AI that actually runs in your business
Laojin Global is an operator-built family of AI tools and SaaS: every product runs in our own business before it ships to you. We help people and teams put AI to work — for real, in production.
- 187
- interview guides
- 132
- people tracked
- 259
- practical articles
- 6
- in-house tools
Frontier Daily
Laojin pulls the official channels every day and compiles the past 48 hours of AI news into a short brief — curation only, every item linked to the original source.
10月5日
Daily
4 itemsFrontier Daily (Oct 5): 4 items from Hacker News, Simon Willison
A project runs the 125B-parameter Qwen 3.8 Flash Next on consumer RTX 4090 hardware at 100T/s, drawing 421 upvotes and 224 comments on Hacke…
Frontier Daily (Oct 4): 5 items from Hacker News, Latent Space, Simon Willison
A writer says they quit OpenAI because its culture is broken, drawing 86 upvotes and 137 comments on Hacker News.
Frontier Daily (Oct 3): 7 items from Hacker News, Anthropic, OpenAI and more
The creator of Redis released ds4 for running LLMs locally, drawing 57 upvotes and 5 comments on Hacker News.
Frontier Daily (Oct 2): 10 items from OpenAI, Hacker News, One Useful Thing and more
OpenAI argues that advanced AI may matter most for the routine work behind breakthrough ideas, with execution shaping the next economy and t…
✦ Sources：OpenAI · Anthropic · Google DeepMind · Qwen · Simon Willison · Latent Space · Import AI · One Useful Thing
AI People Heatmap
Who's actually moving AI this week? Scored from media, deep interviews, GitHub and arXiv signals. Free to cite.
channel-acquired
The Home Depot: The best-performing stock in the S&P 500 since IPO (Audio)
🔥 78channel-tim-ferriss
He Started a Psychedelics Company. It Sold for $700M.
🔥 78channel-latent-space
Which GPU Clouds Are Actually Good? | ClusterMAX 3.0
🔥 78channel-mlst
How a Voice Agent Learns the Rhythm of Conversation — Shawn Wen
🔥 78channel-lenny
Head of ChatGPT on Dots, ambient AI, and why most actions on the internet will soon be taken by AI
🔥 78Free to cite · CC BY 4.0
Classics · Latest deep interviews
AI talks and conversations worth rewatching, with Chinese summaries and key points — 3 hours in 10 minutes.
Transformers Can't Think In Physics: Anima Anandkumar on Accelerated Understanding
In this episode, Caltech professor Anima Anandkumar and researcher Benedikt Jenik dive into the gap between large language models and physical world modeling. They argue that current transformers excel at symbolic reasoning but lack an intrinsic grasp of causality, time, and space continuity. The conversation focuses on combining neural operators, generative simulation, and scientific knowledge to build AI systems that truly accelerate discovery—an essential watch for anyone in AI for Science.
Full summary & key pointsOne product, one living page
Tutorials, official links, news and FAQ on one URL we can keep current. Starting with DeepSeek Harness.
DeepSeek Harness
What dsh is, how to install it, and how it sits next to Claude Code and Cursor. Official links, tutorials and news.
Open hubMinimal terminal coding harnessPi Agent
Mario Zechner's minimal open-source coding agent: 4 base tools, TS extensions, skills & SDK embedding.
Open hubOpen multi-surface coding agentOpenCode
Terminal TUI + desktop + IDE open coding agent across any LLM, MIT.
Open hubPersonal AI assistantOpenClaw 🦞
A personal AI assistant on your devices, connecting WhatsApp/Telegram/Slack via a local Gateway.
Open hubSelf-improving autonomous agentHermes Agent ☤
Nous Research's self-improving agent: learning loop, persistent memory, any model.
Open hubTerminal coding agent ecosystemClaude Code Ecosystem
Anthropic's official terminal agent (closed), with a largely open Skills/hooks/MCP ecosystem — the benchmark.
Open hubAI tools, organized by use case
196 hand-verified AI and going-global tools, categorized — stop gambling on search results.
Browse by topic
Interviews, tutorials and picks across the same topic, aggregated.
LLMs
LLM capabilities, products and best practices
Agents
Building, orchestrating and shipping AI agents
Open Source
Open models, frameworks and ecosystems
Infra
Compute, inference engines and serving
Reasoning
Model reasoning, chain-of-thought and deep thinking
AI Safety
Alignment, bias and responsible AI
Computer Vision
Visual understanding, generation and robotics
Multimodal
Understanding across text, image, audio and video
Reinforcement Learning
RL, reward models and self-play
Engineering
APIs, prompting and system design
AI implementation, with receipts
Front-line AI workflows and hard-won lessons, written up as articles. No concepts — only plays that ran.
Frontier Daily (Oct 5): 4 items from Hacker News, Simon Willison
A project runs the 125B-parameter Qwen 3.8 Flash Next on consumer RTX 4090 hardware at 100T/s, drawing 421 upvotes and 224 comments on Hacke…
Frontier Daily (Oct 4): 5 items from Hacker News, Latent Space, Simon Willison
A writer says they quit OpenAI because its culture is broken, drawing 86 upvotes and 137 comments on Hacker News.
Frontier Daily (Oct 3): 7 items from Hacker News, Anthropic, OpenAI and more
The creator of Redis released ds4 for running LLMs locally, drawing 57 upvotes and 5 comments on Hacker News.
The 365 product family
Sales, growth, developer workflows — the key links of putting AI into production. Open any product to see how it runs in our own business.
Sellenca
The AI sales copilot inside WhatsApp Web
Learn more365AIOrg
Cross-platform AI agent orchestration
Learn moreAllModelsAPI
One key for many models
Learn more365Loopa
AI growth operations, driven by evidence
Learn more365 Ops
An AI operations system for Shopify (internal)
Learn more365Skill
An agent-skills lab: 14 in-house skills
Learn more365Skill: everything we build is open
discover-project-skills · five-step-dev · gws-workspace · shopify-theme-delivery — every skill runs in our own business first.
Want a hand putting AI into production?
Training, consulting and custom deployment — built on the products and workflows we run ourselves.