Pragmatic Anthropomorphism, or: How to Talk to an Autocompleting Cricket - blog.lmorchard.com

kept by solacepaper sep 28

Treating LLMs as collaborative partners via pragmatic anthropomorphism improves their utility, not because they understand, but because it steers their latent space toward useful outputs.

Auto mode for Claude Code
Claude Code's new auto mode uses a Claude Sonnet 4.6 classifier to grant permissions based on intent, with configurable default allow and soft-deny filters.
also on Claude Sonnet, Anthropic, #ai-agents, #ai-coding, #developer-tools
Ask HN: What default model do you use and why?
Hacker News users share and justify their preferred default AI models, citing factors like reasoning quality, cost, privacy, and task suitability.
also on Claude Sonnet, Anthropic, #ai, #developer-tools
You have to beat the models at something
Engineers must leverage deep codebase familiarity and clear technical communication to add value beyond what LLMs can reliably produce today.
also on GPT-4, Anthropic, #ai-coding, #developer-tools
Anthropic/OpenAI may be spending more than $1000 for every $100 you pay them – R&A IT Strategy & Architecture
Coding with LLMs like Claude Code may cost providers $1000+ for every $100 in subscription revenue, making true agentic coding economically unsustainable.
also on GPT-4, Anthropic, #ai-coding, #developer-tools