Aksel Sterri
@AkselSterriPhilosophy, emerging technologies, markets & health. Research director @Langsikt_. Previously: @Harvard, @UniofOxford, @UniOslo, @OsloMet
Similar User
@kjetilba
@Civita_
@jasnoen
@gudmundh
@anineki
@lillasolhusvik
@Minervanett
@Manifesttanke
@achalse
@jonwesselaas
@vanebo
@Braanen
@Klungtveit
@nilaug
@emiers
In this blog post at @ethicsinthenews, I argue that AI suffers from a transition paradox. blog.practicalethics.ox.ac.uk/2022/04/ai-and…
Excited to share a new paper titled Effective Mitigations for Systemic Risks from General-Purpose AI. This paper would not have been possible without my brilliant and hard-working co-authors @Annemieke_Br, Tim Schreier, @NoemiDreksler, @valeriapuligna2, and @RishiBommasani
While I wholeheartedly agree, I think this is often where economics empirical evidence stops and thus fails. Partly we need to address whether there are institutional incentives to get policing right, what the right approach is to organizing and motivating bureaucratic change,…
One of the most consistent research findings in the CJ space is that hiring more police reduces crime. We need more and better policing, not less — and that will require more funding.
Happy birthday @givingwhatwecan! Proud member!
Happy Birthday to @givingwhatwecan, which turns 15 today! 🥳 When @tobyordoxford and I started GWWC 15 years ago, we had just 23 members, each of us pledging at least 10% of our income to effective charities until we retire. Now, over 9,000 people have taken this pledge...…
Sjekk SMS-er du får i dag! Kanskje får du være med i Framtidspanelet? I går var @EirikMofoss på NRK Debatten og diskuterte hva Norge bør gjøre med sine ekstraordinære inntekter. Denne samtalen er imidlertid ikke forbeholdt de som deltar i TV-debatter. framtidspanelet.no
Climate change terminology does not influence willingness to take climate action Our new paper (N=6,132, across 63 countries) suggests that focusing on subtle terminology in climate messaging is not an effective use of resources. sciencedirect.com/science/articl… Led by @dgoldwert &…
«Å ignorere AGI-risiko fordi vi har andre presserende problemer nå, er som å droppe investeringer i pandemiberedskap fordi folk trenger hofteoperasjoner.» 💡 @AkselSterri , @EirikMofoss dn.no/innlegg/fordum…
I'm on the front page of The Times today with @jeffrsebo - see if you can guess which story. Thanks @whippletom How I wish I could control headlines!
Our paper on the ”Class Ceiling in Politics” is now out in the APSR with open access. For an overview of the key findings see this thread! x.com/ollefolke/stat…
Just published on APSR First View: "The Class Ceiling in Politics" by Olle Folke (@ollefolke) and Johanna Rickne (@johannarickne). cambridge.org/core/journals/…
1/10 Today we're launching FrontierMath, a benchmark for evaluating advanced mathematical reasoning in AI. We collaborated with 60+ leading mathematicians to create hundreds of original, exceptionally challenging math problems, of which current AI systems solve less than 2%.
Having written extensively about how real wages have risen strongly for working class Americans since 2019, let me be clear about a few things, on both the facts, and how we should interpret them. 🧵
Den 13. november vil 40 000 få en SMS med invitasjon til å delta i Framtidspanelet, og deretter vil et lotteri trekke ut de heldige. Er du mellom 16 og 26 år er det enda større sjanse for å bli med, vi vil nemlig nå ut til de unge. 💡🗣️🌏
Sign me up!
Can LLMs be induced to deviate from optimal gameplay in a simple game by threats of pain or promises of pleasure? And does the probability of deviating depend on the intensity of the promised pleasure/pain? According to our new paper, released today (arxiv.org/abs/2411.02432), the…
OpenAI’s biggest rival is shaking things up. Anthropic invited 200+ elite hackers to their SF headquarters to see what’s possible with Claude Here’s what we saw at the @AnthropicAI x @MenloVentures Builder Day Hackathon (🧵):
The UK may not be the leader in AI, but at least they seem to be #1 in the West on non-consensual facial recognition - yay!
Big Brother proudly announces Amazon-powered facial recognition in UK rail hubs to observe passengers’ emotions. Critics call it intrusive—how absurd! Results? People are annoyed en route to work, frazzled homebound.
Taiwan’s TSMC Arizona fab to begin 4 nm production in December taiwannews.com.tw/news/5964282
"Nearly 75% of all politicians adhere to what we call "democratic realism" - they think that voters are ignorant, short-term focused, egocentric, and vote based on deeply held identities rather than policy positions. This happens in every country we study."
Now in @apsrjournal: What do politicians think about their voters? Fielding face-to-face surveys to 982 sitting politicians in 11 countries, and surveys of 12,000 citizens, we find they have remarkably consistent - and cynical - theories of voters: /1 doi.org/10.1017/S00030…
"If open model capabilities reach the previous frontier of closed models within a year (or less), this leaves a relatively short window to understand capabilities at the frontier before they exist in models that anyone can download."
1/ Are open-weight AI models catching up to closed models? We did the most in-depth investigation to date on the gaps in performance and compute between open-weight and closed-weight AI models. Here’s what we found: 🧵
United States Trends
- 1. #UFC309 57,8 B posts
- 2. #MissUniverse 145 B posts
- 3. Brian Kelly 9.502 posts
- 4. Mac Miller N/A
- 5. Beck 15 B posts
- 6. Jim Miller 5.788 posts
- 7. #AEWCollision 11,9 B posts
- 8. Feds 37,4 B posts
- 9. Mizzou 6.912 posts
- 10. Nebraska 12,5 B posts
- 11. Romero 18,2 B posts
- 12. Onama 2.800 posts
- 13. Tennessee 40,6 B posts
- 14. Wisconsin 45,3 B posts
- 15. Gators 11,5 B posts
- 16. Louisville 7.329 posts
- 17. Dinamarca 9.309 posts
- 18. Dalton Knecht 3.299 posts
- 19. #LAMH N/A
- 20. #Svengoolie 2.228 posts
Who to follow
-
Kjetil B. Alstadheim
@kjetilba -
Civita
@Civita_ -
Jan Arild Snoen
@jasnoen -
Gudmund Hernes
@gudmundh -
Anine Kierulf
@anineki -
Lilla Sølhusvik
@lillasolhusvik -
Minerva
@Minervanett -
Manifest Analyse
@Manifesttanke -
Andreas C. Halse
@achalse -
Jon Wessel-Aas
@jonwesselaas -
Ove A. Vanebo
@vanebo -
Bjørgulv Braanen
@Braanen -
Harald S. Klungtveit
@Klungtveit -
Nils August Andresen
@nilaug -
Emil André Erstad
@emiers
Something went wrong.
Something went wrong.