Machine Learning Engineer @ PayPal
4760 karmaJoined Jul 2019Working (0-5 years)New York, NY, USAsunyshore.substack.com



I'm interested in effective altruism and longtermism broadly. The topics I'm interested in change over time; they include existential risks, climate change, wild animal welfare, alternative proteins, and longtermist global development.

A comment I've written about my EA origin story

Pronouns: she/her

"It is important to draw wisdom from many different places. If we take it from only one place, it becomes rigid and stale. Understanding others, the other elements, and the other nations will help you become whole." —Uncle Iroh


Sorted by New


Philosophize This!: Consciousness
Mistakes in the moral mathematics of existential risk - Reflective altruism
EA Public Interest Tech - Career Reviews
Longtermist Theory
Democracy & EA
How we promoted EA at a large tech company
EA Survey 2018 Series
EA Survey 2019 Series


Topic Contributions

I can speak for myself: I want AGI, if it is developed, to reflect the best possible values we have currently (i.e. liberal values[1]), and I believe it's likely that an AGI system developed by an organization based in the free world (the US, EU, Taiwan, etc.) would embody better values than one developed by one based in the People's Republic of China. There is a widely held belief in science and technology studies that all technologies have embedded values; the most obvious way values could be embedded in an AI system is through its objective function. It's unclear to me how much these values would differ if the AGI were developed in a free country versus an unfree one, because a lot of the AI systems that the US government uses could also be used for oppressive purposes (and arguably already are used in oppressive ways by the US).

Holden Karnofsky calls this the "competition frame" - in which it matters most who develops AGI. He contrasts this with the "caution frame", which focuses more on whether AGI is developed in a rushed way than whether it is misused. Both frames seem valuable to me, but Holden warns that most people will gravitate toward the competition frame by default and neglect the caution one.

Hope this helps!

  1. ^

    Fwiw I do believe that liberal values can be improved on, especially in that they seldom include animals. But the foundation seems correct to me: centering every individual's right to life, liberty, and the pursuit of happiness.

Thank you for posting this! I've been frustrated with the EA movement's cautiousness around media outreach for a while. I think that the overwhelmingly negative press coverage in recent weeks can be attributed in part to us not doing enough media outreach prior to the FTX collapse. And it was pointed out back in July that the top Google Search result for "longtermism" was a Torres hit piece.

I understand and agree with the view that media outreach should be done by specialists - ideally, people who deeply understand EA and know how to talk to the media. But Will MacAskill and Toby Ord aren't the only people with those qualifications! There's no reason they need to be the public face of all of EA - they represent one faction out of at least three. EA is a general concept that's compatible with a range of moral and empirical worldviews - we should be showcasing that epistemic diversity, and one way to do that is by empowering an ideologically diverse group of public figures and media specialists to speak on the movement's behalf. It would be harder for people to criticize EA as a concept if they knew how broad it was.

Perhaps more EA orgs - like GiveWell, ACE, and FHI - should have their own publicity arms that operate independently of CEA and promote their views to the public, instead of expecting CEA or a handful of public figures like MacAskill to do the heavy lifting.

I've gotten more involved in EA since last summer. Some EA-related things I've done over the last year:

  • Attended the virtual EA Global (I didn't register, just watched it live on YouTube)
  • Read The Precipice
  • Participated in two EA mentorship programs
  • Joined Covid Watch, an organization developing an app to slow the spread of COVID-19. I'm especially involved in setting up a subteam trying to reduce global catastrophic biological risks.
  • Started posting on the EA Forum
  • Ran a birthday fundraiser for the Against Malaria Foundation. This year, I'm running another one for the Nuclear Threat Initiative.

Although I first heard of EA toward the end of high school (slightly over 4 years ago) and liked it, I had some negative interactions with EA community early on that pushed me away from the community. I spent the next 3 years exploring various social issues outside the EA community, but I had internalized EA's core principles, so I was constantly thinking about how much good I could be doing and which causes were the most important. I eventually became overwhelmed because "doing good" had become a big part of my identity but I cared about too many different issues. A friend recommended that I check out EA again, and despite some trepidation owing to my past experiences, I did. As I got involved in the EA community again, I had an overwhelmingly positive experience. The EAs I was interacting with were kind and open-minded, and they encouraged me to get involved, whereas before, I had encountered people who seemed more abrasive.

Now I'm worried about getting burned out. I check the EA Forum way too often for my own good, and I've been thinking obsessively about cause prioritization and longtermism. I talk about my current uncertainties in this post.

A commenter on this thread said it should have been a top-level post rather than a QT. Throwing in my vote for this feature.

Questioning the new "EA is funding constrained" narrative

I recently saw a presentation with a diagram showing how committed EA funding dropped by almost half with the collapse of FTX, based on these data compiled by 80k in 2022. Open Phil at the time had a $22.5 billion endowment and FTX's founders were collectively worth $16.5 billion.

I think that this narrative gives off the impression that EA causes (especially global health and development) are more funding-constrained than they really are. 80k's data excludes philanthropists that often make donations in an EA-aligned way, such as the Gates Foundation's global health program. As of 2022, Gates' endowment was worth $67.3 billion,[1] and global health and development accounted for more than half of its spending that year.[2] The Gates Foundation's global health program, at least, seems to make grants in a cost-effectiveness-driven way, so it arguably should count as EA funding.

  1. ^
  2. ^

Great start, I'm looking forward to seeing how this software develops!

I noticed that the model estimates of cost-effectiveness for GHD/animal welfare and x-risk interventions are not directly comparable. Whereas the x-risk interventions are modeled as a stream of benefits that could be realized over the next 1,000 years (barring extinction), the distribution of cost-effectiveness for a GHD or animal welfare is taken as given. Indeed:

For interventions in global health and development we don't model impact internally, but instead stipulate the range of possible values. This intervention is assumed to cost between <lower bound> and <upper bound> per DALY averted.

So I'd be keen to see more granular modeling of the benefits of these interventions, especially over longer time scales. For example, cash transfers have not only immediate benefits to their recipients, but also a multiplier effect on the economy: according to this 80k episode, $1 in cash transfers produces $2.50 in additional economic output for the community. This could ultimately put the economy on a higher growth path. What effect would a $1000 donation to GiveDirectly or AMF have over the next 20-50 years?

Similarly, for an animal welfare intervention such as a corporate cage-free campaign, the long-term effects would depend on how long the cage-free policy is expected to last, how well it's enforced, etc. This would be undoubtedly complicated to model, but would help make these interventions easier to compare with traditionally "longtermist" interventions.

There's now a Netflix adaptation of it! And the ending reminded me of FTX 😜

Anyone can create a linkpost for an 80k episode. Though it might be extra convenient to have a way to automatically create a linkpost with a pre-filled summary of the linked page and a top-level comment with your thoughts.

Content warning: Israel/Palestine

Has there been research on what interventions are effective at facilitating dialogue between social groups in conflict?

I remember an article about how during the last Israel-Gaza flare-up, Israelis and Palestinians were using the audio chatroom app Clubhouse to share their experiences and perspectives. This was portrayed as a phenomenon that increased dialogue and empathy between the two groups. But how effective was it? Could it generalize to other ethnic/religious conflicts around the world?

Load more