martinald's submissions

1.		Reducing codebase cognitive debt through quizzes? (martinalderson.com)
		2 points by martinald 14 days ago \| past \| 1 comment
2.		Watch out for cache read costs (martinalderson.com)
		4 points by martinald 35 days ago \| past
3.		I'm (mostly) picking models on speed now, not intelligence (martinalderson.com)
		1 point by martinald 43 days ago \| past
4.		The first known runaway AI agent – or a bad marketing stunt? (martinalderson.com)
		3 points by martinald 54 days ago \| past \| 1 comment
5.		Winners and losers in the coming AI margin collapse (martinalderson.com)
		2 points by martinald 60 days ago \| past
6.		GLM 5.2 and the coming AI margin collapse (martinalderson.com)
		694 points by martinald 70 days ago \| past \| 469 comments
7.		Expert-aware quantisation: near-Q4 quality at near-Q2 size? (martinalderson.com)
		3 points by martinald 83 days ago \| past
8.		xAI is looking more like a datacentre REIT than a frontier lab (martinalderson.com)
		692 points by martinald 3 months ago \| past \| 552 comments
9.		Is datacentre sovereignty that important? (martinalderson.com)
		2 points by martinald 3 months ago \| past
10.		Managed agents are the new Lambda (martinalderson.com)
		1 point by martinald 4 months ago \| past
11.		New Claude Code programmatic usage restrictions (twitter.com/i)
		52 points by martinald 4 months ago \| past \| 42 comments
12.		Local LLM Speed Calculator (martinalderson.com)
		2 points by martinald 4 months ago \| past
13.		Open weights are quietly closing up – and that's a problem (martinalderson.com)
		5 points by martinald 4 months ago \| past \| 1 comment
14.		29th August 2026: A Scenario (martinalderson.com)
		3 points by martinald 4 months ago \| past
15.		Map of track defects across the Spanish rail network (vatard.com)
		5 points by martinald 4 months ago \| past
16.		Figma's woes compound with Claude Design (martinalderson.com)
		122 points by martinald 4 months ago \| past \| 98 comments
17.		UK total wind generation record beaten today (robinhawkes.com)
		57 points by martinald 5 months ago \| past \| 33 comments
18.		Using agents and Wine to move off Windows (martinalderson.com)
		2 points by martinald 6 months ago \| past
19.		Why Claude's new 1M context length is a big deal (martinalderson.com)
		4 points by martinald 6 months ago \| past
20.		Is the AI Compute Crunch Here? (martinalderson.com)
		2 points by martinald 6 months ago \| past
21.		Why on-device agentic AI can't keep up (martinalderson.com)
		2 points by martinald 6 months ago \| past
22.		Using OpenCode in CI/CD for AI pull request reviews (martinalderson.com)
		2 points by martinald 6 months ago \| past
23.		Which web frameworks are most token-efficient for AI agents? (martinalderson.com)
		3 points by martinald 6 months ago \| past
24.		Who fixes the zero-days AI finds in abandoned software? (martinalderson.com)
		1 point by martinald 6 months ago \| past
25.		Anthropic's 500 vulns are the tip of the iceberg (martinalderson.com)
		2 points by martinald 6 months ago \| past \| 1 comment
26.		Attack of the SaaS Clones (martinalderson.com)
		2 points by martinald 7 months ago \| past \| 1 comment
27.		Self-Improving Claude.md Files (martinalderson.com)
		2 points by martinald 7 months ago \| past
28.		Two kinds of AI users are emerging (martinalderson.com)
		355 points by martinald 7 months ago \| past \| 341 comments
29.		Turns out I was wrong about TDD (martinalderson.com)
		3 points by martinald 7 months ago \| past
30.		Why sandboxing coding agents is harder than you think (martinalderson.com)
		2 points by martinald 7 months ago \| past \| 3 comments
		More