Urgent.News

One page, thousands of outlets. See who else covered it.

Editions

AI

OpenAI says the changes to its model training will increase compute overhead by 20% of observed inference workload; the increase will not be handed to customers (Thomas Claburn/The Register)

Expanded multistage chain of thought monitoring makes frontier model work more expensive — OpenAI on Tuesday said its decision …

We haven't written up this one. Techmeme has the full story — the link below goes straight to it.

This story

This is one outlet's version. Read the fullest account.

Read the original at theregister.com →

More in AI

No one should be too old for AI

Every morning, an 85-year-old American woman greets her artificial intelligence (AI) companion, ElliQ. It reminds her to take her medication, encourages her to stay active and helps her remain…

A 2-Token Prompt and a 39,966-Token Bill: Measuring What My Agent Actually Costs

There is a small cluster of posts going around right now about auditing your LLM invoice, and about how cost calculators get the numbers wrong.

  • Author discovered Claude CLI's token discrepancy in gitcommit.py script
  • Default output format omitted crucial token details, leading to inaccurate cost calculations
  • Adding --output-format json flag revealed 39,966 billed input tokens vs. 2 prompt tokens

More from Wednesday 19 August →