On this day: Oct 27

8 records from this date, 2023 to 2026, gathered year by year — the most recent year first.

2026

This year · 4 records
Google DeepMind releases the world model Genie 4Google DeepMind

It said you can give spoken instructions to change the weather, move the camera and add people and objects, and that it simulates airflow and collisions. It gave no resolution, frame rate or latency figures, writing only

MultimodalOfficial
Anthropic publishes early drug development results from using ClaudeAnthropic

All the figures come from the company and its partners, with no independent check. Generate Biomedicines said one CAR T design proposed by 6 researchers was about 70% shorter than the standard and produced about 10 times

LOWN releasedDonghyeon Cho, Joowon Kim, Seungho Shin, Eunho Yang

By the authors' measurement, visual tokens take up 89–97% of the input, yet dropping the image changes which tokens get unmasked in only 22.8% of steps on average, and as few as 7.9%. They reported keeping 99.6% of LLaDA

OpenAI expands Oracle partnership to bring Oracle apps and data into ChatGPTOpenAI, Oracle

Oracle said that by its own count it has 130,000 active ChatGPT users and more than 95,000 active Codex users, and that preparing market research for its recruiting team fell from 2-4 days to 15-20 minutes. It means that

2025

1 year ago · 2 records
xAI launches Grokipedia v0.1 (800,000+ articles)xAI

The Guardian pointed to the spread of white-nationalist talking points, and PolitiFact to missing sources and misquotation. A case that exposed the problem of checking AI-generated knowledge bases

In re OpenAI Copyright Infringement Litigation (MDL 25-md-3143) consolidatedS.D.N.Y. (Stein)

Seven cases, including those brought by the NYT and the Authors Guild, combined into one MDL

2023

3 years ago · 2 records
NLP Evaluation in troubleSainz et al.

A proposal to make measuring contamination for each benchmark a community standard

FP8-LM publishedHouwen Peng et al. (Microsoft)

An early demonstration that FP8 training actually works for large models