Monday, 16 November 2026 · today’s edition
English
the bservatory

We read everything, so you don’t have to.

Last 24 hours: 3,024 items read from 696 sources · 150 new notes in the Book

AI in Management

AI for people who run organizations: the day’s news, and what experts say in podcasts and talks.

In this group

Models / Anthropic releases Claude Fable 5.2, says it beats Opus 5.5 on most measures but recommends Opus for most tasks

Anthropic released Claude Fable 5.2 on November 9 and said it beats Claude Opus 5.5 in almost every capability it measured, while recommending Opus as the starting point for most coding and knowledge tasks. Anthropic reports Fable scored 49.1% on FrontierCode 2.0 at Xhigh effort, against 45.5% for Opus at Max, and 69.9% against 62.3% on FrontierSWE v2. Fable's standard API rates are double Opus's, according to The New Stack.

The New Stack

Daily AI News

Video · AI daily, 16-Nov-26

AI daily, 16-Nov-26

Podcasts and talks about AI, 15-Nov-26

Agentic Software Engineering

How people build software with agents: the models, the tools and the research.

In this group

AI Science Corner

Research papers of the week on AI and the fields around it, in plain English.

In this group

Teaching robots spatial understanding as a lasting skill, not a one-off

Researchers built a robot control model that learns about three-dimensional space from a wide mix of robot experience, not just from the task it is later trained on. They combined many different robot datasets and kept the spatial knowledge in place during later fine-tuning. In real-world tests, the robots handled grabs and placements when objects or the table moved.

arXiv · AI research

Ethiopia

Politics, security, the economy, banks. Ministries and media read in Amharic and English.

In this group

Get these issues by email or Telegram, at the hour you choose.

Free. One link to change or stop at any time.

Subscribe