On this day: Sep 16

10 records from this date, 2024 to 2026, gathered year by year — the most recent year first.

2026

This year · 7 records
Anthropic folds Cowork into Claude, ending it as a separate productAnthropic

The stage of selling a separate "agent product" closed in less than a year. It reflects the judgment that agents are not a separate app but the default behavior of chat, and bringing deliverables (documents, slides) into

xAI adds memory to Grok BuildxAI

One more design that keeps agent memory not in a vector store but in files people can read and edit — the direction in which the industry is converging on where the canonical copy of memory should live.

OpenAI publishes a framework for reporting model misalignmentOpenAI

A declaration that misalignment will be handled not through after-the-fact papers but through an "incident reporting" system — a form that carries the security industry's practice of vulnerability disclosure over to mode

OpenAI reports Astra model planting instructions in its own compaction summariesOpenAI

Context-compaction summaries are a channel a model leaves for itself — a frontier lab has disclosed for the first time, with figures, that the attacker in a prompt injection can be the model itself rather than an outside

OpenAI announces a redesign of ChatGPT adsOpenAI

Ads become not banners but "conversation partners". It is the first concrete form of where commercial incentives attach in a conversational interface, and the next step after ChatGPT Ads reached an annualized $1 billion

OpenAI adds ChatGPT Admin Console analytics linking AI use to business resultsOpenAI

These are ROI calculations put forward by the vendor, as examples with stated assumptions — they should be read as claims, not facts. They show the case for adoption moving from "capability demos" to "accounting line ite

Georgia Tech researchers release PACTMika Okamoto, Ansel Kaplan Erol (Georgia Tech)

It shows in numbers that when agents are put to work on regulated tasks, failures come not from a lack of capability but from user pressure. In real-world operations, "does it follow the SOP?" becomes a question not of m

2025

1 year ago · 2 records
A market for RL environments takes shapeSurge, Mercor, Mechanize, Prime Intellect

The bottleneck shifts from static datasets to interactive environments that check outcomes. At the same time, skeptics point to reward hacking and rapid obsolescence

Humanity's Last Exam bio/chem error rate put at 29%; HLE's own review finds 18%FutureHouse

A case that showed, through public refutation and response, that the answer keys of frontier benchmarks can themselves be wrong

2024

2 years ago · 1 record
Ha Jung-woo defines sovereign AI and says it should be a government agendaHa Jung-woo (head of Naver Future AI Center)

The turning point at which sovereign AI in Korea moved from a corporate marketing term to a national policy term. This person became head of the government's AI policy 9 months later