What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

AI Coding and the Bug Selection Problem

Connections

Related to

Agentic Coding: LLM Agents, Testing, and Hallucination Risks Agentic coding and hallucination risks address the reliability of AI-generated code, while the bug selection problem addresses the parallel issue of whether AI-identified bugs actually get fixed — both highlight gaps between AI capability and real-world impact
AI is removing the middle class of software engineering Both examine how AI coding tools reshape software engineering roles and incentives — one focusing on middle-class elimination, the other on how unlimited bug-finding capability is underutilized due to organizational dynamics

Supports

Ironies of Automation The 'Ironies of Automation' describes how automation creates new failure modes and human oversight gaps; the bug selection problem similarly shows that AI capability doesn't translate to improved outcomes due to organizational and human behavioral constraints
Understanding is the new bottleneck 'Understanding is the new bottleneck' argues human comprehension limits AI utility; the bug selection problem is a direct case study where human cognitive and organizational limits prevent leveraging AI's technically superior bug-finding capacity
Why Junior Engineers Still Matter in the AI Era The bug selection problem implies human judgment about which bugs matter remains critical, reinforcing the argument that junior engineers (and human decision-making generally) still have indispensable roles even when AI can technically outperform them

Develops into

8 Myths of Software Development with AI Both directly address myths and misconceptions about AI in software development — the bug selection problem is a concrete illustration of how AI capability does not automatically translate to better software quality

Supported by

Bespoke: A Programming Language for People Who Say Please Bespoke's humorous framing of mutations requiring apologies satirizes real programming pain points (side effects, state mutation) that directly relate to the 'bug selection problem' in AI coding — both highlight how programmer intent and compiler/AI interpretation diverge
SWE-bench Science: Coding Agents for Engineering Tasks SWE-bench Science provides empirical evidence for the 'bug selection problem' thesis — by benchmarking which engineering tasks coding agents can actually resolve, it illuminates the systematic biases in what AI coding tools succeed and fail at
The AI Operating Layer The bug selection problem in AI coding is a governance/oversight challenge that Coder's centralized management platform aims to mitigate through enterprise controls
Six Months Writing Code Exclusively With AI Agents Both address practical realities of AI-assisted coding from a developer perspective; the six-month experience directly illustrates the 'bug selection problem' where AI agents introduce subtle issues that require human judgment
DHH on Omarchy acceleration DHH's 'shallow problems' framing aligns with the bug selection problem — when AI can solve surface-level bugs quickly, the remaining hard problems are ones requiring deeper judgment

Challenged by

Domain-Driven Agents The bug selection problem in AI coding stems partly from poor semantic understanding of codebases; domain-driven agents propose a structured solution to give AI models better context, potentially reducing this failure mode
Autonomous Mode in Kiro Web for Technical Debt Kiro's autonomous PR submission for technical debt directly engages the 'bug selection problem' — autonomous agents must decide which issues to prioritize, a non-trivial choice with significant downstream consequences
SWE-2: Advanced Coding Model with Improved Cost-Capability Tradeoff SWE-2's improved coding capability directly confronts the 'bug selection problem' thesis — as models get better at coding tasks, the nature of bugs AI introduces vs. solves shifts

Slideops: Repository Slide Deck Generator Both address the gap between what code says and what it does — Slideops by detecting doc/implementation drift, and Bug Selection Problem by noting AI tends toward certain bug patterns; both concern code fidelity and trust

Type:	link
First published:	Saturday, 22 August 2026 at 15:44 UTC
Location:	https://nolanlawson.com/2026/08/16/you-can-just-choose-how-many-bugs-you-want-now/
URL:	https://mattwood.fyi/i/f4d01fd9

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections