Article
45990 discoveries · Hand-picked discoveries from the open web, newest first.
-
‘Hen’ Review: A Cockeyed Comedy Sees Europe’s Migrant Crisis Through the Eyes of a Chicken
Executive-produced by Ari Aster, György Pálfi's new film evokes "EO" and "Flow" with its animal POV on humanity's moral failures.
IndieWire · EN
-
‘North of North’ Season 2 Review: Netflix’s Defiantly Nice Arctic Sitcom Is for the Warmhearted F**k-Ups
Indie Spirit and Gotham Award nominee Anna Lambe shines even brighter in a second season that sees her single mom double down on "leveling up" — at work, in love, and as a contributing member of her cherished Inuk community, which, like her, needs all the help it can get.
IndieWire · EN
-
‘Wild Horse Nine’ Review: Martin McDonagh’s Prickly but Unconvincing CIA Satire Recasts ‘In Bruges’ as the Stuff of a Classic Western
John Malkovich and Sam Rockwell are two CIA agents vacationing on Easter Island in McDonagh's sardonic look at American interventionism.
IndieWire · EN
-
‘The Weed Eaters’ Lands Expanded Release from Yellow Veil Pictures: New Zealand Cannibal Movie Sets Winter Rollout
Exclusive: Director Callum Devlin's feature gets expanded stateside distribution after debuting in the Letterboxd Video Store in September.
IndieWire · EN
-
‘La Bola Negra’ Wins Denver Film Festival Top Prize as More Regional Fests Boost Oscar Hopefuls
The Sly Stallone behind-the-scenes biopic "I Play Rocky" won the narrative feature audience award, repeating a pattern that started in Toronto.
IndieWire · EN
-
‘Everytime’ Review: Sandra Wollner’s Infinitely Beguiling Portrait of Grief in the Digital Age Is One of the Most Haunting Films of the Decade
Picking up where the likes of "Aftersun" and "Personal Shopper" left off, Wollner's unforgettable third feature explores the modern world's changing relationship to death.
IndieWire · EN
-
Gatorville
In Colorado’s forgotten valley, two siblings face alligators and the ache of leaving youth behind.
Short of the Week · EN
-
Adversarial Training Can Further Hide a Backdoor, but Does It Improve Trigger Identification?
Summary The Sleeper Agents paper demonstrated that adversarial training on red-team prompts failed to remove a planted backdoor. The authors further suggested that it taught the model to "better identify when to act unsafely." This would pose a significant danger, as a standard safety practice would be refining the very behavior it set out to remove. However, the degree of danger depends on exactly what the model is learning during adversarial training. We discuss how, due to saturation of firing rates, more benign scenarios such as the model's backdoor response being generally weakened are al
LessWrong · EN
-
Better honeypots, fewer tries: training an optimizer to make PPO cheat
I trained a small optimizer on synthetic functions, and used it to search gridworld levels where PPO cheats. TL;DR Reward-hacking honeypots are levels that are built by hand where a reward is given to the agent for cheating (e.g. taking a shortcut to the goal). I converted this process into a black box optimization problem where the score is based on how often PPO cheats. I used evolution strategies to train the optimizer on the cheap, synthetic functions, and applied it to the honeypot without any further training. Given the same PPO budget, my best policy (RLBO) found levels where PPO cheate
LessWrong · EN
-
Clashing Intuitions for the Plausibility of Alignment (Illustrated via Differing Causal Models)
Epistemic Status: Brief explanation of an unsettled debate, not novel theory; writing is entirely my own, with diagrams from ChatGPT 5.5. (Originally posted on twitter ) Recent debates have highlighted a divergence between Yudkowskian AI-doomers and AI-whisperers about how capabilities will evolve that is really important. Specifically, j⧉nus thinks that alignment has been going really well, and that we should update significantly from current jagged capabilities. In my view, whether or not that criticism is warranted relies on specific details about future capability trajectories, which then
LessWrong · EN
-
On marrying the egregore
I mentioned in the last post that Kegan's lumping of all "relationships" including identification with one's community into stage 3 has some interesting implications. Basically, people invent an individual named "everyone", use this as a cognitive shortcut, and then become subject to this cognitive shortcut which can take on a life of its own as an egregore: The big divergence between my description and Kegan's starts where I ended describing stage 3 as about being subject to dyadic relationships and stage 4 about making these object in serving larger group coordination. Kegan calls stage 3 ab
LessWrong · EN
-
AI #193: Progress May Dramatically Accelerate
Anthropic is warning us. Its modified S-1 now says that ‘we may imminently meet our Responsible Scaling Policy’s definition of ‘dramatically accelerated progress attributable to AI.’ That is a highly scary sentence to have a company put in its S-1. This could happen. For now I appreciate the relative quiet. Perhaps many were focused the election. I’m sure the quiet will not last long. We have a new biggest crypto hack, as Binance got taken for billions of dollars. The big debate of the week was covered in Scott Alexander’s Against SURA Seems Deeply Confused . OpenAI updates its Model Spec. It’
LessWrong · EN
-
Linkpost to METR’s Anthropic Speedup Assessment
TLDR: roughly a 2x speedup, with a range of 1.3-3x. Model 4 looks very scary. https://metr.org/blog/2026-11-05-anthropic-ai-rd-acceleration-assessment/ Discuss
LessWrong · EN
-
Is it better to directly or indirectly ban ASI?
Senator Bernie Sanders has proposed directly banning development of ASI . Given the current scientific understanding of AI alignment, it is highly sensible to ban ASI and then consider revoking this ban if future scientific breakthroughs (e.g., “solving interp”) make it possible to safely develop ASI, or if there is truly no way to coordinate on everyone avoiding developing ASI. Some people have proposed banning ASI indirectly instead. For instance, one could mandate: Developing AI with superhuman capabilities is not permitted unless you can demonstrate that the AI systems you are developing w
LessWrong · EN
-
METR: Independent Assessment of AI R&D Acceleration at Anthropic in October 2026
Executive summary From August 12 to October 29, 2026, METR staff performed a high-access evaluation at Anthropic to assess the degree to which AI is accelerating AI R&D. If AI systems started to significantly accelerate further AI development, this could lead to a sudden emergence of extreme and risky capabilities . Anthropic offered unprecedented access to internal data on their AI R&D process, including a variety of metrics on pretraining, reinforcement learning (RL), training compute in each domain, benchmark scores for internal models, AI speedup analyses, forecasts of a future model’s met
LessWrong · EN
-
Thursday assorted links
1. Economics job market site. 2. Data on the flattening of consumer culture (NYT). One-sided, but still of interest. 3. A new and better measure of how much of educational value is signaling. 4. Inside joke concerning the Pope that only some Beatles fans will get. 5. U.S. productivity growth is looking up. 6. NYT […] The post Thursday assorted links appeared first on Marginal REVOLUTION .
Marginal Revolution · EN
-
Love at first photo?
Romantic interest on swipe-based dating apps (SBDAs) is highly concentrated. This can frustrate both popular users who feel overwhelmed by a glut of attention, and less popular users who feel dejected by a lack of it. Is the concentration of interest exacerbated by the platform design of dating apps? I provide causal evidence on the […] The post Love at first photo? appeared first on Marginal REVOLUTION .
Marginal Revolution · EN
-
Inconvenient brands
For an upcoming blog post, I’m collecting a list of small businesses that are worth bypassing Amazon for. Holiday shopping is around the corner, and sometimes people want to put in the effort to buy something that takes more than a click. Folks like Burlap and Barrel, Gramercy Tools or Askinosie Chocolate–direct trade small businesses. […]
Seth's Blog · EN
-
How Paternity Leave Policies Help Moms, Babies, and Families
When our second child was born, my spouse had several weeks of paternity leave under his parental leave policy at work. I’m a certified family life educator, and parental leave comes up regularly in my teaching and community work. So I was aware of the benefits his leave could have for our family—in theory. Still, I could not have imagined how significantly his leave would end up affecting our family and our connections with community. During his leave time, my spouse took our oldest child to playdates and birthday parties. He quickly became connected with our small-town network of families wi
Greater Good · EN
-
Russell Brand is convicted of rape, and judge says he'll get a prison term
A prosecutor described Brand as a ruthless, calculating, manipulative "sexual predator." Brand, 51, denied the charges, saying he had never forced anyone to do something against their will.
NPR Arts · EN
-
Min Jin Lee on 'American Hagwon' and why it takes a decade to write each novel
Lee's new novel is about the promise and perils for Koreans who pursue a Western education. Her previous book, the 2017 novel Pachinko , was a finalist for a National Book Award.
NPR Arts · EN
-
With humor, heart and homunculi, 'Small Prophets' is a gentle marvel
A man calls forth magical spirits in an effort to find his long-missing girlfriend in this new British comedy-fantasy. Original and surprising, Small Prophets imbues daily life with poetic touches.
NPR Arts · EN
-
Feminism for a World on Fire by Natasha Walter audiobook review – an urgent call for collective action
The writer and activist finds galvanising energy in grassroots initiatives in the face of inequality and the climate crisis ‘Throughout the world confidence in progress has shattered,” notes Natasha Walter. “The certainty – even the hope – that human rights equality might be within everyone’s reach has disintegrated … Everywhere in the world, from west to east, south to north, women are struggling to win and to protect their rights, and to live in safety. They keep coming up against misogynistic violence that never disappears.” In Feminism for a World on Fire, Walter, an activist and academic,
Guardian Culture · EN
-
‘A hero’s origins’: the true story of the first drag queen presidential candidate
In off-Broadway show Ms Blakk for President, Oscar-winning Moonlight writer Tarell Alvin McCraney explores a slice of under-appreciated queer history At the height of the Aids crisis, Chicago activist Terence Alan Smith put on his drag persona, Joan Jett Blakk, and announced she was running for president. Her slogan: “Lick Bush in ‘92!” Her platform: fire everyone in the White House and paint it pink, reroute the military budget towards education and healthcare, then give the land back to Indigenous people “and let them take it from there”. The campaign did not advance. But Blakk did make it o
Guardian Culture · EN