Is it too hard to do good through EA?
By Subha @ 2025-03-03T12:54 (+55)
This is a Draft Amnesty Week draft. As I prepare to post this, I’m happy to see that it’s “Draft Amnesty Week” – I feel less unsure about initiating a thread without having read as much and heard as much from the conversations already underway, as I’d have liked to. From the quick scan of posts tagged in “Building Effective Altruism”, I can see that this post from Luke Freeman on “Big Tent” EA in May 2022 touches on some of the problems I’ve covered and triggered a lot of debate. I hope the deeper exploration of solutions below can trigger action! |
Commenting and feedback guidelines: I'm posting this to get it out there. I'd love to see comments that take the ideas forward. |
TLDR:
The Effective Altruism (EA) landscape is complex and overwhelming, making it difficult for newcomers, especially mid-career professionals, to engage. This complexity leads to low conversion rates from website visitors to active participants.
- Improve the 80,000 Hours website:
- Introduce career stage-based filtering of resources
- Implement tiered calls to action for different time commitments
- Create an integrated EA ecosystem “hub of hubs” map
- Leverage technology to simplify discovery and suggest possible next steps
- Showcase a wider range of EA-aligned career paths
- Expand and diversify support systems:
- Scale 1:1 advising with options for peers and buddy cohorts
- Leverage AI for personalized recommendations and to reduce cognitive overload
- Develop engaging and accessible content:
- Create high-impact videos on key topics
- Integrate content from various EA YouTubers
By simplifying, personalizing, and integrating the EA engagement experience, the platform will reduce barriers to entry, increase understanding, and motivate action. This will lead to a significant increase in effective donations, volunteering, and advocacy.
Add your thoughts if there are other aspects of the problem or solutions to consider.
Thanks Jay Tate for the incredible support in writing this out!... and @Nina Friedrich🔸 and @Lynn_Tan for your feedback on previous drafts.
1. My initial challenges with EA 
My first brush with the Effective Altruism (EA) community was in 2021 - when a colleague referred to 80,000 Hours in passing. This was at a time when I was actively exploring ways to get back to the social impact sector.
My first look at 80,000 Hours was fascinating… in the ~15 years that I’d toggled between the social and corporate worlds till then, never had I come across a social impact space that brought in all the logic, evidence and rational aspects that I’d till then associated only with the corporate world.
The first look showed that I’d need to ‘invest’ time to understand and meaningfully engage in EA. So I put it off to get to over a weekend…and I did get back to it eventually, a few weekends later - as it was no easy task given the full-time job and all the responsibilities of a mother that I was juggling.
I read a lot more on 80,000 Hours than I had on any other volunteering site - but I also recall a feeling of being overwhelmed by how much I didn’t know. I worked through the anxiety and worked through most of the first three sections of the 8-section Career Guide till I got to the section on clarifying my Strategic Focus areas, which I read as demanding a degree of complexity that was more than I could deal with. So I then applied for 1:1 advising for help to work through the rest of the 11-page guide and get on a better track.
Having spent tens of hours getting better informed, I was surprised when I was rejected at the point when I was highly motivated and help could have been most valuable.
But I was still very motivated, and when I applied for a job with an EA org, I was again surprised with my first rejection note. I persisted, and applied for a second EA org position, then a third, then a fourth. However, eventually by the time I received the last rejection from the ~10 jobs I’d applied to, I was pretty much done with EA. At the risk of sounding immodest, I believed I had a lot to offer and nobody seemed to want any of that Career Capital. The history of success in the ~20 years of my career till then hadn’t prepared me for this.
I moved on to volunteer full-time with a non-EA non-profit for ~2 years.
Only during Summer 2024 - when I started thinking of a transition – did I take another shot at engaging with the EA community. Yes, I’d been at a loss to learn from what had been almost a dozen rejections from the EA community… but I still identified with the philosophy and respected the community.
2. The difference with High Impact Professionals
I’d followed High Impact Professionals (HIP) with interest, given its focus on enabling mid-career professionals like me. And, the Fall 2024 cohort of the Impact Accelerator Program (IAP) turned out to be a fantastic opportunity for me to take a different approach, which has been far more effective for me.
One eye-opening insight for me occurred during Week 2 of the program, during a reading on Mapping our Impact Path: Seeing the Absorbency scale gave me a simple and wonderfully helpful overview of multiple paths to impact that I had been missing during all my previous efforts.
Seeing this relatively simple diagram is when it hit me that in my enthusiasm to move back to the social impact space, I’d prematurely focused exclusively on ‘working in EA’ as the highest potential path to impact. I had registered the Earning-To-Give model as an option when working on my Career Guide – but during and after the onslaught of EA job rejections, my lack of a framework for learning from each rejection left me unproductively stuck.
However, as I worked through the wonderful 6 weeks of the IAP – with the ultra-distilled readings and simpler action-worksheets, I can now see, in hindsight that it was my miss - that I didn’t move on from Plan A to even a Plan B let alone a Plan Z. Things I’ve learned in the IAP could have helped me avoid becoming merely tired from seemingly fruitless time and work already invested - and helped me reframe my disappointment with how blind the space seemed to the career capital I had to offer.
Features of the IAP program that could have helped me sooner include:
- A really well-curated and distilled reading list… just really a lot less reading… with enough cross-references to dig deeper only if one wants to
- Action Plan Templates which parsed the journey of exploration into micro-steps that were most relevant to where I am right now
- Personalized feedback and recommendations from my Cohort + Facilitator on the plan / alternatives to consider etc.
The micro-steps helped me narrow from an initial list of 30+ options to 3 prioritized choices in my Plans A, B and Z.
And from the conversations with others in the IAP I’ve discussed this problem with, it appears my story is not unique. Seven of eleven participants I’ve talked with agreed that the process of exploring the EA philosophy + community and identifying best ways to engage demands a lot of time and effort.
This makes me wonder about possibilities such as:
- How might 80,000 Hours leverage and learn from the IAP?
How might High impact Professionals itself distill and expand the reach of the IAP beyond cohort participants, perhaps making some components available for busy career professionals who don’t feel they can commit to a cohort-based program?[1]
3. Toward a Theory of Change
Reflecting on this has also led me to a broader type of question:
- To what extent might the very rigor or complexity of EA become an entry barrier to many (especially mid-career folks) who want to do the most good they can?
- Might they be more likely to increase their impact if their paths to engage were simpler?
In particular, what might be needed to bring the benefit of the IAP distillation and simplicity to potentially every one of the ~1M visitors who visit 80,000 Hours every year? The 80,000 Hours website is a treasure trove of content on the Effective Altruism philosophy and years of research and exploration. However, engagement rates[2] are lower compared to Idealist - a US-focused website that supports users interested in nonprofits.
Is there value in augmenting, or perhaps even re-framing the 80,000 Hours welcome-page content to better meet mid-career professional where they are, such as:
Feature simple calls to action, e.g., arranged from just 5 minutes to “if you have a bit more time” to “I’m all in. What can I do?”.[3]
- Allow Users to select modified versions of the 80,0000 Hours Landing Page by indicating Mid-Career Stage: Allow users to filter resources by career stage (e.g., "early career," "mid-career," "late career").
- Feature High Impact Professionals prominently on a Mid-Career version of the 80,000 Hours landing page.
Create an integrated view of the EA ecosystem: there are many ‘hubs’ of information within EA for a mix of cause areas, career stage, channels of engagement (for AI safety, Animal Advocacy, mid-career professionals, effective giving etc.). Build a hub-of-hubs which draws an integrated view of the different pathways for engagement[4].
- Expand the 1:1 advising system 10x:
- in order to expand 1:1 availability from 80,000 Hours perhaps explore creating a lower tier of near-peer or volunteer advisors.
- Facilitate a one-time “buddy” system or support-group system for bringing together people at comparable stages in their EA journey.
- Consider referring site visitors to the PeterSinger AI for quick answers to questions about Animal Rights, Global Poverty, Applied Ethics.
- Beyond "full-time" EA Roles: Given the low likelihood of EA-organizations generating full-time roles for the ~3K professionals already a part of the HIP Talent Directory (and possibly ~10K professionals with interest in this space per AIM estimates), highlight this demand-supply skew from early on (for instance - put the Absorbency Scale for Money-Network-Skills, front & center on the homepage). Encourage exploration of other channels of engagement - including earning-to-give, networking, volunteer and EA-adjacent careers.
- When 80,000 Hours begins making YouTube videos in 2025, consider placing special emphasis on highly effective videos that address key stumbling-point topics, such as:
- high-impact donations (regardless of what else you do).
- redeploying Career Capital from non-EA to EA-organizations.
- volunteering as possible stepping-stone to full-time EA activity.
- reskilling for EA.
- Finally, drawing on the leverage of AI or even older technology and UX tools to simplify such explorations; - as demonstrated by PeterSinger AI; is there value in a solution that
- curates user-friendly educational content on EA principles, cause areas, and opportunities better adapted to the limited time of busy professionals.
provides personalized recommendations and guidance to support individual exploration and action planning[5]?
The core reco is to make the 80,000 Hours website even better for current users. But more importantly, this new approach could bring in a lot more people from the wider world of charitable giving. This could lead to much bigger impact in several ways:
- Donations: How can we encourage more of the huge amount of money given to charity each year (around $550 billion in the US alone, and much more globally) to go to the most effective causes? [6]
- Volunteering: How can we get more of the 850 million people who volunteer each year to focus their efforts on high-impact opportunities? [7]
- Advocacy: How can we make our advocacy efforts ten times more effective in reaching and influencing policymakers?
- Careers: How can we inspire ten times more people to start or switch to careers that have a big positive impact?
Theory of Change: Increasing EA Engagement for Mid-Career Professionals
Situation | The Effective Altruism (EA) landscape is complex with the array of choices across cause areas, functional areas and modes of engagement (money, skill, network) (not counting geo and level of experience!). The complexity of navigating these options - aggravated by the text-heavy interface - creates barriers to entry for newcomers, particularly mid-career professionals. This results in low conversion rates from website visitors to active EA participants. | Final Goals | To significantly increase engagement in effective altruism - particularly mid-career professionals, leading to a substantial rise in effective donations, volunteering, advocacy, and career choices aligned with EA principles. |
Inputs and | Outputs | Change mechanism | Outcomes | Impacts |
Improve 80,000 Hours Website:
Expand and Diversify Support Systems:
Develop Engaging and Accessible Content:
|
|
|
|
|
Evidence assessment | Moderate. Evidence from website traffic data, engagement statistics, and previous attempts at simplifying EA engagement (e.g., the 80,000 Hours quiz) suggest that fragmentation and complexity are barriers to engagement. See Appendix for further evidence of the problem. Further research and data analysis will be needed to strengthen the evidence base (eg.: assess EA-engagement/impact of applicants to the 80,000 hrs advising program and/or IAP who did not make it to the programs.) |
Assumptions |
| Possible unintended consequences |
|
4. Your thoughts?
While 'solving' this was one of the early motivations for writing 'this', the exploration of the solutions makes it clear that adding a new hub is the least efficient way to solve for the challenges flagged above including the lack of a hub of hubs. So, I'm putting this out there to see if the questions merit debate - and if yes, attract the right attention from others already in the ecosystem to solve for it! I'm happy to stay engaged in ideation and exploring ways to solve the problems. If any of this resonates -- you could share your experience/thoughts/recommendations to build the discussion and possibly help build the solution!
Appendix - Evidence of Relatively Low Engagement for 80,000 Hours
Low Conversion: Despite ~1 million annual visitors to 80,000 Hours (10 million total)[8], only a tiny fraction actively engage with EA:
Mid-Career Disconnect: ~80% of 80,000 Hours visitors are 25+[11], already 5-10 years into their careers, yet resources are often geared towards younger audiences. This is aggravated by the fragmented experience of exploration – where one may need to hop from 80000 hrs to Giving-What-We-Can to Successif/Hive other such sub-area specific hubs. Each of these are needed – but is there an opportunity to create easier visibility to these resources and simplify the effort needed to identify the best paths for deeper exploration?
Comparatively lower engagement: A manifestation of the lower engagement in 80,000 Hours could be in the lower engagement stats compared to Idealist - which has more traffic, longer average time on site, and lower bounce rate (i.e., leaving the site after viewing only one page) despite being a US-focused site for individuals looking to do / more good[12] and not a more global site like 80,000 Hours:
Traffic: Idealist.org receives more traffic (1.9 million visits till Nov 2024) than 80,000hours.org (1.1 million visits till Oct 2024).
Idealist.org Global Rank: #30,358
80000hours.org Global Rank: #52,586
Time on Site: Visitors spend an average of 3 minutes 15 seconds on Idealist.org, compared to 2 minutes 11 seconds on 80000hours.org.
User Engagement:
Idealist.org has a lower bounce rate (41.37%) than 80000hours.org (49.6%). (Bounce rate is the percentage of visitors who enter a website and then leave rather than continuing to view other pages within the same site.)
Visitors view more pages per visit on Idealist.org (4.18) than 80000hours.org (3.30).
- ^
Perhaps similar to how Blue Dot Impact offers both cohort-based and non-cohort ways to take its courses.
- ^
Bounce rate of 49.6% compared to 41.37% for Idealist; also users spend an average of 2 minutes 11 seconds on 80000 Hours - lower than 3 minutes 15 seconds on Idealist. See Appendix for additional evidence of lower engagement for 80,000 Hours than the Idealist site.
- ^
For example, see The Climate Reality Project for an illustration of calls-to-action arranged in increasing levels of effort and commitment.
- ^
For example, see Landscape Map: AI Safety World
- ^
For example, see Navigation & Guidance in the Age of AI: 5 Trends to Watch for an overview of how AI is being leveraged in career guidance; or on a broader basis - leverage of technology similar to those used by platform models like Airbnb which match demand and supply using criteria/preferences to build personalized recommendations
- ^
- ^
- ^
- ^
- ^
- ^
- ^
gergo @ 2025-03-05T07:57 (+4)
Thanks for writing this up!
Scale 1:1 advising with options for peers and buddy cohorts
For those interested in AIS, aisafety.com has a list of advisors to whom you can talk.
Karen Singleton @ 2025-03-04T04:47 (+3)
Thanks for sharing your post on what I think is an important topic.
I like how you have clearly stated the problem and you make a great point about the complexity of engaging with EA, especially for mid-career professionals. I agree the HIP programme seems to be an effective way to navigate this space more smoothly, and I can sympathise with the challenges you’ve highlighted.
Building on your ideas regarding career stage-based filtering and personalised recommendations, making it easier for newcomers to find relevant pathways without cognitive overload, could there be an interactive "guided exploration" tool—perhaps an AI-driven assistant or decision-tree quiz—that helps users quickly assess their best starting point based on their background, interests and constraints? I wonder if anyone has experimented with similar tools before or are there existing resources that could be adapted for this?
Subha @ 2025-03-04T19:10 (+2)
Thanks for your response Karen.
Interestingly - the idea of building an AI-led solution was the starting point of this journey for me :-) I'm not pursuing it for a number of reasons (especially - it's sub-optimal to add another organization in what I see as already a crowded ecosystem which is tough to navigate). However, for those interested in pursuing this -- this article has a useful overview of current usage of AI-tools in supporting Career Counselling in colleges - Navigation & Guidance in the Age of AI: 5 Trends to Watch.
Also, at a broader level in Analytics modeling -- this falls under the broad purview of matching-problems -- which form the core of personalized recommendation engines in most platform-model businesses. I'm not tech-enough to detail this further -- but can help interested folks connect with experts if they'd like to learn more.