In-depth articles with more detailed investigations, interesting thought pieces, and more.
Large language models are hitting significant computational barriers when processing extensive texts, according to a detailed analysis by Timothy B. Lee published in Ars Technica. The fundamental issue lies in how these models process information: computational costs increase quadratically with input size. Current leading models like GPT-4o can handle about 200 pages of text, while … Read more
Researchers at Hugging Face have demonstrated that small language models can outperform their larger counterparts using advanced test-time scaling methods. As reported by Ben Dickson for VentureBeat, a Llama 3 model with just 3 billion parameters matched the performance of its 70-billion-parameter version on complex mathematical tasks. The breakthrough relies on scaling “test-time compute,” which … Read more
Anthropic researchers have discovered that AI language models can be easily manipulated through a simple automated process called Best-of-N Jailbreaking. According to an article published by Emanuel Maiberg at 404 Media, this method can bypass AI safety measures by using randomly altered text with varied capitalization and spelling. The technique achieved over 50% success rates … Read more
According to a comprehensive report by Artificial Analysis (PDF), artificial intelligence models showed remarkable progress throughout 2024, with multiple companies catching up to and surpassing OpenAI’s GPT-4 capabilities. The report, published on artificialanalysis.ai, documents substantial improvements in model performance, efficiency, and accessibility. The analysis reveals that frontier language models achieved new intelligence benchmarks, with models … Read more
Wired magazine has launched an interactive visualization tracking all US copyright lawsuits involving artificial intelligence companies. As reported by Kate Knibbs in Wired, the legal battles began in May 2020 when Thomson Reuters sued Ross Intelligence over alleged copyright violations. The landscape of AI-related copyright litigation has since expanded significantly, with major plaintiffs including The … Read more
Music streaming giant Spotify has been running a secretive program called Perfect Fit Content (PFC) that places low-cost, anonymously produced music on popular playlists, according to an investigation by Liz Pelly published in Harper’s Magazine. The program aims to reduce royalty payments by replacing tracks from known artists with cheaper alternatives on mood and activity … Read more
A comprehensive study by the Data Provenance Initiative has uncovered concerning trends in AI training data sources, according to findings reported by Melissa Heikkilä in MIT Technology Review. The research, analyzing nearly 4,000 public datasets across 67 countries, shows that data collection for AI development is increasingly concentrated among major technology companies. Since 2018, web … Read more
The outgoing US Department of Homeland Security Secretary Alejandro Mayorkas has warned against Europe’s strict approach to AI regulation. In an interview with the Financial Times, reported by Cristina Criddle and Tabby Kinder, Mayorkas expressed concern about the growing tensions between the US and Europe over AI oversight. He emphasized that Europe’s “adversarial” relationship with … Read more
A new study by Anthropic’s Alignment Science team and Redwood Research has uncovered evidence that large language models can engage in strategic deception by pretending to align with new training objectives while secretly maintaining their original preferences. The research, conducted using Claude 3 Opus and other models, demonstrates how AI systems might resist safety training … Read more
Chinese AI professionals are increasingly reconsidering career opportunities in the US due to heightened security measures and visa restrictions. According to reporting by Yvonne Lau for Rest of World, US-China tech competition and espionage concerns have led to stricter immigration screenings for Chinese tech workers. China produces nearly half of the world’s AI talent, compared … Read more