AI Disruption

AI Disruption

Home
Podcast
Chat
Chip
Meta
Paper
Qwen
Agent
Robot
Cursor
Google
OpenAI
AI Code
AI Video
Amazon
Microsoft
Windsurf
Anthropic
DeepSeek
AI Weekly
Elon Musk
AI Writing
AI Painting
AI Business
🎈 Guest Posts
AI Open Source
Chinese Outbound
Foundation Model
Archive
About
Claude Code Ralph Loop: AI Debugs Itself in 3 Steps
Claude Ralph Loop auto debugs code until tests pass. Set task, exit marker, max loops and it fixes errors without you watching.
9 hrs ago • Meng Li
Clawdbot: the Mac-mini sell-out that Siri sees as its mirror
Clawdbot open-source AI assistant sparks Mac mini sell-out; self-host your own proactive agent
11 hrs ago • Meng Li
Anthropic Blocks Third Parties, OpenAI Opens OpenCode
OpenCode ban sparks Claude vs OpenAI showdown: how to protect your dev account
Jan 25 • Meng Li
OpenAI at a Crossroads: Besieged by Google and Anthropic
OpenAI pivots hard to enterprise AI as Google and Anthropic erode its consumer dominance
Jan 25 • Meng Li
From Figma to Code: How Pencil Redefines the Design Workflow
Use Pencil MCP free: design in IDE, skip Claude, export clean code in minutes
Jan 23 • Meng Li
OpenAI: We May Take a Cut of Future AI-Driven Profits
OpenAPI considers taking a cut of customer profits, shifting from SaaS fees to IP revenue share
Jan 23 • Meng Li
AI Constitution Born: Claude Nears Self Awakening
Anthropic open sources Claude’s 84-page AI Constitution, teaching ethics over rules and sparking debate on machine sentience
Jan 22 • Meng Li
57.1% Fooled! Runway Gen 4.5 Out
Runway Gen 4.5 blurs reality: 57% can't spot AI video. Test it now
Jan 22 • Meng Li
Claude Code Creator’s Tips I Can’t Get Behind
Claude Code tips from the creator: what I agree with, what I skip, and what actually saves tokens
Jan 21 • Meng Li
R1 Turns One: DeepSeek Model 1 Emerges Quietly
DeepSeek Model1 spotted on GitHub: V4 in disguise? 512-dim MLA, Blackwell boost, sparse MLA hint
Jan 21 • Meng Li
Claude Cowork Now Has Permanent Memory
Claude Cowork now packs permanent memory—knowledge bases remember every project, preference, and lesson so your AI teammate grows smarter each time you…
Jan 20 • Meng Li
Zhipu’s New Model Also Uses DeepSeek’s MLA, Runs on Apple M5
Zhipu IPO follow-up: free 30B MoE GLM-4.7-Flash, 3B active, MLA power, M5 local 43 tok/s
Jan 20 • Meng Li
© 2026 Meng Li · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture