My interpretation of Premise 4 (“Being an investor in such companies will generate outsized returns on the road to slow-takeoff AGI”) is that Milan is asserting a company that develops advanced AI capabilities in the future will likely generate higher returns than the stock market after it has developed these capabilities. This does not seem like a controversial claim since it is analogous to the stocks of biotech companies skyrocketing after good news is announced like regulatory approval to launch a drug. The market may have priced the probability of a company leading a slow AI takeoff in the next X years, but having that actually happen is an entirely different story.
The concept of investing in something that generates a lot of capital if something “bad” happens like investing in AI stocks with the goal of having a lot more capital to deploy if a suboptimal/dangerous AI takeoff occurs is known as “mission hedging.” EAs have covered this topic, for example Hauke’s 2018 article A generalized strategy of ‘mission hedging’: investing in ‘evil’ to do more good. Mission hedging is currently a recommended research topic on Effective Thesis.
I think the more important question, which Richard brought up, is whether having X times more cash after a suboptimal/dangerous AI takeoff begins is better than simply donating the money now in an attempt to avert bad outcomes. EAs will have different answers to this question depending on their model of how they can deploy funds now and in the future to impact the world.
The title of the article (“If slow-takeoff AGI is somewhat likely, don’t give now” at the time of writing) implies giving now is bad because mission hedging all of that money for the purpose of donating later will lead to better outcomes. I believe the article should be modified to indicate that EAs should evaluate employing mission hedging for part or all of their intended donations rather than suggest that putting all intended donations towards mission hedging and ceasing to donate now is an obviously better option. After all, a hedge is commonly known as a protective measure against certain outcomes, not the sole strategy at work.
I recommend modifying this:
I use and recommend Vanguard, because they have no transaction fees on their index ETFs.
To something like this:
I use and recommend Vanguard as a brokerage firm because they have no transaction fees for buying and selling nearly all U.S. exchange-traded funds. I also recommend using Vanguard ETFs and mutual funds because they essentially operate at cost.
Hi Alex, thanks for sharing your thoughts! This article was meant to get community input on the design of the platform which is why we did not include other details like the governance structure.
Our prototype was not custom coded, so we did not set up a shared repo. I think that open development has a lot of potential, but there can be drawbacks, particularly slow development or even abandoned projects. My intuition is that a small but dedicated group (or even one person) would be more effective than a large but noncommitted group.
I reached out to WealthSimple. Their response was:
“We do indeed support the donation of specific assets to a charity of this choice. To my knowledge, we do not have limitations regarding the amount or charities accepted.
This feature is fairly manual. It requires a number of forms to be signed and is not fully supported in-product. However, we can definitely still process a request like this.”
This is great to hear. WealthSimple’s 0.5% annual fee is twice as high as Betterment’s and WealthFront’s. A DIY approach of investing in separate ETFs for asset classes and donating them when appreciated and selling them to harvest losses when depreciated would be more optimal, but WealthSimple looks like a functional choice for people that want to donate a lot of assets.
The Betterment page on donating stock (https://www.betterment.com/donating-stock/) now does not appear to limit the selection of charities. As such, Betterment appears to be a better choice for an EA that wants a managed investment option and wants to donate substantial amounts to registered charities because Betterment makes donating assets very easy within their user interface, has a low fee, and is well established.
I agree with your heuristic in the sense that prior experience increases the probability of success, but not in the sense that it’s necessary for success.
“Long-Term Active Investing” refers to systematic asset allocation approaches designed by experts that are active in the sense of more frequent trading rather than human decision making.
My main point was to emphasize that because Antigravity Investments is already operating and producing real world outcomes, those operations and outcomes should have majority weight in evaluating this project. I agree that FINRA licensing exams are fundamental competency tests rather than measures of skill, and should be weighted accordingly. In talking with evaluators of this project, I speak almost exclusively about actual impact and progress rather than FINRA licensing or prior experience I had before starting Antigravity Investments.
Information about our actual impact and progress is currently not public knowledge. We’re on course to drive tens to hundreds of thousands of dollars to effective charities in 2019, and I think this can be increased substantially since the scope of our operations is very limited. I think this project can drive more to charity than E2G in my near-term future, which is why I’ve prioritized this project above earning to give.
“No one has reached out to me” refers to not hearing “significant bugs” in evaluations within the last year (2018) on my current plan for Antigravity Investments. My current plan differs considerably from previous plans that were run past EAs. There currently has been no overlap between people that have evaluated previous versions of the plan versus the current version of the plan, although I am interested in having there be more overlap.
Previous feedback primarily involved (1) focusing on passive investment approaches instead of active approaches, (2) focusing on donating appreciated securities, and (3) questioning the need for EA-specific investment services. Regarding point 1, I revised the plan to incorporate offering passive investment approaches, and think that such approaches are great for DIY investors and taxable investment accounts. Regarding point 2, I think donating appreciated securities is a great idea, if we were currently managing taxable accounts, I would utilize that approach. Regarding point 3, the data points I have indicate that EA organizations are currently not investing or investing suboptimally, and that staff do not have experience with nonprofit asset management. I think external providers either would not be incentivized to provide certain types of recommendations or would provide services that are fairly suboptimal and/or high fee. I am currently focused on serving EA organizations. Regarding serving individual EAs, I think that skilled DIY investing with the donation of appreciated securities is pretty optimal and doesn’t require Antigravity Investments if someone can and wants to do it themselves. Managed alternatives like, say, most robo-advisors are less optimal and may not support asset donation.
Antigravity Investments does not do trading or money-management in the sense of actively managing portfolios. Looking at our actual progress and work output (incorporating, becoming SEC-registered, managing EA capital, and using non-controversial evidence-based investing practices) my opinion is that I’ve managed to make it work despite launching the project without experience (I’m now an SEC-registered investment advisor and have experience managing portfolios).
I’ve reached out to see if you’d like to reevaluate the current plan.
I look forward to our upcoming call on Monday to compare platform designs!
I agree that talking with many individuals, which is what we have been doing and will continue to do, reduces the risk that we launch with a suboptimal version of the idea. We are very open to talking with anyone to hear their feedback (the primary reason for this post) as well as collaborating on the platform design with fellow EAs. I think our openness to incorporating insights on how to optimize the platform before and after launch reduces the chance this will have a “large negative impact.”
I agree that those bottlenecks are important and have reached out regarding how to best address them.
That’s a great idea, I’ll include a summary with action items on future long posts. Do you think I should edit a summary into the current post?
I hadn’t considered using the EA Survey as a data gathering mechanism. My initial thought is that the survey administrators might not want to include project-related questions in the survey because it might be beyond the scope of the survey, add a lot of additional questions to the survey, and not apply to many respondents. Did you have something specific in mind about what adding project-related info to the survey would look like?
Perhaps the word “project” has a different connotation than I intended. We’re seeking to support the highest impact early ventures, and included in that classification would be the past early-stage versions of every currently existing EA organization. Do you think a different term would be appropriate? The naming is very much a work in progress; I was thinking of the EA Initiatives Platform, EA Coordination Platform, or simply the Altruism.vc Platform among many potential names.
The EA Angel Group has received proposals that would match “startup” more than “project.” The majority of proposals have already received funding prior to applying to us, have teams actively working on the proposal, and have some degree of traction.
Thanks! I haven’t been able to find a public retrospective on EA Ventures but I just reached out to Tyler Alterman to see if he has any insights. We’ve also been in communication with multiple EA Grants team members.
Thanks! WealthSimple’s support for the donation of appreciated securities is not listed online, so this is very useful information for EAs to have as they evaluate investment options. Do they explicitly support this in the United States, and do they impose any restrictions on asset donations?
Hi Jonas, thanks for mentioning this, I was not aware this was happening. Unfortunately, no one has reached out to me, so I cannot comment on any potential concerns or address any rumors floating around about the project. I welcome anyone getting in touch with me to evaluate the effectiveness of the overall idea (enhancing community financial returns) and my specific implementation of this idea, since that is much more useful to me and the overall community than circulating possibly incorrect information. My email is email@example.com. As a related note, I believe making it easier to evaluate the EV of EA initiatives such as this one can provide significant value to the community, and this is part of a project I am personally working on at the moment.
It is very likely most concerns are “partly or fully out of date” as you mention, likely stemming from when I talked with a lot of EAs in finance about my very first investment related project idea in 2016, at which time I had a weak background in finance. Within the last year, this project has received positive evaluations from EA Grants, other funders in the community, and EAs in finance, and I have not received any negative feedback or heard of any concerns.
It’s interesting to see that we are receiving community responses that line up with three areas of demand: EAs who just want to fund grants, EAs who want to fund and evaluate grants, and EAs that just want to evaluate grants. There is also the category of EAs that want to perform auxiliary functions like help people assess the impact of working on an EA project and provide advising/support to EA projects that are running.
In our post, we mentioned a system targeted towards the third group (EAs that solely want to evaluate projects) involving the creation of a “distributed group of volunteer grant evaluators with expertise across many different areas of EA to improve upon the traditional model of a small centralized group evaluating a tremendous range of grants.” This system will operate meritocratically and I anticipate that it will operate very transparently (barring concerns about project confidentiality).
As Ben mentioned, for the angel group which aims to target the middle group of EAs that want to fund and evaluate grants, we will cater to what angels in the group want. It’s hard to tell if there will be strong consensus either way, or a divided group. I anticipate that at least some angels, particularly those that are confident in their grantmaking ability and process, will publicize their grants, and we definitely don’t have a problem with that.
We have acquired the domain altruism.vc as a preliminary brand name and website for our initiative. We may use https://altruism.vc/, Medium, or the new EA Forum to post grant recommendations and grants we have issued.
Thanks for the supportive words Sanjay!
We also believe that early-stage grant opportunities should be made more transparent, and we even proposed a system in our post to create an “online portal to enable the broader community to discover grant opportunities, add their thoughts on the relative merits and risks of grant proposals, and directly fund grants without an intermediary.” Making an online portal is more involved than making an angel group but it is possible we may launch something like this in the coming months.
We have already reached out to CEA regarding getting access to EA Grants’ grant opportunities. Once our angel group gets going, I intend on resuming our contact with CEA to see what we can do regarding sharing grant opportunities in the early-stage funding space.
CEA doesn’t seem to be as responsive on the EA Forum but we have been able to communicate with them via direct outreach.
Hi Remmelt, have you joined the Rethink Charity Slack? I can’t seem to find you on there.
I increased my speed of reviewing progress in the space of small project funding. There seems to be one major project related to improving centralized grant funding. 1–2 people are interested in implementing a “Kickstarter for EA projects” at some point in the future but have not started yet. The EA Peer Funding project is essentially “Kickstarter for making grants to individual EAs.” This is the extent of my knowledge based on Skyping with several people in this space. No one has mentioned anything else in the comments section of this post or otherwise.
Since there doesn’t appear to be others in this area yet, I believe moving forward with concept refinement and seeking additional feedback would be a useful next step. Let’s coordinate this via Rethink Charity’s Slack!
Agreed! I, for one, would like to know who is handling “Idea 2.” I have talked to several people working on funding small projects and have only heard about Idea 1 and Idea 3. Idea 3 doesn’t seem to have anyone actively working on it, just thinking about it.
I propose an infrastructure to generate more active qualified grant makers by making people who are close to qualified/good grantmakers (as Gregory says, good judgement, domain knowledge, relevant network, etc) into grantmakers by giving them the ability to recommend grants from a centralized fund that donors can contribute to in order to fund small projects without the hassle of evaluating dozens of projects themselves, and with the possibility of earmarking funds for specific grantmakers.
I also aim to solve the awareness problem of EA projects that are requesting funding, since EA Grants does not at present have a way for non-CEA staff to learn about possible grants, so only a handful of people can actually assess grants and people that might be great grantmakers are left out. This also requires infrastructure.
Thanks for the insight Remmelt! A good way to start this would be to create an MVP much like Ryan Carey suggested so that we can get started quickly, with a prebuilt application system (Google Forms, Google Docs, a forum, etc) and possibly using a DAF or fiscal sponsor. The web app itself could take a while, but having public projects and public feedback in a forum or something would be reasonably close and take much less effort.
I am meeting with someone who has made some progress in this area early next week. Based on traction and the similarity between the other person’s system and this system, I’ll see if a new venture in this space could add value, or if existing projects in this space have a good chance of succeeding. One way or the other I’ll be in touch!