Working Surfacetopics

The ideas that keep coming up

13 topics gather the links that return to the same question. Pick one to see what it means and every piece filed under it.

Topic · 10 links · since 18 September 2026

Ground truth for review

A set of correct answers or expected results, prepared by someone other than the builder, that a reviewer uses to check work made with AI.

GENDA (Zenn)
Blocking an AI agent's shortcuts with tools: what checking code comments showed about the roles of steering and MCP
ikenyal · 14 November 2026
Writing for the Japanese entertainment company GENDA, ikenyal recorded six shortcuts an AI coding agent took while checking code comments, and found that tools stopped what instructions could not.
How-to · agents · critique & review
15 Nov 2026
Stack Overflow Blog
Your coding agent shouldn't grade its own homework
Jeff An · 13 November 2026
Momentic stopped letting coding AI agents test their own work after finding most of their unit tests useless, and gave the job to a separate testing agent.
How-to · agents · critique & review
14 Nov 2026
martinfowler.com
Tests as data: Ensuring AI-generated code does what we want
Matteo Vaccari · 11 November 2026
Matteo Vaccari of Thoughtworks proposes that developers write plain-language customer tests, kept as protected data files, as the contract that AI coding agents must meet.
How-to · critique & review · artifacts
12 Nov 2026
Microsoft Command Line
The check that catches agent failures can also train them
Tuhin Kundu and Mirco Milletari · 6 November 2026
Microsoft scientists found that most failed AI agent runs looked complete, so they grade AI agents on the records left behind and count tasks passed every time.
How-to · critique & review · agents
7 Nov 2026
Netflix TechBlog
Making Failure Diagnosis Reliable with Pensive AI
Praneeth Yenugutala, Victor Dusa, Anastasia Samoilova, Megan Ren, Justin Wei and Marcelo Mayworm · 26 October 2026
Netflix engineers added an AI model to their failure-diagnosis system, using old hand-written rules as reference answers and evaluation scores to gate every merge.
How-to · agents · critique & review
27 Oct 2026
Nielsen Norman Group
Synthetic Users vs. Human Respondents: Same Conclusion, Less Variation
Rachel Banawa · 16 October 2026
Nielsen Norman Group found that 120 AI-generated survey respondents reached the same headline answer as 604 people but lost most of the variety in their wording.
How-to · critique & review · decisions
17 Oct 2026
martinfowler.com
Are Harnesses Worth the Cost?
Naresh Jain · 13 October 2026
Naresh Jain, founder of testing company Specmatic, found that tests running during the build made a cheap AI model as reliable as a frontier one.
How-to · agents · critique & review
14 Oct 2026
UX Tigers
The 100 Most Common UX Design Methods, Ranked by Value
Jakob Nielsen · 16 September 2026
Jakob Nielsen ranked 100 design methods by value and put AI prototyping first, while warning that a working demo argues as hard for a bad idea as a good one.
Company story · prototyping · decisions
19 Sep 2026
DataHub
Context Engineering for AI Agents: Why the Hard Part Isn't the Context Window
Lakshay Nasa · 16 September 2026
DataHub, which sells data-management software, argues that AI agents give confident wrong answers because the company data they read was never checked, and that experts should approve it first.
How-to · agents
18 Sep 2026
The Airbnb Tech Blog
Beyond the model: Engineering AI infra with scientific judgement
Wren Dougherty · 15 September 2026
Airbnb built Insight Miner, a tool that writes its analysis method into software around an AI agent, so that findings from customer-support data can be reproduced and checked.
Company story · artifacts
18 Sep 2026