and increasing our impact. It will provide initial momentum to the leadership of events to raise the remaining funds, enabling us to expand our reach. The endowment will help ensure that Envision is able to survive leadership transitions by reducing the burden on the new leader to spend their time raising money, and provide a catalysis for potential leaders to feel like joining Envision will have a substantial impact. The interest from the Endowment will initially be used primarily to help Envision establish a second hub in Boston, sustain our conference, and continue to run our high impact trips. The endowment will not fully fund any of these activities, but it will provide vital initial momentum the organizers of these events can use to raise additional funds and run the events.

 

Spending from the endowment will be regulated to ensure that it is used effectively every year. The president will create a spending plan for the annual interest from the endowment that will be approved by the board of previous presidents. No more than 50% of the interest can be used for a single event and the principal will remain untouched, and additional funds from events will contribute to the endowment’s principal.



To ensure that the endowment continues to be used effectively, the charter will mandate that the endowment will be donated to an organization within Effective Altruism dealing with existential risk, to be determined by the Board of Trustees, if Envision fails to host a successful Envision Conference with success defined by the following criteria:

  • At least 100 participants from 10 or more institutions

  • At least 10 speakers, with the following conditions:

    • At least half are not from Princeton.

    • At least two are affiliated in some way with EA.

    • At least one is speaking on existential risk.

  • At least $15,000 raised from corporate sponsorship and Princeton support

    • Excess funds will be redirected to the endowment if not necessary for the conference.

 



Envision Conference 2017

 

Envision Conference 2017’s initial progress shows exciting evidence that Envision has been effective in engaging young leaders and cultivating prestige. The momentum from the first Envision Conference enabled us to acquire a better speaker roster than last year including: three accomplished professors at MIT; the former CTO of the FTC and professor at Harvard; an executive at a major government contractor with hundreds of millions in annual revenue; accomplished researchers in Artificial Intelligence; the CEO of a successful government-backed startup; and a former member of the National Security Council.

 

We’ve also adjusted our focus to more explicitly separate the trends in technology that we’re focused on from their collective implications -- these can be found on our website. Additionally, this year’s conference theme is Action, to reflect an increased concern with concrete positive results from both Envision’s and our participants’ efforts and to fill what we see as the main gap in current discourse.

 

We’ve expanded from 150 participants to 200 and have already attracted many highly accomplished young leaders before we’ve started publicizing the conference. Applicants in our early bird round include Ph.D candidates from top universities, founders of funded startups with up to 7 figures in revenue, winners of prestigious science competitions including Intel ISEF, founders of nonprofits with hundreds of members, student partners at VC firms, and other qualified applicants. The majority of applicants previously unaffiliated with Envision reported that they heard of it from a recommendation from a friend. We will begin full-scale marketing within the next week.

 

The theme of Action for Envision Conference 2017 reflects our increasing focus on providing measurable value: speakers will give concrete proposals on what we can collectively do or what participants can individually do to ensure the safe and beneficial development of technology.. Participants will have access to resources like speaker office hours, high-profile companies and startups seeking recruits, and entrepreneurial mentorship to help them take definite action on what they learn at Envision Conference.

 

How EA can help

 

By continuing to grow our flagship conference, hosting trips to engage members of the Princeton community with industry leaders, and establishing another hub of Envision we’re confident that, with the necessary resources, Envision can continue to have a high counterfactual impact on how future leaders develop, implement, and regulate technology. Our ability to effectively do these things is contingent upon finding support.

 

The easiest way effective altruists can support Envision is by helping us spread news of our conference to qualified participants. The application is currently open and will close on October 9th. We welcome telling peers about our conference, posting on other forms about Envision, helping us reach relevant mailing lists, or other ways of spreading Envision.

 

Envision Conference benefits from partnerships with both nonprofits and corporate sponsors. If you’re involved with or know of an institution that would be a good match to partner with Envision, please connect us.

 

The most impactful way you can support Envision is by contributing to our endowment. The endowment will enable us to run our events better, expand and better measure our impact, and ensure longevity. As I previously wrote, the endowment will be donated to EA if Envision no longer effectively accomplishes its mission, as determined by specific, measurable criteria. As an officially registered 501c3 your donations are tax deductible.

 

If you’d like to support Envision through any of these ways, I can be reached at ajs9 < a t > princeton ( d o t ) edu or by direct message.

 

Summary

 

Envision is an undergraduate group working to fundamentally change how young leaders view research and innovation. We recognize that the difference between technology bringing human extinction and improving our lives will depend on how it’s developed and implemented. We target young people likely to be leaders to maximize the impact we can have on how technology will be developed in the future, and work to cultivate a safety-conscious mindset.

 

This post described what Envision has accomplished so far, what we’ve learned about the best ways to carry out our mission, and how we’re adjusting our strategy to maximize on the areas we’re effective.

 

Envision has momentum and is at the point where we have the potential for a significant, lasting impact. Our ability to secure this impact will depend on if we are able to secure Envision’s longevity.

 

We welcome any feedback on Envision’s plans or suggestions on the best way we can accomplish our mission.



Comments (3)

Comment author: zdgroff 03 October 2017 05:31:36PM 0 points [-]

This seems like an excellent program. If the program is selective at all or you have outreach that is limited and could be varied randomly, have you thought of doing a randomized evaluation of some sort? I've realized that even when we're working on the far future, there are often intermediate outcomes (e.g. changed attitudes) to focus on, and we should test our impacts on those as rigorously as possible.

Comment author: aspencer 07 October 2017 05:29:40PM 0 points [-]

Our conference does have an application process. To provide accommodation for participants, travel reimbursement, and events that allow participants to engage with speakers we can only have a limited number of participants. Right now, we seek to maximize the impact that the conference has by selecting participants who are likely to be leaders in the future. We've narrowed down criteria we believe are indicative of becoming future leaders.

That said, I do agree with your point and I worry about our ability to accurately predict who will become a leader in the future. There are many possible paths to becoming a leader in the future, and what you do in college may not reflect well what you'll do in the future. Beyond that, applications are subjective and we likely have a high false negative rate. Some people are better at writing applications, and skill in writing applications doesn't imply (I think) the applicant will be successful in the future.

We've considered doing randomized evaluation, but haven't yet because we don't know of a rigorous way to measure how we did selecting participants that will have an impact in the future. An impact in the future would, by construction, have to be measured years after the conference. We'd have to track randomized applicants and applicants we selected and have a way to measure their impact, which would be challenging if not infeasible. Right now, the best heuristic we have for identifying applicants that will be leaders in the future is by looking at what they're doing now.

The act of having an application also provides self selection for applicants; many people have communicated to us that completing the application is a barrier to entry. The application itself selects for a certain level of interest in the conference (the interest required to complete the application).

We like to believe that this approach is better than random, but we do admit that it may not be.

Comment author: John_Maxwell_IV 28 September 2017 09:00:01PM 0 points [-]

It sounds like you are doing great work. Congratulations on everything you've accomplished.

You've probably thought about this already, but have you thought much about leveraging the popularity of your large events to increase attendance at your small events? (E.g. encouraging attendees to the large events to sign up for a mailing list where the small events are advertised.)

An idea for seeding new Envision chapters is to get people who are starting new chapters to precommit to weekly Skype calls so you can keep them inspired and figure out where they're failing. Or more broadly, do some case studies of campus student organizations that have successfully seeded chapters in lots of different schools, and try to figure out what they're doing right.

For what it's worth, I think I would relax the conditions for the case where your endowment gets donated away. I can easily imagine you guys missing just 1 or 2 of the criteria you specified and still being an effective organization. Of course you'll have to talk to your funders about that.