NunoSempere
Bio
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>
Posts 114
Comments1256
Topic contributions14
I followed up with classic 2019 EA forum post Aligning Recommender Systems as Cause Area with a podcast with Ivan Vendrov here: https://x.com/NunoSempere/status/1965820448123629986
Here is an endpoint that takes a google doc and turns into a markdown file, including the comments. https://docs.nunosempere.com. Useful for automation, e.g., I downloaded my browser history, extracted all google docs, summarized them, and asked for a summary & blindspots.
Here is the executive summary and few sections for this week's brief on global risks, by my team @ Sentinel.
- Geopolitics: Trump and Putin met in Alaska to discuss the Ukraine war. Forecasters’ estimate of the chance of a ceasefire by October dropped from 27% pre-summit to 9%.
- Biorisks: The chikungunya virus continues to spread, including in France and the UK.
- Tech and AI: Meta’s policies explicitly allowed its AI chatbots to “engage a child in conversations that are romantic or sensual.”
- And more: Three soldiers were killed and four others injured in a drone attack by FARC dissidents on the Colombian military.
Geopolitics: Trump/Putin summit in Alaska
Trump and Putin met in Alaska to discuss the Ukraine war. Before the summit, forecasters estimated a 27% (15% to 40%) probability of a ceasefire by October 1. After the summit, our forecast dropped to a 9% probability (2% to 30%).
Before the summit, Zelensky told his European counterparts that he would be willing to formally cede territory that Russia already occupies in exchange for freezing the conflict along the current frontlines, while Putin was demanding that Ukraine withdraw its troops from the entirety of the Donbas in exchange for a freeze everywhere else (a demand Zelensky rejected). Trump also threatened “severe consequences” for Russia if a ceasefire wasn’t agreed, presumably in the form of more military aid to Ukraine and further sanctions and secondary tariffs on Russia. Still, based on the failure of previous talks, our forecasts for the chance of a ceasefire were significantly below 50%, although the fact that Trump and Putin were meeting at all was setting high expectations (with Trump claiming a 75% chance that the summit would be successful).
After the summit, forecasters consistently reduced their estimates. Trump didn’t take the chance to exert pressure, with US Secretary of State Marco Rubio arguing that further sanctions would only prolong the war. Instead, Trump adopted Putin’s position of aiming for a comprehensive peace agreement, which would likely take much longer to achieve, as there are many issues to resolve. With Putin proposing a land swap (potentially slanted 10-to-1 in favour of Russia) that puts Russian soldiers on the Kyiv side of the extensive defensive lines Ukraine has built, some of our forecasters believe there is a big gulf between the two sides’ positions, a sentiment echoed by Rubio. Others think that withdrawal from the Donbas might be acceptable to Ukraine if there are robust security guarantees for Ukraine as part of a peace deal, most likely in the form of European boots on the ground in the country.
rest of the brief here.
Instead of letting the social fabric collapse, everyone suddenly turns their ire on one person, the victim. Maybe this person is a foreigner, or a contrarian, or just ugly. The transition from individuals to a mob reaches a crescendo. The mob, with one will, murders the victim (or maybe just exiles them).
Maybe this person is a contrarian, but Girard also argues that the scapegoat effect is greater if the person is like any other member of the public, because then this scares the participants more because "it could have been me"
Executive summary for this week's global risks roundup: