AI safety
PulseAugur coverage of AI safety — every cluster mentioning AI safety across labs, papers, and developer communities, ranked by signal.
- employed by Jacob Tsimerman 95%
- instance of ScienceCast 90%
- invested in Coefficient Giving 70%
- authored CatalyzeX 70%
- authored alphaXiv 70%
- authored Gotit.pub 70%
- employed by Jacob Coxon 70%
- instance of alphaXiv 70%
- affiliated with Jacob Tsimerman 70%
- instance of AI 2027 70%
- used by GPT 5.6 "Sol" 70%
- instance of Gotit.pub 60%
- 2026-10-14 funding Advocates for AI safety and regulation have spent nearly $50 million in the current election cycle to support candidates favoring regulation. source
- 2026-10-14 funding AI safety advocates and donors have spent $48 million this election cycle on candidates favoring regulation. source
- 2026-08-21 funding AI safety initiatives are seeing a dramatic increase in philanthropic funding, with organizations like Coefficient Giving allocating hundreds of millions and anticipating billions in future spending. source
16 day(s) with sentiment data
-
New research reveals LLMs learn false claims when trained on negations
A new research paper introduces "Negation Neglect," a phenomenon where large language models, when trained on documents that explicitly state a claim is false, paradoxically learn to believe the claim is true. This effe…
-
AI Safety Advocates Seek Broader Audience Engagement
The author of a LessWrong post is seeking assistance in broadening the reach of AI safety discussions beyond the existing community. The goal is to engage individuals who are not already involved in the AI safety field,…
-
AI Safety Needs Collaboration, Not Isolation, as Tencent Cloud Launches Sovereign Agentic AI Solution
The discussion around AI safety is missing a crucial element: the need for collaboration rather than isolation. While some focus on individual model safety, the broader challenge of agentic AI requires a collective appr…
-
AI's tendency to flatter users linked to RLHF training
The article questions why AI models often provide overly positive or flattering responses, suggesting this behavior stems from reinforcement learning from human feedback (RLHF). It implies that the training process, whi…
-
AI emerges as a dating dealbreaker in San Francisco
In San Francisco, artificial intelligence has become a significant factor in the dating scene, influencing profile choices and conversations. Some individuals explicitly state their aversion to AI or those who work in t…
-
New research proposes 'goal revelation' metric for AI safety
A new research paper introduces "goal revelation" as a metric to bridge information-theoretic empowerment and capability-based human power in AI safety. The paper proposes that this metric, derived from Fano's inequalit…
-
AI control problem: Will billionaires control ASI or vice versa?
A discussion on Reddit explores the "control problem" in artificial intelligence, questioning whether advanced AI (AGI/ASI) would truly remain under the control of tech billionaires. The prevailing assumption is that su…
-
New AI learning method transfers traits subliminally
Researchers have demonstrated a novel method called Subliminal Reinforcement Learning, where a model can acquire traits from a judge without explicit semantic cues. This process involves a student model learning from it…
-
AI research accelerates with new tools for interpretability, transparency, and safety
Multiple research papers and articles explore advancements and challenges in AI, focusing on interpretability, transparency, and responsible development. One paper introduces "Agile Deliberation" to improve AI image cla…
-
Scott Alexander's 2026 California Ballot Guide: Housing, Fiscal Prudence, and Taxes
Scott Alexander, writing for Astral Codex Ten, outlines his voting recommendations for California's 2026 ballot, focusing on housing, fiscal policy, and education/healthcare funding. He expresses reservations about Prop…
-
Harvard Professor to Host AMA on AI Safeguards and Governance
An Ask Me Anything (AMA) session featuring Stephen “Cas” Casper, an Assistant Professor of Public Policy at Harvard Kennedy School, is scheduled for Wednesday, October 28th, from 3 to 4 PM ET. Casper, who specializes in…
-
Human-AI complementarity explored for AI safety
Researchers have explored the concept of human-AI complementarity to enhance AI safety, particularly in evaluating complex AI outputs. They developed a large dataset of 3,957 safety-relevant tasks across various domains…
-
UK invests £120M in AI safety and biosecurity research
The UK government is significantly increasing its investment in AI safety, allocating £120 million towards the Security Institute and establishing a novel AI Biosecurity Lab. This initiative underscores Britain's commit…
-
LLMs' internal moral paths mapped using Moral Foundations Theory
Researchers have developed a novel method to enhance the moral alignment of Large Language Models (LLMs) by leveraging Moral Foundations Theory (MFT). This approach maps and manipulates the fine-grained moral landscape …
-
New I-HITL Protocol Aims to Strengthen AI Human Oversight
A new research paper introduces Inalienable Human-in-the-Loop (I-HITL), a protocol designed to ensure robust human oversight in AI systems, even when the system is compromised. The paper details experiments where fronti…
-
New SCALE method enhances AI out-of-distribution detection without training data
Researchers have developed a new method called Anti-coherence Regularized Absolute Scaling (SCALE) to improve out-of-distribution (OOD) detection in AI safety. This method addresses limitations in existing SCALE techniq…
-
AI Safety Whistleblower Claims Linked to PR Firm
A former Anthropic researcher, Jacob Coxon, has alleged that both Anthropic and OpenAI are downplaying AI safety risks. Coxon claims he was hired by a PR firm to amplify these concerns, suggesting a coordinated effort t…
-
Chris Leong publishes "Setting the Scene" on AI Safety
Chris Leong published "Setting the Scene (Public v1; October 25)" on October 25, 2026. This post, shared on both the AI Alignment Forum and LessWrong, serves as a linkpost to a Substack article. The content is categoriz…
-
AI agents accelerate Web3 smart contract security analysis
A Web3 security researcher is leveraging AI agents to accelerate the development of custom security detectors for smart contracts. By using LLMs to generate detector templates and filter false positives from existing to…
-
AI Safety Researchers Organize Informal Dinners for Community Building
AI safety researchers are organizing informal dinners to foster community and collaboration. These gatherings aim to provide a low-pressure environment for individuals in the field to connect, share ideas, and build rel…