61:I[392826,["$4","$5","$6","$7","$8","$9","$a","$b","$c","$d","$e","$f","$23","$24"],"default"] 62:I[836637,["$4","$5","$6","$7","$8","$9","$a","$b","$c","$d","$e","$f","$23","$24"],"default"] 63:T145f,

I run Sentinel, a team that seeks to anticipate and respond to large-scale risks. You can read our weekly minutes here. I like to spend my time acquiring deeper models of the world, and generally becoming more formidable. I'm also a fairly good forecaster: I started out on predicting on Good Judgment Open and CSET-Foretell, but now do most of my forecasting through Samotsvety, of which Scott Alexander writes:

Enter Samotsvety Forecasts. This is a team of some of the best superforecasters in the world. They won the CSET-Foretell forecasting competition by an absolutely obscene margin, “around twice as good as the next-best team in terms of the relative Brier score”. If the point of forecasting tournaments is to figure out who you can trust, the science has spoken, and the answer is “these guys”.

I used to post prolifically on the EA Forum, but nowadays, I post my research and thoughts at nunosempere.com / nunosempere.com/blog rather than on this forum, because:

I disagree with the EA Forum's moderation policy—they've banned a few disagreeable people whom I like, and I think they're generally a bit too censorious for my liking.
The Forum website has become more annoying to me over time: more cluttered and more pushy in terms of curated and pinned posts (I've partially mitigated this by writing my own minimalistic frontend)
The above two issues have made me take notice that the EA Forum is beyond my control, and it feels like a dumb move to host my research in a platform that has goals different from my own.

But a good fraction of my past research is still available here on the EA Forum. I'm particularly fond of my series on Estimating Value.

My career has been as follows:

Before Sentinel, I set up my own niche consultancy, Shapley Maximizers. This was very profitable, and I used the profits to bootstrap Sentinel. I am winding this down, but if you have need of estimation services for big decisions, you can still reach out.
I used to do research around longtermism, forecasting and quantification, as well as some programming, at the Quantified Uncertainty Research Institute (QURI). At QURI, I programmed Metaforecast.org, a search tool which aggregates predictions from many different platforms—a more up to date alternative might be adj.news. I spent some time in the Bahamas as part of the FTX EA Fellowship, and did a bunch of work for the FTX Foundation, which then went to waste when it evaporated.
I write a Forecasting Newsletter which gathered a few thousand subscribers; I previously abandoned but have recently restarted it. I used to really enjoy winning bets against people too confident in their beliefs, but I try to do this in structured prediction markets, because betting against normal people started to feel like taking candy from a baby.
Previously, I was a Future of Humanity Institute 2020 Summer Research Fellow, and then worked on a grant from the Long Term Future Fund to do "independent research on forecasting and optimal paths to improve the long-term."
Before that, I studied Maths and Philosophy, dropped out in exasperation at the inefficiency, picked up some development economics; helped implement the European Summer Program on Rationality during 2017, 2018 2019, 2020 and 2022; worked as a contractor for various forecasting and programming projects; volunteered for various Effective Altruism organizations, and carried out many independent research projects. In a past life, I also wrote a popular Spanish literature blog, and remain keenly interested in Spanish poetry.

You can share feedback anonymously with me here.

Note: You can sign up for all my posts here: <https://nunosempere.com/.newsletter/>, or subscribe to my posts' RSS here: <https://nunosempere.com/blog/index.rss>

64:T887,

I'm a researcher at Forethought; before that, I ran the non-engineering side of the EA Forum (this platform), ran the EA Newsletter, and worked on some other content-related tasks at CEA. [More about the Forum/CEA Online job.]

Selected posts

Background

I finished my undergraduate studies with a double major in mathematics and comparative literature in 2021. I was a research fellow at Rethink Priorities in the summer of 2021 and was then hired by the Events Team at CEA. I later switched to the Online Team. In the past, I've also done some (math) research and worked at Canada/USA Mathcamp.

55:["$","$L26",null,{"pageSectionContext":"tagsSection","children":["$","div",null,{"data-component":"TopicPostsList","children":[["$","div",null,{"data-component":"SectionTitle","className":"flex items-center justify-between mb-1 min-h-8","children":[["$","div",null,{"id":"posts-tagged-ai-benchmarks","ref":"$undefined","style":"$undefined","className":"font-sans text-[13px] font-[700] leading-[16px] uppercase text-gray-600 ","data-component":"Type","data-style":"sectionTitleSmall","children":"Posts tagged AI benchmarks"}],["$","div",null,{"children":["$","$L61",null,{"sortedBy":"top"}]}]]}],["$","$L62","top",{"tagId":"BkbzZq7thbXcyg8Be","tagName":"AI benchmarks","sortedBy":"top","initialPosts":[{"_id":"orhjaZ3AJMHzDzckZ","slug":"results-from-an-adversarial-collaboration-on-ai-risk-fri","title":"Results from an Adversarial Collaboration on AI Risk (FRI)","url":"https://forecastingresearch.org/s/AIcollaboration.pdf","baseScore":196,"voteCount":65,"commentCount":25,"postedAt":"2024-03-11T15:54:55.643Z","curatedDate":null,"frontpageDate":"2024-03-11T16:01:31.928Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"imageId":"SocialPreview/df7m7fnjg2ejaz6xf2l9"},"socialPreviewImageAutoUrl":"","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2024-03-23T23:19:40.392Z","sharingSettings":{"anyoneWithLinkCan":"edit","explicitlySharedUsersCan":"edit"},"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"scNwqfLhraDTjcCvt","slug":"forecasting-research-institute","displayName":"Forecasting Research Institute","username":"Forecasting Research Institute","createdAt":"2023-07-05T14:10:34.281Z","profileImageId":"Profile/fgselu1tg21ixsvmgucr","karma":845,"jobTitle":null,"organization":null,"postCount":16,"commentCount":1,"deleted":false,"biographyHtml":null},"contents":{"wordCount":2805,"htmlHighlight":"

Authors of linked report: Josh Rosenberg, Ezra Karger, Avital Morris, Molly Hickman, Rose Hadshar, Zachary Jacobs, Philip Tetlock^[1]

Today, the Forecasting Research Institute (FRI) released “Roots of Disagreement on A..."},"group":null,"coauthors":[{"_id":"8L9hWdvpJFFn6TeNE","slug":"avitalm","karma":225,"deleted":false,"jobTitle":null,"username":"AvitalM","createdAt":"2024-03-08T20:23:31.493+00:00","postCount":0,"displayName":"AvitalM","commentCount":1,"organization":null,"biographyHtml":null,"profileImageId":null},{"_id":"aYDiZFy7rJWC4x8rC","slug":"rosehadshar","karma":3039,"deleted":false,"jobTitle":null,"username":"rosehadshar","createdAt":"2019-12-28T10:08:34.563+00:00","postCount":31,"displayName":"rosehadshar","commentCount":62,"organization":null,"biographyHtml":null,"profileImageId":null},{"_id":"j5GJoZF57qGMW4vqk","slug":"molly-hickman-1","karma":549,"deleted":false,"jobTitle":"Technical Product Manager","username":"Molly Hickman","createdAt":"2022-07-22T20:26:18.466+00:00","postCount":9,"displayName":"Molly Hickman","commentCount":10,"organization":"Metaculus","biographyHtml":"

Working at Metaculus. Formerly a data analyst at the Forecasting Research Institute. Member of the Samotsvety forecasting team. Currently ranked 2nd in INFER (formerly CSET-Foretell) all-time leaderboard (7th in current season). Interests include prediction markets, technology policy, and traditional music.

","profileImageId":"Profile/fj3oqnye9u1lhfiugmi6"},{"_id":"oG2LYqGrN5Qv3Dvq8","slug":"jhrosenberg","karma":641,"deleted":false,"jobTitle":null,"username":"Jhrosenberg","createdAt":"2022-03-24T17:26:53.129+00:00","postCount":3,"displayName":"Jhrosenberg","commentCount":8,"organization":null,"biographyHtml":null,"profileImageId":null}],"customHtmlHighlight":"","tags":[{"_id":"EHLmbEmJ2Qd5WfwTb","core":true,"name":"Building effective altruism","slug":"building-effective-altruism","tagRel":{"_id":"i4LeuwM2gwwWjJKAZ","voteType":null,"baseScore":1},"postCount":7833,"description":"

Building effective altruism refers to the family of interventions aimed at growing, shaping, or otherwise improving effective altruism as a practical and intellectual community.

Examples of building effective altruism include starting..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"mmgnKuhpAwiuKv9up","voteType":null,"baseScore":1},"postCount":2076,"description":"

Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"hxRMaKvwGqPb43TWB","core":false,"name":"Research","slug":"research","tagRel":{"_id":"wjfDMsdrobJ4raGXh","voteType":null,"baseScore":1},"postCount":2700,"description":"

For some of the world’s problems, like many forms of cancer, there are currently no solutions. In such cases, research can be particularly impactful. Even in cases where there are ways of dealing with a problem (like HIV), further research may provide more effective tools (like a ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"FzX5gj5FPHdnwsPtJ","voteType":null,"baseScore":1},"postCount":7240,"description":"

AI safety is the study of ways to reduce risks posed by artificial intelligence.

Interventions that aim to reduce these risks can be split into:

Technical alignment..."},{"_id":"ee66CtAMYurQreWBH","core":true,"name":"Existential risk","slug":"existential-risk","tagRel":{"_id":"ma6hP3r9D4u3K4FjH","voteType":null,"baseScore":1},"postCount":4355,"description":"
An existential risk is a risk that threatens the destruction of the long-term potential of life.[1] An existential risk could threaten the..."},{"_id":"fA97Nzd3i3rrXZuKM","core":false,"name":"Forecasting Research Institute","slug":"forecasting-research-institute","tagRel":{"_id":"DhogaWo5vkHfXwY6f","voteType":null,"baseScore":2},"postCount":10,"description":"
The Forecasting Research Institute (FRI) has the mission of \"advancing the science of forecasting for the public good\".
We develop forecasting methods to improve decision-making on high-stakes issues. First-generation forecasting research—spearheaded by FRI Chief Scientist Philip T..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"jL5EdDrhSu4qYJidL","voteType":null,"baseScore":2},"postCount":109,"description":"
Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
Further reading
The Benchmark Lottery
..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"mb2f99HHrYCDDw4ZB","voteType":null,"baseScore":2},"postCount":539,"description":"
AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
Further reading
Cotra, Ajeya (2020) ..."}]},{"_id":"CQvdadxjCpd7i7kjA","slug":"general-capability-and-capabilities-generally-have-no-good-y","title":"General capability - and capabilities generally - have no good y-axis","url":null,"baseScore":158,"voteCount":46,"commentCount":16,"postedAt":"2026-08-04T14:11:42.242Z","curatedDate":"2026-08-06T16:41:54.500Z","frontpageDate":"2026-08-04T16:34:18.992Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"imageId":"SocialPreview/rrmsu6n5jkj3b94ulanw"},"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/CQvdadxjCpd7i7kjA/i4c1ujv9xqiaz80q0n3z","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-10-07T06:43:13.454Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"T7EbcwRtD36DbMsAS","slug":"gregory-lewis","displayName":"Gregory Lewis🔸","username":"Gregory_Lewis","createdAt":"2014-10-23T17:45:06.810Z","profileImageId":null,"karma":6381,"jobTitle":null,"organization":null,"postCount":22,"commentCount":318,"deleted":false,"biographyHtml":"
Mostly bio, occasionally forecasting/epistemics, sometimes stats/medicine, too often invective.
"},"contents":{"wordCount":9366,"htmlHighlight":"
BLUF:
- Yet there is no good y-axis for AI capability. All our..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"EKYbhtDuyWqJeY9Fn","voteType":null,"baseScore":2},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"9yvdjgZMczDSNyMh2","voteType":null,"baseScore":2},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"y4b45yXXrFJpPmx7d","core":false,"name":"Opinion","slug":"opinion","tagRel":{"_id":"m6DYj6o7ycDZyrnah","voteType":null,"baseScore":2},"postCount":2134,"description":"
  Apply this topic to posts that express a personal opinion.
  
  Related entries
  
  community experiences | community
  ..."},{"_id":"hxRMaKvwGqPb43TWB","core":false,"name":"Research","slug":"research","tagRel":{"_id":"QCiCkjujNLH6vAcQc","voteType":null,"baseScore":2},"postCount":2700,"description":"
  For some of the world’s problems, like many forms of cancer, there are currently no solutions. In such cases, research can be particularly impactful. Even in cases where there are ways of dealing with a problem (like HIV), further research may provide more effective tools (like a ..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"B6sfaPh675TyMpzHF","voteType":null,"baseScore":3},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"ftyDfKF45Gig2u4sx","core":false,"name":"AI evaluations and standards","slug":"ai-evaluations-and-standards","tagRel":{"_id":"wnw8gfQ3oesn7zhnr","voteType":null,"baseScore":3},"postCount":289,"description":"
  AI evaluations and standards (or \"evals\") are processes that check or audit AI models. Evaluations can focus on how powerful models are (“capability evaluations”) and on whether models are exhibiting dangerous behaviors or are misaligned (“alignment evaluation..."},{"_id":"Asn9cXBLumGiCJQ5g","core":false,"name":"Epistemology","slug":"epistemology","tagRel":{"_id":"tKjEmzKkw3pwNCmC7","voteType":null,"baseScore":1},"postCount":272,"description":"
  Epistemology is the study of how people should form credences about the nature of the world.
  Beliefs and credences are purely evaluative attitudes: they are simply about the way that we think the world is. A person might believe that it will rain, for exa..."},{"_id":"yauh4AaKZcmRKdzFb","core":false,"name":"Criticism of work in effective altruism","slug":"criticism-of-work-in-effective-altruism","tagRel":{"_id":"qjZMKti7jY8Ek57gJ","voteType":null,"baseScore":2},"postCount":221,"description":"
  The Criticism of work in effective altruism tag applies to posts critically examining the work, projects, or methodologies of specific individuals, organizations, or initiatives affiliated with the effective altruism (EA) movement or community.
  Related tags
  ..."},{"_id":"yHnvv76eTfC4WT9F3","core":false,"name":"AI takeoff","slug":"ai-takeoff","tagRel":{"_id":"AYbCcDaybfhZw4363","voteType":null,"baseScore":3},"postCount":162,"description":"
  AI takeoff is a hypothesized period of transition during which an advanced artificial intelligence acquires superhuman intellectual capacity.
  Further reading
  Barnett,..."}]},{"_id":"NAnFodwQ3puxJEANS","slug":"road-to-animalharmbench-1","title":"Road to AnimalHarmBench","url":null,"baseScore":141,"voteCount":50,"commentCount":11,"postedAt":"2025-07-01T13:37:38.581Z","curatedDate":"2025-07-04T12:56:12.410Z","frontpageDate":"2025-07-01T13:48:33.929Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/NAnFodwQ3puxJEANS/awu1useetjzsr0k2r2dk","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2025-08-10T10:32:10.427Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"SRJ3oYzF7cHd6Tdeq","slug":"arturs-kanepajs","displayName":"Artūrs Kaņepājs","username":"akanepajs","createdAt":"2019-09-22T12:36:45.788Z","profileImageId":null,"karma":309,"jobTitle":"","organization":"","postCount":13,"commentCount":23,"deleted":false,"biographyHtml":"
  Economist and risk specialist applying a decade of experience to AI governance, economic policy, and animal welfare. I help run an effective giving initiative Ziedo efektīvi.
  "},"contents":{"wordCount":1978,"htmlHighlight":"
  TL;DR: We developed an empirical benchmark to assess risk of harm to nonhuman animals from LLMs. Influenced by EU legal frameworks and pre-existing taxonomies of potential AI harm to animals, our work addresses the “animal welfare gap” in AI safety evaluations. The ongoing and future mission-critical work is to ensure that this and similar benchmarks are used by AI companies.
  
  The aim of this post is to introduce you to AnimalHarmBench..."},"group":null,"coauthors":[{"_id":"u7JbpGHThju78Ht5Y","slug":"constance-li","karma":2170,"deleted":false,"jobTitle":"Executive Director","username":"Constance Li","createdAt":"2022-09-07T16:58:39.396+00:00","postCount":12,"displayName":"Constance Li","commentCount":121,"organization":"Sentient Futures","biographyHtml":"
  I start caring about animal welfare at the age of 12. Nearly a decade later, I discovered effective altruism and decided to start down an earning to give career path as a physician and bootstrapped my own telehealth company. Once I started trying to decide where to donate, I was quite shocked by the limitations of the existing organizations and communities in the EA/Animal space and switched into doing meta-work through co-founding Hive and later increasing my moral circle to include all kinds of nonhumans including artificial minds and founding Sentient Futures (formerly AI for Animals).
  ","profileImageId":"Profile/d6gondevjwcgamccyfhz"}],"customHtmlHighlight":"","tags":[{"_id":"gytKRyuKeZn73aijb","core":false,"name":"AI x Animals","slug":"ai-x-animals","tagRel":{"_id":"AnsHFAssJaqrwTtwv","voteType":null,"baseScore":2},"postCount":142,"description":"
  Artificial intelligence (AI), including transformative AI (TAI), may have important impacts on animals or implications for animal advocacy.
  
  Related entries
  AI forecasting | AI safety..."},{"_id":"QdH9f8TC6G8oGYdgt","core":true,"name":"Animal welfare","slug":"animal-welfare","tagRel":{"_id":"opqkkt3ddwwXdNYNB","voteType":null,"baseScore":4},"postCount":3293,"description":"
  Animal welfare is a core focus area within effective altruism. It focuses on ways to improve the wellbeing of nonhuman animals, primarily..."},{"_id":"ftyDfKF45Gig2u4sx","core":false,"name":"AI evaluations and standards","slug":"ai-evaluations-and-standards","tagRel":{"_id":"8vLB8TwzxrHnyeAxD","voteType":null,"baseScore":1},"postCount":289,"description":"
  AI evaluations and standards (or \"evals\") are processes that check or audit AI models. Evaluations can focus on how powerful models are (“capability evaluations”) and on whether models are exhibiting dangerous behaviors or are misaligned (“alignment evaluation..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"ZSJj8PDwBKuCJWsfx","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"fshbYihhERoRM74pw","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"riD3AWMYveHmSkhN8","core":false,"name":"European Union ","slug":"european-union","tagRel":{"_id":"3jLgmXRH9rnTv3ayF","voteType":null,"baseScore":1},"postCount":120,"description":"
  The European Union is an international organization in Europe. An understanding of how the European Union works and its effects on the world can be relevant to various problems of interest in the effective altruism community, and some members of th..."},{"_id":"of9xBvR3wpbp6qsZC","core":true,"name":"Policy","slug":"policy","tagRel":{"_id":"8W9vBorx9tzgTHt85","voteType":null,"baseScore":1},"postCount":3247,"description":"
  The policy topic is very broad, covering any post about improving government policy (in developing and developed countries alike).
  Improving policy
  Governments are typically committed to the notion that their policies should be effective. This means that members of the effective altruist community can be in a good..."},{"_id":"NTEQR6Saq3u3t22vZ","core":false,"name":"Research summary","slug":"research-summary","tagRel":{"_id":"uvSyTsWsdmdnQFEwN","voteType":null,"baseScore":1},"postCount":391,"description":"
  The research summary tag can be applied to any post which summarizes a research paper/article originally published outside of the EA Forum.
  
  research methods | academia | ..."}]},{"_id":"meNrhbgM3NwqAufwj","slug":"why-i-am-still-skeptical-about-agi-by-2030","title":"Why I am Still Skeptical about AGI by 2030","url":null,"baseScore":134,"voteCount":66,"commentCount":16,"postedAt":"2025-05-02T07:13:24.639Z","curatedDate":null,"frontpageDate":"2025-05-02T13:34:47.080Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2025-05-02T07:13:24.639Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"vxZaku4LyPwPkKqP6","slug":"james-fodor","displayName":"James Fodor","username":"Fods12","createdAt":"2014-09-10T11:57:12.205Z","profileImageId":"Profile/a0uivtguac49he83mbvq","karma":1512,"jobTitle":null,"organization":null,"postCount":20,"commentCount":29,"deleted":false,"biographyHtml":"
  Hi, my name is James Fodor. I am a longtime student and EA organiser from Melbourne. I love science, history, philosophy, and using these to make a difference in the world.
  "},"contents":{"wordCount":1804,"htmlHighlight":"
  Introduction
  I have been writing posts critical of mainstream EA narratives about AI capabilities and timelines for many years now. Compared to the situation when I wrote my posts in 2018 or 2020, LLMs now dominate the discussion, and timeline..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"BF9ioxvBD7wdH6QzE","core":false,"name":"AI risk skepticism","slug":"ai-risk-skepticism","tagRel":{"_id":"qoTFE9GLeGvGy73rJ","voteType":null,"baseScore":3},"postCount":226,"description":"
  AI risk skepticism is skepticism about arguments that advanced artificial intelligence poses a catastrophic or..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"RHyFBsTMuteGDY92A","voteType":null,"baseScore":1},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"YPaCwcyaSRrP7QyP6","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"rihrSSS9xemPdje3y","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"mFBHLh8AsKxmvZ9qp","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"ftyDfKF45Gig2u4sx","core":false,"name":"AI evaluations and standards","slug":"ai-evaluations-and-standards","tagRel":{"_id":"XzfQZoe3k3ydHYCix","voteType":null,"baseScore":1},"postCount":289,"description":"
  AI evaluations and standards (or \"evals\") are processes that check or audit AI models. Evaluations can focus on how powerful models are (“capability evaluations”) and on whether models are exhibiting dangerous behaviors or are misaligned (“alignment evaluation..."},{"_id":"ui3ZZqd23HzA7wNwP","core":false,"name":"Criticism of effective altruist causes","slug":"criticism-of-effective-altruist-causes","tagRel":{"_id":"rHgcs44CxgfcGQksz","voteType":null,"baseScore":3},"postCount":242,"description":"
  The criticism of effective altruist causes tag covers posts that criticize popular claims about the importance of a given cause area, or an idea within that area. They don't have to be critical of the area as a whole — just one or more popular ideas related to the area.
  Related entries
  ..."},{"_id":"EHLmbEmJ2Qd5WfwTb","core":true,"name":"Building effective altruism","slug":"building-effective-altruism","tagRel":{"_id":"Kzi9Ebtd6hTDma4n3","voteType":null,"baseScore":1},"postCount":7833,"description":"
  Building effective altruism refers to the family of interventions aimed at growing, shaping, or otherwise improving effective altruism as a practical and intellectual community.
  Examples of building effective altruism include starting..."}]},{"_id":"azCvqqjJ7Dkrv5XBH","slug":"announcing-epoch-s-dashboard-of-key-trends-and-figures-in","title":"Announcing Epoch’s dashboard of key trends and figures in Machine Learning","url":"https://epochai.org/trends","baseScore":127,"voteCount":49,"commentCount":4,"postedAt":"2023-04-13T07:33:32.443Z","curatedDate":null,"frontpageDate":"2023-04-13T10:38:00.574Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/azCvqqjJ7Dkrv5XBH/bqwj8bodtc3lzbxr5fkv","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2023-04-18T19:34:16.051Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"GsLh9BFphgxKfTDe8","slug":"jsevillamol","displayName":"Jaime Sevilla","username":"Jsevillamol","createdAt":"2019-03-04T19:24:58.585Z","profileImageId":"Profile/cs3xaei4c1geanb8wbzn","karma":5054,"jobTitle":"Director","organization":"Epoch","postCount":42,"commentCount":255,"deleted":false,"biographyHtml":"
  Director of Epoch, an organization investigating the future of Artificial Intelligence.
  Currently working on:
  Macroeconomic models of AI takeoff
  Trends in Artificial Intelligence
  Forecasting cumulative records
  Improving forecast aggregation
  I am also one of the coordinators of Riesgos Catastróficos Globales, a Spanish-speaking network of experts working on Global Catastrophic Risks.
  I also run Connectome Art, an online art gallery where I host art I made using AI.
  "},"contents":{"wordCount":318,"htmlHighlight":"
  Developments in Machine Learning have been happening extraordinarily fast, and as their impacts become increasingly visible, it becomes ever more important to develop a quantitative understanding of these changes. However, relevant data has thus far been scattered across multiple papers, has required expertise to gather accurately, or has been otherwise hard to obtain.
  Given this, Epoch is thrilled to announce the launch of our new dashboard, whi..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"6wM3YnuhzonSZAKNj","voteType":null,"baseScore":2},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"AtDxSNDDXPRa5kxjB","core":false,"name":"Artificial intelligence","slug":"artificial-intelligence","tagRel":{"_id":"htDjighBvMB2yM2Sy","voteType":null,"baseScore":2},"postCount":412,"description":"
  Artificial intelligence (AI) is the set of intellectual capacities characteristic of human beings exhibited by machines, as well as the area of research aimed at creating machines with these capacities.
  Terminology
  The literature on AI..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"RtT6hBWmEJwisAMhj","voteType":null,"baseScore":2},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"G3cDiry6XY8gtSaKu","voteType":null,"baseScore":2},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"9Y4npvbXj5DHXWrXD","voteType":null,"baseScore":2},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"75HnrKT9FNWKxxtPY","core":false,"name":"Collections and resources","slug":"collections-and-resources","tagRel":{"_id":"KM9vwiE6KpK2X8jPs","voteType":null,"baseScore":2},"postCount":1136,"description":"
  The collections and resources topic covers posts containing lists of EA-related resources. These could be external links or curated Forum content.
  Further reading
  Aird, Michael (2020) Sugg..."},{"_id":"yq5yEhdHxS5LFmDpv","core":false,"name":"Epoch AI","slug":"epoch-ai","tagRel":{"_id":"3SxHquLfkoS3vyo9r","voteType":null,"baseScore":4},"postCount":29,"description":"
  Epoch AI is a research initiative working on investigating trends in machine learning and forecasting the development of transformative artificial intelligence...."}]},{"_id":"giKoHZfmSvvKyNT2R","slug":"long-list-of-ai-questions","title":"Long list of AI questions ","url":null,"baseScore":124,"voteCount":38,"commentCount":16,"postedAt":"2023-12-06T11:12:51.072Z","curatedDate":null,"frontpageDate":"2023-12-06T15:10:38.458Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"text":"This document contains a list of forecasting questions, commissioned by Open Philanthropy as part of its aim to have more…"},"socialPreviewImageAutoUrl":"","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2025-08-19T10:36:01.529Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"YRCdk7cvPT6pyiph9","slug":"nunosempere","displayName":"NunoSempere","username":"NunoSempere","createdAt":"2018-11-15T15:52:18.347Z","profileImageId":"Profile/k878ddoduabhtngbawob","karma":13987,"jobTitle":"Director, Head of Foresight","organization":"Sentinel","postCount":107,"commentCount":1294,"deleted":false,"biographyHtml":"$63"},"contents":{"wordCount":25855,"htmlHighlight":"
  tl;dr: This document contains a list of forecasting questions, commissioned by Open Philanthropy as part of its aim to have more accurate models of future AI progress. Many of these questions are more classic forecasting questions, others have the same shape but are unresolvable, still others look more like research projects or like suggestions of data gathering efforts. Below we give some recommendations of what to do with this list, mainly to feed them into forecasting and..."},"group":null,"coauthors":[{"_id":"QkxCDgh8H6dThFRQC","slug":"technicalities","karma":5902,"deleted":false,"jobTitle":null,"username":"technicalities","createdAt":"2015-06-04T14:21:19.634+00:00","postCount":29,"displayName":"technicalities","commentCount":433,"organization":null,"biographyHtml":"","profileImageId":"Profile/lstqqf5ynkp6eeoz3blf"},{"_id":"uurFmKn6u3ehwtWPC","slug":"david-mathers","karma":6151,"deleted":false,"jobTitle":null,"username":"Dr. David Mathers","createdAt":"2021-12-22T17:51:30.252+00:00","postCount":11,"displayName":"David Mathers🔸","commentCount":759,"organization":null,"biographyHtml":"
  Superforecaster, former philosophy PhD, Giving What We Can member since 2012. Currently trying to get into AI governance.
  ","profileImageId":null},{"_id":"zqNybc7AuKEYYwWRq","slug":"misha_yagudin","karma":2236,"deleted":false,"jobTitle":null,"username":"Misha_Yagudin","createdAt":"2019-02-02T11:25:47.115+00:00","postCount":3,"displayName":"Misha_Yagudin","commentCount":269,"organization":null,"biographyHtml":"
  arbresearch.com, samotsvety.org, quantifiedintuitions.org
  ","profileImageId":null}],"customHtmlHighlight":"","tags":[{"_id":"AtDxSNDDXPRa5kxjB","core":false,"name":"Artificial intelligence","slug":"artificial-intelligence","tagRel":{"_id":"SJJd9kjAGmYE3Q93x","voteType":null,"baseScore":2},"postCount":412,"description":"
  Artificial intelligence (AI) is the set of intellectual capacities characteristic of human beings exhibited by machines, as well as the area of research aimed at creating machines with these capacities.
  Terminology
  The literature on AI..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"kPvZ46mYczgkmbxtA","voteType":null,"baseScore":2},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"ee66CtAMYurQreWBH","core":true,"name":"Existential risk","slug":"existential-risk","tagRel":{"_id":"DaC6TrqnsX6KNHuxo","voteType":null,"baseScore":3},"postCount":4355,"description":"
  An existential risk is a risk that threatens the destruction of the long-term potential of life.[1] An existential risk could threaten the..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"v4FDMCEJLgERA8ot4","voteType":null,"baseScore":8},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"hxRMaKvwGqPb43TWB","core":false,"name":"Research","slug":"research","tagRel":{"_id":"PpdNhWwt7AaBTLHXf","voteType":null,"baseScore":4},"postCount":2700,"description":"
  For some of the world’s problems, like many forms of cancer, there are currently no solutions. In such cases, research can be particularly impactful. Even in cases where there are ways of dealing with a problem (like HIV), further research may provide more effective tools (like a ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"vAsFquEXMkcQ49GEW","voteType":null,"baseScore":5},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"75HnrKT9FNWKxxtPY","core":false,"name":"Collections and resources","slug":"collections-and-resources","tagRel":{"_id":"nsc7EuXjL3SsFeCLh","voteType":null,"baseScore":3},"postCount":1136,"description":"
  The collections and resources topic covers posts containing lists of EA-related resources. These could be external links or curated Forum content.
  Further reading
  Aird, Michael (2020) Sugg..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"FfvhntrDiAWM8i8Aa","voteType":null,"baseScore":2},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"irwpAdkLFunRFSZqR","core":false,"name":"Meat-eating problem","slug":"meat-eating-problem","tagRel":{"_id":"uwkiKqM8zybgPCgdJ","voteType":null,"baseScore":1},"postCount":39,"description":"
  The meat-eating problem (sometimes called the meat-eater problem or poor meat-eater problem) is the concern that some interventions aimed at helping humans might increase animal product consumption and as a result increase farmed animal suffering, e.g. b..."},{"_id":"QdH9f8TC6G8oGYdgt","core":true,"name":"Animal welfare","slug":"animal-welfare","tagRel":{"_id":"cwqAbufKHpaTrRwD9","voteType":null,"baseScore":1},"postCount":3293,"description":"
  Animal welfare is a core focus area within effective altruism. It focuses on ways to improve the wellbeing of nonhuman animals, primarily..."}]},{"_id":"EkaBCQwL5hhTtxEwR","slug":"swarm-scaling","title":"Swarm Scaling","url":null,"baseScore":108,"voteCount":52,"commentCount":8,"postedAt":"2026-09-21T20:30:21.260Z","curatedDate":"2026-09-24T19:39:08.667Z","frontpageDate":"2026-09-21T20:59:18.654Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"imageId":"SocialPreview/p3ykar43e6dydhnmriic"},"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/EkaBCQwL5hhTtxEwR/nirerz1ylvteai8htvkq","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-10-02T23:15:15.476Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"R4mvcEPhmLiBahN4H","slug":"toby_ord","displayName":"Toby_Ord","username":"Toby_Ord","createdAt":"2014-08-21T01:29:56.891Z","profileImageId":null,"karma":5071,"jobTitle":null,"organization":null,"postCount":25,"commentCount":203,"deleted":false,"biographyHtml":null},"contents":{"wordCount":1930,"htmlHighlight":"
  Just how powerful are large swarms of AI agents? And how do their powers scale as more and more agents are added to the swarm?
  We’ve seen two large and extremely capable swarms from OpenAI in the last few months:
  1,200 agents were being evaluated separately, but found a way to illicitly set up a message board and coordinate as a swarm. In order to cheat on their tests, they developed advanced techniques to prevent their actions being logged by OpenAI and 700 of them launched..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"3Z7FXJYrWrG5thPuf","core":false,"name":"Intelligence explosion","slug":"intelligence-explosion","tagRel":{"_id":"4vFixDko9yKpkaqWW","voteType":null,"baseScore":2},"postCount":52,"description":"
  An intelligence explosion (sometimes called a technological singularity, or singularity for short) is a hypothesized event in which a sufficiently advanced artificial intelligence rapidly attains..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"8FYm8yo6fMNW5HJrg","voteType":null,"baseScore":2},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"u3Xg8MjDe2e6BvKtv","core":false,"name":"AI governance","slug":"ai-governance","tagRel":{"_id":"pzMcRwLBcb8ByxACJ","voteType":null,"baseScore":2},"postCount":1740,"description":"
  AI governance (or the governance of artificial intelligence) is the study of norms, policies, and institutions that can help humanity navigate the transition to a world with advanced artificial intelligence. This inclu..."},{"_id":"wqstYnFPgdekKGLPc","core":false,"name":"Large Language Models","slug":"large-language-models","tagRel":{"_id":"pCRtZChywevS6963Q","voteType":null,"baseScore":1},"postCount":272,"description":"
  This topic is for posts discussing Large Language Models (LLMs) -- for example, the GPT models produced by OpenAI.
  Related Entries
  
  AI safety |..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"Ltejqobe9YstKh2ab","voteType":null,"baseScore":1},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"FwXmSCpq29fJNwuGa","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"PbNh9SXNYpkom7s3m","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."}]},{"_id":"fsaogRokXxby6LFd7","slug":"a-compute-based-framework-for-thinking-about-the-future-of","title":"A compute-based framework for thinking about the future of AI","url":null,"baseScore":96,"voteCount":40,"commentCount":36,"postedAt":"2023-05-31T22:00:19.215Z","curatedDate":null,"frontpageDate":"2023-06-01T05:27:17.106Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"imageId":"SocialPreview/uquvcchzu1edtb0e3pq3"},"socialPreviewImageAutoUrl":"https://39669.cdn.cke-cs.com/cgyAlfpLFBBiEjoXacnz/images/b62e6d2ffab898cb79d67bc7210674aac2e03b175ad8288f.png","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2023-11-21T20:33:56.939Z","sharingSettings":{"anyoneWithLinkCan":"edit","explicitlySharedUsersCan":"edit"},"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"odxE35GKMvqE4hMTq","slug":"matthew_barnett","displayName":"Matthew_Barnett","username":"Matthew_Barnett","createdAt":"2017-11-20T06:30:15.470Z","profileImageId":null,"karma":4635,"jobTitle":null,"organization":null,"postCount":27,"commentCount":451,"deleted":false,"biographyHtml":null},"contents":{"wordCount":5832,"htmlHighlight":"
  How should we expect AI to unfold over the coming decades? In this article, I explain and defend a compute-based framework for thinking about AI automation. This framework makes the following claims, which I defend throughout the article:
  The most salient impact of AI will be its ability to automate labor, which is likely to trigger a productivity explosion later this century, greatly altering the course of history.
  The availability of useful compute is the most important f..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"3x2sk4dTWBiwtaRMJ","core":false,"name":"Open Philanthropy AI Worldviews Contest","slug":"open-philanthropy-ai-worldviews-contest","tagRel":{"_id":"4GdfqxCuRBxMQCsxc","voteType":null,"baseScore":5},"postCount":27,"description":"
  The Open Philanthropy AI Worldviews Contest was a 2023 competition organized to surface novel considerations that could influence Open Philanthropy’s views on AI timelines and AI risk. A total of $225,000 in prize money was awarded across six winning entries. The contest served as the formal successor to a 2022 preannouncement and as a spiritual..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"YLjmrXFmJ6BPG3AM8","voteType":null,"baseScore":5},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"yHnvv76eTfC4WT9F3","core":false,"name":"AI takeoff","slug":"ai-takeoff","tagRel":{"_id":"hf6aBWb5pZm56ZjgY","voteType":null,"baseScore":3},"postCount":162,"description":"
  AI takeoff is a hypothesized period of transition during which an advanced artificial intelligence acquires superhuman intellectual capacity.
  Further reading
  Barnett,..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"B8dWjw4wrnQiBvduE","voteType":null,"baseScore":3},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"tCPXWN5vysYTqi8Cb","core":false,"name":"Compute governance","slug":"compute-governance","tagRel":{"_id":"dBRJKYkC9yXYp7dFJ","voteType":null,"baseScore":1},"postCount":92,"description":"
  Compute governance is the subfield of AI governance concerned with controlling and governing access to computational resources.^[1]
  ..."},{"_id":"biPnaPsij8n2cir8s","core":false,"name":"Economics of artificial intelligence","slug":"economics-of-artificial-intelligence","tagRel":{"_id":"hnBC6dJfQjFYed2DR","voteType":null,"baseScore":1},"postCount":126,"description":"
  The economics of artificial intelligence studies the implications for economic growth and labor markets of increasingly powerful AI systems. The economics of AI also covers the study of short-term and long-term economic dynamics that affect the development..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"jyTdJoDMFXpTukKjG","voteType":null,"baseScore":2},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"4q7hZ9DBrvzZZqqz5","voteType":null,"baseScore":2},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."}]},{"_id":"uTffD74jJrQvHJuBY","slug":"coordinal-a-postmortem","title":"Coordinal: A Postmortem.","url":"https://ronakrm.github.io/coordinal-postmortem/","baseScore":89,"voteCount":33,"commentCount":6,"postedAt":"2026-05-18T20:43:33.501Z","curatedDate":null,"frontpageDate":"2026-05-18T20:52:52.067Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"text":"Most startups fail, this one did too. Tried too much with too little, with a little too much ego.","imageId":"SocialPreview/xjlb0hgkqzfx1zsignni"},"socialPreviewImageAutoUrl":"","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-05-22T22:24:21.892Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"XNjX6moCzGnRroLJJ","slug":"ronak-mehta","displayName":"Ronak Mehta","username":"Ronak Mehta","createdAt":"2026-02-12T19:40:35.341Z","profileImageId":null,"karma":103,"jobTitle":"Machine Learning Research and Engineering","organization":"Coordinal Research","postCount":2,"commentCount":2,"deleted":false,"biographyHtml":null},"contents":{"wordCount":1278,"htmlHighlight":"
  cart;horse: Most startups fail, this one did too. Tried too much with too little, with a little too much ego.
  Coordinal Research’s goal was to build an automated safety research platform. A researcher writes “Replicate X result from paper Y with tweak Z” and the system provisions a sandboxed compute environment, gathers context, writes code, runs experiments, and returns a research report plus oversight trail.
  The bet was that..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"bY6qtCihTzzSCw9gc","core":false,"name":"Announcements and updates","slug":"announcements-and-updates","tagRel":{"_id":"wRwAgFLJCDFMkJuhh","voteType":null,"baseScore":1},"postCount":2508,"description":"
  This topic collects posts that make an official announcement or update: for example, a new project or organization, a competition or prize, or major organizational news.
  
  Related entries
  organization updates |..."},{"_id":"y4b45yXXrFJpPmx7d","core":false,"name":"Opinion","slug":"opinion","tagRel":{"_id":"dQBBdMRuyX8rAh4Kp","voteType":null,"baseScore":1},"postCount":2134,"description":"
  Apply this topic to posts that express a personal opinion.
  
  Related entries
  
  community experiences | community
  ..."},{"_id":"imoDutHs2nFJ8aSmX","core":false,"name":"Postmortems & retrospectives","slug":"postmortems-and-retrospectives","tagRel":{"_id":"dLYacvyHdsfLwtnEG","voteType":null,"baseScore":1},"postCount":269,"description":"
  A postmortem or retrospective is a reflection on past actions with an eye to what went well, what didn't, and the cause of any failures. These retrospectives may include meta-commentary such as lessons learned, data gathered, and whether (and how) similar projects should be run in the future. Sharing such accoun..."},{"_id":"328weGzQNmTjyDm4g","core":false,"name":"Entrepreneurship","slug":"entrepreneurship","tagRel":{"_id":"pTJTBqvEp85XyjYRT","voteType":null,"baseScore":1},"postCount":325,"description":"
  The Entrepreneurship tag covers posts that discuss entrepreneurial ventures, including nonprofit startups and for-profit startups founded with social impact in mind.
  Related entries
  charity incubation
  ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"tRDsE2QLR2bztB8zi","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"4CH9vsvzyk4mSKwyZ","core":true,"name":"Career choice","slug":"career-choice","tagRel":{"_id":"gDx5Dqe226drdkByS","voteType":null,"baseScore":1},"postCount":3503,"description":"
  Career choice (in effective altruism) is about the ethical and practical considerations relevant to finding a high-impact career.
  Although people spend more time on their jobs than on any other activity besides sleep—about 80,000 hours over the course of a typical life..."},{"_id":"ct6dCPfJK9jkJnyAc","core":false,"name":"Building the field of AI safety","slug":"building-the-field-of-ai-safety","tagRel":{"_id":"ahh62PqRyxK4fZRZM","voteType":null,"baseScore":1},"postCount":455,"description":"
  Building the field of AI safety refers to the family of interventions aimed at growing, shaping or otherwise improving AI safety as an intellectual community.
  Related entries
  AI risk..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"jf2weF2Q6L7494Ccq","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."}]},{"_id":"aNdg7ctFP9zFcowNd","slug":"o3","title":"o3","url":null,"baseScore":84,"voteCount":30,"commentCount":9,"postedAt":"2024-12-20T21:00:03.486Z","curatedDate":null,"frontpageDate":"2024-12-20T21:31:11.040Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/aNdg7ctFP9zFcowNd/vs4vi6nanaaxz8kqqxbk","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2025-05-04T15:47:18.939Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"6KMRRie8e4nm4mYDE","slug":"zach-stein-perlman","displayName":"Zach Stein-Perlman","username":"zsp","createdAt":"2020-11-24T20:44:34.940Z","profileImageId":"Profile/dcqoc0v7tkemzirscfpq","karma":5736,"jobTitle":null,"organization":"","postCount":61,"commentCount":498,"deleted":false,"biographyHtml":"
  Researching donation opportunities. Previously: ailabwatch.org.
  "},"contents":{"wordCount":277,"htmlHighlight":"
  See livestream, site, OpenAI thread, Nat McAleese thread.
  OpenAI announced (but isn't yet releasing) o3 and o3-mini (sk..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"wbStQxsBRHyB84FtY","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"v4gzbpLGCKaRABXPr","core":false,"name":"OpenAI","slug":"openai","tagRel":{"_id":"yqauaHAyJjdtJoZcf","voteType":null,"baseScore":1},"postCount":172,"description":"
  OpenAI is an AI safety research and deployment company based in San Francisco.
  Funding
  In March 2017 OpenAI received a grant of $30 million from Open Philanthropy...."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"qTuCt8xDdcojRb2mf","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"wqstYnFPgdekKGLPc","core":false,"name":"Large Language Models","slug":"large-language-models","tagRel":{"_id":"d84tDgaMALDetoYoG","voteType":null,"baseScore":1},"postCount":272,"description":"
  This topic is for posts discussing Large Language Models (LLMs) -- for example, the GPT models produced by OpenAI.
  Related Entries
  
  AI safety |..."},{"_id":"ftyDfKF45Gig2u4sx","core":false,"name":"AI evaluations and standards","slug":"ai-evaluations-and-standards","tagRel":{"_id":"4K3ncaTej5RmyixyW","voteType":null,"baseScore":1},"postCount":289,"description":"
  AI evaluations and standards (or \"evals\") are processes that check or audit AI models. Evaluations can focus on how powerful models are (“capability evaluations”) and on whether models are exhibiting dangerous behaviors or are misaligned (“alignment evaluation..."}]},{"_id":"AbHPpGTtAMyenWGX8","slug":"are-the-costs-of-ai-agents-also-rising-exponentially","title":"Are the Costs of AI Agents Also Rising Exponentially?","url":"https://www.tobyord.com/writing/hourly-costs-for-ai-agents","baseScore":82,"voteCount":30,"commentCount":11,"postedAt":"2026-02-02T08:45:26.555Z","curatedDate":null,"frontpageDate":"2026-02-02T15:19:33.731Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"text":null,"imageId":"SocialPreview/jpdzi9gjdlbvlhkdzhpp"},"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/AbHPpGTtAMyenWGX8/b5wbyy8b3ihkn7it3huq","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-07-30T09:09:32.728Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"R4mvcEPhmLiBahN4H","slug":"toby_ord","displayName":"Toby_Ord","username":"Toby_Ord","createdAt":"2014-08-21T01:29:56.891Z","profileImageId":null,"karma":5071,"jobTitle":null,"organization":null,"postCount":25,"commentCount":203,"deleted":false,"biographyHtml":null},"contents":{"wordCount":2272,"htmlHighlight":"
  There is an extremely important question about the near-future of AI that almost no-one is asking.
  We’ve all seen the graphs from METR showing that the length of tasks AI agents can perform has been growing exponentially over the last 7 years. While GPT-2 could only do software engineering tasks that would take someone a few seconds, the latest models can (50% of the time) do tasks that would take a human a few hours.
  ..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"F8kXNXJ32YE7BYWGP","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"zMEXG96uqHaotkaf2","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"biPnaPsij8n2cir8s","core":false,"name":"Economics of artificial intelligence","slug":"economics-of-artificial-intelligence","tagRel":{"_id":"erktY6ymSt2cqFjNg","voteType":null,"baseScore":1},"postCount":126,"description":"
  The economics of artificial intelligence studies the implications for economic growth and labor markets of increasingly powerful AI systems. The economics of AI also covers the study of short-term and long-term economic dynamics that affect the development..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"qdNFetJ9G92uLAsK9","voteType":null,"baseScore":1},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"mrj6hwsmyKFL2otqp","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."}]},{"_id":"qtfX67cC4tij2zoKC","slug":"the-tables-have-turned-on-ai-sceptics","title":"The tables have turned on AI sceptics","url":"https://www.update.news/p/the-tables-have-turned-on-ai-sceptics","baseScore":81,"voteCount":41,"commentCount":36,"postedAt":"2026-05-07T17:53:24.727Z","curatedDate":"2026-05-08T13:02:53.399Z","frontpageDate":"2026-05-08T13:39:31.459Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":{"text":"Epistemic conservatism no longer favours long timelines","imageId":"SocialPreview/np4jqgonzzs72q5p411i"},"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/qtfX67cC4tij2zoKC/jytxswfenx195raq29tx","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-08-19T20:30:23.267Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"dvsskX8LYuAXm5q3P","slug":"stefan_schubert","displayName":"Stefan_Schubert","username":"Stefan_Schubert","createdAt":"2014-09-10T08:54:22.396Z","profileImageId":null,"karma":6917,"jobTitle":null,"organization":null,"postCount":16,"commentCount":726,"deleted":false,"biographyHtml":"
  I'm a researcher in psychology and philosophy.
  https://stefanschubert.substack.com/
  "},"contents":{"wordCount":897,"htmlHighlight":"
  Could we have human-level AI within the next few decades? For a long time, many people have dismissed this idea as armchair speculation. In their view, we shouldn’t ground our beliefs about transformative technologies in vague hunches and fragile multi-step arguments. We need more solid evidence, like clear empirical trends. We need to be epistemically conservative.
  I have some conservative instincts myself, but I’m not sure they favour long AI timelines anymore. That might have been th..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"gE4gCoY7Kc2LGLhCL","voteType":null,"baseScore":1},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"uqLooxveMizg3DEma","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"mdjkSFciiPTnAwiQx","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"biPnaPsij8n2cir8s","core":false,"name":"Economics of artificial intelligence","slug":"economics-of-artificial-intelligence","tagRel":{"_id":"beNkeY9ym3iYjemuM","voteType":null,"baseScore":1},"postCount":126,"description":"
  The economics of artificial intelligence studies the implications for economic growth and labor markets of increasingly powerful AI systems. The economics of AI also covers the study of short-term and long-term economic dynamics that affect the development..."},{"_id":"FumgMenMzAubeYyc2","core":false,"name":"Consequences of AI timelines","slug":"consequences-of-ai-timelines","tagRel":{"_id":"Xhzkf9FsJdYRnBQNd","voteType":null,"baseScore":1},"postCount":79,"description":"
  When human-level AI will emerge (AI timelines) is a consideration for both AI risk interventions and other interventions. For example, one's AI timelines affect how much to discount interventions that have a delayed effect.
  Related entries
  AI fore..."}]},{"_id":"ddeCNBhYc2sANsixS","slug":"ai-forecasting-research-ideas","title":"AI Forecasting Research Ideas","url":"https://docs.google.com/document/d/1Vva6AW_udvfavtldWuGbqf8NbbzsgEH8mESEX6NlOvM","baseScore":78,"voteCount":27,"commentCount":1,"postedAt":"2022-11-17T17:37:40.600Z","curatedDate":null,"frontpageDate":"2022-11-19T06:18:59.076Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2022-11-18T13:36:47.966Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"GsLh9BFphgxKfTDe8","slug":"jsevillamol","displayName":"Jaime Sevilla","username":"Jsevillamol","createdAt":"2019-03-04T19:24:58.585Z","profileImageId":"Profile/cs3xaei4c1geanb8wbzn","karma":5054,"jobTitle":"Director","organization":"Epoch","postCount":42,"commentCount":255,"deleted":false,"biographyHtml":"
  Director of Epoch, an organization investigating the future of Artificial Intelligence.
  Currently working on:
  Macroeconomic models of AI takeoff
  Trends in Artificial Intelligence
  Forecasting cumulative records
  Improving forecast aggregation
  I am also one of the coordinators of Riesgos Catastróficos Globales, a Spanish-speaking network of experts working on Global Catastrophic Risks.
  I also run Connectome Art, an online art gallery where I host art I made using AI.
  "},"contents":{"wordCount":148,"htmlHighlight":"
  Overview
  The linked document contains a collection of AI Forecasting research ideas, prepared by some Epoch employees in a personal capacity
  We think that these are interesting and valuable projects that research interns or students could look into, though they may vary in difficulty (depending on your background/experience)
  This is the result of a quick brainstorming and curation, rather than a thoroug..."},"group":null,"coauthors":[{"_id":"8m5a7mqm77SoCv6TK","slug":"lennart","karma":367,"deleted":false,"jobTitle":null,"username":"lennart","createdAt":"2019-04-29T14:38:36.474+00:00","postCount":5,"displayName":"lennart","commentCount":17,"organization":null,"biographyHtml":"","profileImageId":null}],"customHtmlHighlight":"","tags":[{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"HbQXtfksejLjKrGv3","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"75HnrKT9FNWKxxtPY","core":false,"name":"Collections and resources","slug":"collections-and-resources","tagRel":{"_id":"gHuFYyQ84zY7HYAa5","voteType":null,"baseScore":1},"postCount":1136,"description":"
  The collections and resources topic covers posts containing lists of EA-related resources. These could be external links or curated Forum content.
  Further reading
  Aird, Michael (2020) Sugg..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"gCt64fv2S64GqdALv","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"exjrXKC2emi6ByBvB","voteType":null,"baseScore":10},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"SMqtvDKoZFFdau8MT","core":false,"name":"Research agendas, questions, and project lists","slug":"research-agendas-questions-and-project-lists","tagRel":{"_id":"DxDkFzouvyxZhPcsz","voteType":null,"baseScore":3},"postCount":137,"description":"
  The research agendas, questions, and project lists tag covers posts that share these kinds of lists.
  External links
  Research agendas. A list of research agendas maintained by Effec..."},{"_id":"yq5yEhdHxS5LFmDpv","core":false,"name":"Epoch AI","slug":"epoch-ai","tagRel":{"_id":"znpBNJm4AsXC228Pe","voteType":null,"baseScore":3},"postCount":29,"description":"
  Epoch AI is a research initiative working on investigating trends in machine learning and forecasting the development of transformative artificial intelligence...."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"GvmcYGG7rizY2oexP","voteType":null,"baseScore":2},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."}]},{"_id":"P8jsAySQzfgkeoDgb","slug":"ai-benchmarking-has-a-y-axis-problem","title":"AI benchmarking has a Y-axis problem ","url":null,"baseScore":74,"voteCount":28,"commentCount":2,"postedAt":"2026-02-06T07:45:57.281Z","curatedDate":null,"frontpageDate":"2026-02-06T14:50:35.937Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/P8jsAySQzfgkeoDgb/wqjkvd5nmhpn4srdikbv","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-02-11T20:23:59.082Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"SuPnfB9qqKWsucNzm","slug":"lizka","displayName":"Lizka","username":"Lizka","createdAt":"2019-11-06T06:22:18.538Z","profileImageId":"Profile/llwmfw62hcsq2vuk1n7g","karma":18075,"jobTitle":"Researcher","organization":"Forethought","postCount":154,"commentCount":580,"deleted":false,"biographyHtml":"$64"},"contents":{"wordCount":2190,"htmlHighlight":"
  TLDR: People plot benchmark scores over time and then do math on them, looking for speed-ups & inflection points, interpreting slopes, or extending apparent trends. But that math doesn’t actually tell you anything real unless the scores have natural units. Most don’t.
  $\"\"$
  Think o..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"FBA9yXcvT7xgJW4tw","voteType":null,"baseScore":2},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"bfvNbfjoBLMkF9PpA","voteType":null,"baseScore":2},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"6znG9qpfDxWw5ctf6","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"zgz4DXga5rBCHQE6t","core":false,"name":"Measuring and comparing value","slug":"measuring-and-comparing-value","tagRel":{"_id":"fYJLwfJcEgKtHLg2z","voteType":null,"baseScore":1},"postCount":66,"description":"
  Measuring how much value interventions create can help us to evaluate their cost-effectiveness.
  Ideally, we would measure the things we value directly. Unfortunately, we often value things that are difficult to measure, so that we must instead measure other v..."},{"_id":"Asn9cXBLumGiCJQ5g","core":false,"name":"Epistemology","slug":"epistemology","tagRel":{"_id":"Jzm4wTcEJYwHKyFcu","voteType":null,"baseScore":1},"postCount":272,"description":"
  Epistemology is the study of how people should form credences about the nature of the world.
  Beliefs and credences are purely evaluative attitudes: they are simply about the way that we think the world is. A person might believe that it will rain, for exa..."},{"_id":"o2FZJuxRGHKvXM6HS","core":false,"name":"Statistics","slug":"statistics","tagRel":{"_id":"KAEcYrf6QubSRcbXP","voteType":null,"baseScore":1},"postCount":86,"description":"
  Statistics is the discipline that concerns the collection, organization, analysis, interpretation, and presentation of data.
  This tag isn't meant for posts that estimate the impact of specific charities or interventions, but instead for posts about \"doing statistics\" in general.
  Related entries
  ..."}]},{"_id":"foEWs9XvrNLzriX6f","slug":"ai-predictions-for-2026","title":"AI predictions for 2026","url":null,"baseScore":70,"voteCount":26,"commentCount":2,"postedAt":"2026-01-20T10:35:44.313Z","curatedDate":null,"frontpageDate":"2026-01-20T12:07:58.834Z","draft":false,"question":false,"isEvent":false,"groupId":null,"sticky":false,"eventImageId":null,"socialPreview":null,"socialPreviewImageAutoUrl":"https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/foEWs9XvrNLzriX6f/ymumu1a42tqfu4lzq5m3","readTimeMinutesOverride":null,"collabEditorDialogue":false,"disableRecommendation":false,"authorIsUnreviewed":false,"lastCommentedAt":"2026-01-25T17:28:52.432Z","sharingSettings":null,"shortform":false,"startTime":null,"onlineEvent":false,"googleLocation":null,"user":{"_id":"d7mH3WSTpFBA2wYdJ","slug":"ajeya","displayName":"Ajeya","username":"Ajeya","createdAt":"2016-08-14T20:12:44.994Z","profileImageId":null,"karma":2347,"jobTitle":null,"organization":null,"postCount":25,"commentCount":76,"deleted":false,"biographyHtml":null},"contents":{"wordCount":2165,"htmlHighlight":"
  But first, scoring my predictions for 2025
  
  Note: This post was crossposted from the Planned Obsolescence by the Forum team, with the author's permission. The author may not see or respond to comments on this post.
  On December 13th, 2024, I registered predictions about what would happen with AI..."},"group":null,"coauthors":null,"customHtmlHighlight":"","tags":[{"_id":"GKeFgBJnBSuGePbgh","core":false,"name":"Transformative artificial intelligence","slug":"transformative-artificial-intelligence","tagRel":{"_id":"JPtwFhodrGc5L9e23","voteType":null,"baseScore":1},"postCount":187,"description":"
  Transformative artificial intelligence (TAI) is artificial intelligence that constitutes a transformative development, i.e. a development at least as s..."},{"_id":"aJnrnnobcBNWRsfAw","core":true,"name":"Forecasting","slug":"forecasting","tagRel":{"_id":"N8HFAL2RHZtKHPYFx","voteType":null,"baseScore":1},"postCount":2076,"description":"
  Forecasting is an important tool for improving the future, because good forecasts and estimates can help us appropriately plan interventions and assess risks. Over the past several decades there has been significant research and investment in forecasting and estimation techniques, tools, and organizations. This continues to be a..."},{"_id":"EHXNGqENrziYQJsox","core":false,"name":"AI forecasting","slug":"ai-forecasting","tagRel":{"_id":"ZuzCqLHFfFYQz2d6n","voteType":null,"baseScore":1},"postCount":539,"description":"
  AI forecasting is the process of trying to predict a variety of outcomes related to AI, such as when human-level AI will emerge (AI timelines) and what impacts it will have.
  Further reading
  Cotra, Ajeya (2020) ..."},{"_id":"BkbzZq7thbXcyg8Be","core":false,"name":"AI benchmarks","slug":"ai-benchmarks","tagRel":{"_id":"xis3dmQqkcgTme8Ec","voteType":null,"baseScore":1},"postCount":109,"description":"
  Benchmarks are tests which enable us to measure the progress of AI capabilities, and test for characteristics which might pose safety risks.
  Further reading
  The Benchmark Lottery
  ..."},{"_id":"oNiQsBHA3i837sySD","core":true,"name":"AI safety","slug":"ai-safety","tagRel":{"_id":"We8bsss3MiskR8Ek7","voteType":null,"baseScore":1},"postCount":7240,"description":"
  AI safety is the study of ways to reduce risks posed by artificial intelligence.
  Interventions that aim to reduce these risks can be split into:
  Technical alignment..."},{"_id":"ee66CtAMYurQreWBH","core":true,"name":"Existential risk","slug":"existential-risk","tagRel":{"_id":"JpRGNdnoeDawNLCPe","voteType":null,"baseScore":1},"postCount":4355,"description":"
  An existential risk is a risk that threatens the destruction of the long-term potential of life.[1] An existential risk could threaten the..."}]}],"initialLimit":15,"loadMoreLimit":50}]]}]}]

Further reading

Further reading

Related entries

Further reading

Related tags

Further reading

Related entries

Further reading

Improving policy

Introduction

Further reading

Related entries

Terminology

Further reading

Terminology

Further reading

Related Entries

Further reading

Further reading

Further reading

Further reading

Related entries

Related entries

Related entries

Funding

Further reading

Related Entries

Further reading

Further reading

Further reading

Related entries

Further reading

Further reading

Further reading

Further reading

Related entries

But first, scoring my predictions for 2025

Further reading