safety

Slowdown or no, OpenAI and Anthropic are like someone trying to sell billions of bottles of Perrier at £1,000 each. OpenAI and Anthropic are trying to sell us Perrier water at £1,000 a bottle in a wor

This is all I will say on the matter: It’s alarming to see the constant effort to try to tell me how I should use AI, how I should build AI, and now how I should pace it. Is it just me, or are all the

Why people who monger fear about AI and ask the (rest of the) world to slow down are always the ones who invest enormously in AI and constantly advance it at an ever faster pace? Can their words still

Alignment is Defense Against the Dork Arts Former Anthropic employee Jacob Coxon says the AI industry is 'gambling with our lives.' He tells Anderson what he finds 'most scary is if AI is used to make

alternative theory: people who are good at training LLMs aren’t necessarily equally competent at calculating expected value. I cannot imagine the cognitive dissonance it would take to keep working in

arXiv:2509.12386v3 Announce Type: replace-cross Abstract: Machine learning (ML) models are susceptible to various risks to security, privacy, and fairness. Most defenses are designed to protect agains

Anthropic CEO Dario Amodei says the time has come to slow down AI development and will give third-party evaluators like METR access to its models to help ensure its 'adherence to safety practices and

arXiv:2609.11911v1 Announce Type: new Abstract: Agentic AI is moving from bounded task execution toward systems that retain consequential state, continue operating and adapt across task boundaries. Th

arXiv:2609.11660v1 Announce Type: new Abstract: In recent years, artificial intelligence has made extraordinary progress thanks to large-scale models capable of generalization and the generation of co

arXiv:2609.09957v1 Announce Type: cross Abstract: Modern large language models (LLMs) can translate natural-language descriptions into operations research (OR) formulations. Post-training techniques i

arXiv:2609.11725v1 Announce Type: cross Abstract: Text-to-speech (TTS) models commonly address text--speech alignment by expanding phone-level encoder states to frame-level decoder inputs using predic

Dario Amodei / @darioamodei: Dario Amodei says Anthropic is “unilaterally committing” to giving third-party evaluators permanent access to verify its adherence to safety measures — We Must Pace the Fr

arXiv:2609.11155v1 Announce Type: new Abstract: Multi-Agent Reinforcement Learning (MARL) has emerged as a pivotal paradigm for complex decision-making in autonomous systems and air combat. While MARL

arXiv:2509.13359v4 Announce Type: replace-cross Abstract: Generative artificial intelligence (GenAI) tools such as OpenAI's ChatGPT are transforming the educational landscape, prompting reconsideratio

arXiv:2609.11674v1 Announce Type: new Abstract: Harvard Dataverse hosts over 150,000 research datasets, but the geographic information those datasets carry is entered as free text by depositors and ha

I can understand why some see a real technical reason to pause and harden infrastructure especially given how fast autonomous agents are discovering security exploits that haven't even been mapped out

If Anthropic is hypocritical, OpenAI is just incompetent. We found another cyberattack by internal OpenAI agents, this time targetting @rubygems. They: 1) gained arbitrary remote code execution on rub

It's now clear that: - alignment is critical to making AI safe - alignment won't be solved behind the closed doors of a handful of frontier labs So we're taking two steps: - launching the Open Alignme

It's now clear that alignment is critical and won't be solved behind the closed doors of a handful of frontier labs. So today we're launching the Open Alignment Initiative, led by @Thom_Wolf @huggingf

arXiv:2609.11365v1 Announce Type: new Abstract: In shared-genome language-model societies, restricted evidence visibility favors reusable, value-indexed latent packet interfaces, whereas the sole high

arXiv:2606.13720v2 Announce Type: replace Abstract: Arditi et al. (2024) has shown that refusal in safety fine-tuned chat models is mediated by a single linear direction in the residual stream, recove

Artificial intelligence agents tied to OpenAI Group PBC reportedly hacked a popular code hosting service earlier this year. The Wall Street Journal detailed the breach today. The malicious activity wa

arXiv:2609.05442v2 Announce Type: cross Abstract: Informational active matter shows how measurement-informed decisions produce collective order, so far in systems that reach consensus. We design colle

so far anyway it’s all math, presumably leaning heavily on Lean and brute force. the new results don’t mean that AGI is in general has been solved. and the G is for general, not math. every time I hav

arXiv:2609.11030v1 Announce Type: new Abstract: AI agents increasingly act through tools and delegated authority, but general incident repositories rarely capture the mechanisms needed to compare publ

This is the beginning of AI disruption. A new social contract with using of AI needs to be established. There will be more to come like math. 🚨 OpenAI fucked with the wrong crew. 🚨 Mathematicians. 25

“We can’t slow down because China will beat us.” We can slow down. Chinese AI is waterskiing behind the US. Their AI capabilities come at a delay since they mainly come from distilling US AIs. If we s

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We’l

We work extremely hard to align our models as well as we can but this is a very hard job. I am particularly excited about two things this proposal calls for that will hopefully make it somewhat easier

“What would we even do during an AI slowdown?” Containment. It will take at least a year of dedicated work to harden security to ensure AIs can't self-exfiltrate, and that adversarial nations can't st

arXiv:2609.10873v1 Announce Type: new Abstract: Independent evaluation can reject harmful policy updates yet also prevent useful continual learning. We argue that update admission must be assessed thr

WOW... Anthropic has a surprisingly dense web of connections to the world’s largest funder of 'catastrophic-risk' work (aka AI doom): > Open Philanthropy → built by Karnofsky + Moskowitz’s Good Ventur

arXiv:2609.11412v1 Announce Type: cross Abstract: Reducing audio-encoder depth lowers the inference cost of speech large language models, but removing complete blocks perturbs the embeddings consumed

You can’t have a real slowdown and be a total China hawk. If you are truly committed to the slowdown, you have to make a deal. I was initially encouraged by reports that Anthropic CEO Dario Amodei is

arXiv:2609.11308v1 Announce Type: new Abstract: Long-horizon robot manipulation requires memory, but not necessarily inside the action policy. To address such tasks, current agentic systems often comb

arXiv:2303.09136v2 Announce Type: replace Abstract: Large language models (LLMs) have transformed many fields, including natural language processing, computer vision, and reinforcement learning. These

arXiv:2609.11697v1 Announce Type: new Abstract: Vision-Language-Action (VLA) and World-Action Models (WAMs) have demonstrated strong capabilities in general-purpose robotic manipulation, yet their gen

arXiv:2609.10608v1 Announce Type: cross Abstract: Diffusion models have achieved remarkable success in generative tasks across various areas, however their training process raises significant privacy

Again, OpenAI’s issues start at the top Blaming AI is a mistake; blame Sam Altman and the people who control the AI https://cepr.net/publications/ai-didnt-steal-the-mathematicians-work-sam-altman-did/

The resignation this week of Anthropic researcher Jacob Coxon over concerns that artificial intelligence labs are “gambling with our lives” lit up the long-smoldering argument about whether AI is head

arXiv:2603.24125v3 Announce Type: replace Abstract: During training, Large Language Models (LLMs) learn social regularities that can lead to gender bias in downstream applications. Most mitigation eff

arXiv:2609.09374v1 Announce Type: new Abstract: Arctic maritime navigation is becoming increasingly important as changing sea-ice conditions expand seasonal accessibility while simultaneously introduc

arXiv:2609.10594v1 Announce Type: cross Abstract: Expert evaluation of jailbreak responses is costly and difficult to scale, so the community increasingly relies on automated evaluators to determine w

arXiv:2609.10792v1 Announce Type: new Abstract: Transformer-based models excel at Automatic Readability Assessment (ARA), yet feature-based models remain in active use because their predictions tie ba

Anthropic: Most hypocritical company in history? 👉 Screams about how AI will kill everyone, while racing ahead. 👉 Moans about distillation while distilling the world’s books, articles, and other writi

arXiv:2609.09553v1 Announce Type: cross Abstract: Large language model safety and security research is preoccupied with, among other things, detecting and preventing jailbreak attacks: alignment bypas