Thoughts on Role Confusion
Researchers Charles Ye, Jasmine Cui, and Dylan Hadfield-Menell found that large language models often ignore explicit role tags like <system> or <user> and instead infer roles from text tone, enabling…
OpenAI is an American AI research laboratory founded in 2015, known for creating ChatGPT, the GPT series of large language models, and DALL-E image generators. It is one of the most influential AI companies in the world.
Researchers Charles Ye, Jasmine Cui, and Dylan Hadfield-Menell found that large language models often ignore explicit role tags like <system> or <user> and instead infer roles from text tone, enabling…
Polymarket odds for Claude Fable 5 restoration have rebounded to 60% by July 1 and 88% by July 31 after code hints and an Amazon Bedrock reappearance, though the update may be overconfident. The incid…
OpenAI and Broadcom unveiled Jalapeño, their first jointly-made AI chip designed to run large language models, claiming performance per watt substantially better than current state-of-the-art. The chi…
OpenAI and Broadcom unveiled Jalapeno, OpenAI's first custom AI chip, designed for inference tasks in ChatGPT and other services. The ASIC chip, developed in nine months, aims to reduce costs and impr…
OpenAI updated GPT-5.5 Instant, its most-used ChatGPT model, to better understand user intent and improve conversation quality. The model now identifies underlying goals, adapts to user feedback, and …
Developers are increasingly using large language models to encode business logic in Markdown prompts, a practice dubbed 'programming in Markdown,' despite the availability of simpler, faster, and more…
Stanford graduates expressed mixed feelings about artificial intelligence during commencement, with some walking out on Google CEO Sundar Pichai's speech. Students cited concerns over ethical developm…
Qualcomm announced a major push into data center infrastructure at its 2026 Investor Day, unveiling a new Dragonfly C1000 CPU with 250+ cores and a High Bandwidth Compute (HBC) architecture that stack…
OpenAI updated ChatGPT's default model to GPT-5.5 Instant in May 2026, replacing GPT-5.3 Instant for all users. The update delivers 52.5% fewer hallucinated claims and 37.3% fewer inaccurate answers, …
OpenAI has taped out its first custom inference chip, named Jalapeño, built with Broadcom. The ASIC is designed to reduce inference costs for serving LLMs, following the vertical-integration strategy …
Shrijith Venkatramana, building git-lrc, explains how scaling laws discovered by OpenAI, Google, DeepMind, and Anthropic made large language models like ChatGPT, Claude, and Gemini possible. The 2020 …
Roughly $27 million was spent in a proxy contest over Assemblyman Alex Bores in New York's 12th Congressional District, with AI-linked super PACs playing a decisive role. AI-centric super PACs tied to…
The Trump administration is pressuring Meta to join a federal program that reviews advanced AI models before public release, citing national security concerns. Meta is the only major US AI developer n…
The University of South Carolina has expanded its deal with OpenAI to $3 million, up from an initial $1.5 million, after demand for ChatGPT Edu access exceeded expectations. The agreement includes ima…
OpenAI unveiled Jalapeño, its first in-house AI chip developed with Broadcom, to reduce reliance on Nvidia and lower costs. The application-specific integrated circuit (ASIC) aims to accelerate OpenAI…
A new AI news aggregator called AI Frontier, hosted at smol.ai, curates top AI developments including a thesis on the 'harness era' beyond the model, OpenAI's GPT-5.5-Cyber for closed-loop patch gener…
Lelu, an authorization engine for AI agents, detects prompt injection and low-confidence decisions to gate agent actions, offering four outcomes: allow, human review, compute (sandbox), or deny. The o…
A GeoScan analysis of 100 leading B2B brands found an average GEO Readiness score of 47 out of 100, with no brand fully optimized for AI search engines. The study reveals that many brands rely on trai…
Two Google AI researchers, Jonas Adler and Alexander Pritzel, are leaving to join Anthropic, following Nobel laureate John Jumper and Gemini co-lead Noam Shazeer. The departures, which wiped $270 bill…
Google is set to lose two more AI researchers to Anthropic, a startup founded by former OpenAI employees. The departures highlight ongoing talent migration from big tech to AI startups.…