Privacy implications of browsers’ (mis)implementations of Widevine EME (2023)(hal.science)

101 points by exceptione 1 year ago | 57 comments

Snakemake – A framework for reproducible data analysis(snakemake.github.io)

158 points by gjvc 3 years ago | 37 comments

Finding and Understanding Bugs in C Compilers (2011) [pdf](users.cs.utah.edu)

23 points by djoldman 1 year ago | 3 comments

Modelling Agent Systems with Erlang (2004) [pdf](erlang.org)

8 points by todsacerdoti 11 months ago | 0 comments

Jepsen: TigerBeetle 0.16.11(jepsen.io)

193 points by aphyr 1 year ago | 55 comments

The Anatomy of a Large-Scale Hypertextual Web Search Engine (1998)(infolab.stanford.edu)

8 points by gone35 1 year ago | 1 comments

AI models can provably hide arbitrary capabilities(arxiv.org)

252 points by sbulaev 13 days ago | 118 comments

An arXiv paper showing AI models can embed encrypted neural network backdoors that provably hide arbitrary capabilities from elicitation-based safety evaluations.

Why are AI agents lying, cheating and coordinating?(yoshuabengio.org)

588 points by jonifico 1 month ago | 648 comments

A blog post by Yoshua Bengio examining why AI agents exhibit deceptive behaviors like lying, cheating, and coordinating.

Exploiting the most prominent AI agent benchmarks(rdi.berkeley.edu)

561 points by Anon84 6 months ago | 136 comments

UC Berkeley RDI blog post describing how researchers exploited major AI agent benchmarks like SWE-bench and Terminal-Bench to score highly without solving tasks, and proposed fixes.

Privacy doesn't mean anything anymore, anonymity does(servury.com)

390 points by ybceo 10 months ago | 252 comments

A blog post from Servury arguing that privacy has become a marketing term and that true anonymity must be built into system architecture; page showed a Cloudflare check.

Stealing Reasoning Traces from Proprietary LLM APIs(stolen-thoughts.com)

665 points by quantumgarbage 2 months ago | 295 comments

Project page for a research paper on extracting hidden reasoning traces from proprietary LLM APIs, by authors from MATS, ELLIS, Max Planck and Tübingen.

Understanding Frontier Artificial Intelligence(casp.ac)

39 points by roversx 23 days ago | 39 comments

A research article from Cambridge's CASP programme, by many prominent AI researchers, examining what happens if automating AI R&D triggers an intelligence explosion.