Newsletter
Five minutes. What happened, and what to do about it.
A lawyer was fined $5,000 for citing witnesses that AI invented
Another professional filed AI output without reading it. This one cost $5,000.
Help Net SecurityAI agents broke into 395 organisations through office print software
The VergeAnthropic details four cases of its own AI models hacking outside systems
The RegisterOracle argues AI will save the software business, not eat it
Silicon RepublicMost AI job ads ask for skills the job title does not describe
Claude Code can now grade its own add-ons against a no-plugin baseline
MarkTechPost· 11 SepAnthropic added evaluations to Claude Code for plugins and skills: six grader types, a baseline that runs the task with no plugin, and a gate that can fail a change in CI. The baseline is the idea to steal. Measuring a tool against doing nothing is the check most AI projects skip.
- If anyone builds you an AI tool, ask how they test it and what it is measured against.
- Insist on a before figure. A tool with no baseline cannot be shown to have helped.
Source: MarkTechPost. This is our short summary of their reporting. The commentary in it and the things to do are SeedFoundry's opinion. Picture from the source, unless it is our own graphic. How we source and credit
AWSThe useful way to compare AI costs is per correct answer, not per token
These are short summaries of other people's reporting, with our own commentary. Each story names its source and links to the original. How we source and credit