2025-09-09 Tue

Linked references — 3 pages

2025-09-08 Mon

AI/Copyright Issue
- #Anthropic #News
  - Anthropic to pay $1.5 billion to authors in landmark AI settlement | The Verge
    - The amount breaks down to smaller payouts expected to be approximately $3,000 per book or work. Lawyers for the plaintiffs said it’s “believed to be the largest publicly reported recovery in the history of US copyright litigation.”
    - the final amount could be higher, in that approximately 500,000 works will likely be paid out, but if the total is higher than that, Anthropic will pay an additional $3,000 per work, and it all depends on the number of claims submitted.

It turns out there are. In particular, check out:

(tools with higher limits listed first)

Tool	Pro‑grade models	Free tier limit	Credit card
Qwen Code	Qwen3-Coder-480B	2,000 requests/day	No
Rovo Dev CLI	Claude Sonnet 4	5M tokens/day (beta)	No
Gemini CLI	Gemini 3 Pro, Gemini 2.5 Pro	Gemini 3 Pro (waitlist/paid), 100 req/day Gemini 2.5 Pro	No
Cursor	GPT-5.1-Codex-Max	Free until Dec 11, 2025 (77.9% SWE-bench)	No
Kilo Code	Claude Opus/Sonnet, Gemini 2.5 Pro, GPT‑4.1	Up to $25 signup credits (one‑time)	Yes
Warp	GPT‑5, Claude Opus 4.1, Claude Sonnet 4, Gemini 2.5 Pro	150 credits/month (first 2 months), then 75/month	No
Trae	Claude 4 Sonnet (Beta), Claude 3.7 Sonnet, GPT‑4.1, GPT‑4o, Gemini 2.5 Pro	10 fast + 50 slow requests/month	No
Amazon Q Developer	Claude Sonnet 4	50 agentic requests/month	Yes
GitHub Copilot	GPT‑4.1, Claude Opus 3.5, Gemini 2.0 Flash, Grok Code Fast 1	50 chat requests + 2,000 completions/month	No
Windsurf	OpenAI, Anthropic, Google, xAI	25 credits/month	Yes
Jules	Gemini 2.5 Pro	15 tasks/day	No
AWS Kiro	Claude 4 Sonnet, Claude 3.7 Sonnet	50 credits/month	No
Qoder	Qwen3-Coder-480B, Claude, GPT, Gemini	Free tier + 2-week Pro trial (1,000 credits)	No

https://github.com/inmve/free-ai-coding#qualifying-prograde-models

Only models achieving >60% on SWE-bench Verified qualify as pro-grade for real-world coding tasks. Below is the current list

Person/inmve
- Person/inmve/coding-with-ai.dev
- GitHub - inmve/free-ai-coding: AI coding tools that give free Claude Opus/Sonnet, GPT-5, Gemini Pro, and other pro-grade models

Person Pages Created
- Person/Orta Therox - Developer known for React Native and TypeScript work
- Person/Indragie Karunaratne - macOS app developer who built apps with Claude Code
- Person/Armin Ronacher - Creator of Flask web framework
- Person/Chris Dzombak - Software engineer and developer

Linked references — 1 page

Linked references — 1 page

AI/Coding
- Thread: Master Claude Code: Proven Daily Workflows from 3 Technical Founders (Real Examples) - YouTube
  - Slides: 2025-07-18 - Claude Code talk at Foundations - YouTube Video Shared Links - Google Drive
  - Some great gems in here. I suppose I took away about a dozen things. Still need to type up my takeaways and point them here.
AI
- #Thought
  - AI is a "plausibility engine" - it generates things that are "plausible." Many things that are plausibly described are correlated with reality or real truth, but not all of them, and not uniformly.
  - A lot of people haven't figured out how to make this property of plausibility useful to them. They think they understand how to use an AI that's driven by a "truth engine," but they haven't yet taught themselves to find utility in an AI that's a "plausibility engine."
  - Those who do learn to drive the plausibility-engine e-skateboard will go many places, some of which are much farther away than they could have visited with their own feet. Those who sit around complaining about how cool it would be if we had a flying "truth engine" e-skateboard will likely not get very far powered by AI - that's just not the AI we have right now.
Audio
- #Filed
  - Ableton/Arrangement View/Track/Link
    - Ableton/YouTube/21/04/Learn Live Linked-Track Editing