ai-development

This thread explains why AI builders should use loops instead of single prompts, emphasizing proper triggers, verification, and stop conditions to build reliable, cost-effective AI systems.

A reflective piece asking what recent AI developments would have seemed most unbelievable in 2020, and what future surprises might await.

The author built a 3D spaceship flying game in Lovable using just one prompt, highlighting how AI simplifies game development.

This article analyzes the reasons behind the performance leap of Zhipu GLM-5.2, suggesting that its 40B activation parameters provide greater effective capacity after accounting for fixed overhead, making RL post-training more effective. It also reviews the history of Chinese AI model development and notes that the large model approach ultimately prevailed.

A developer shares how they replaced a $200/month SaaS with a custom client portal built in under a day using AI tools like Claude and Lovable, highlighting the growing trend of building software instead of buying it.

YoAmigo Studio is a tool that lets you build and ship real applications using AI subscriptions you already have, simplifying app development with existing AI investments.

The author describes setting up an AI dev platform in his homelab using OpenCode Web UI with Git access, enabling AI-assisted maintenance of Docker services via PR review and GitOps deployment.

Matt Pocock shares his seven phases of AI-powered development, emphasizing the need for more structure in the pre-PRD phase.

WeChat released the developer access guide for Mini Program AI development mode (beta), allowing developers to encapsulate Mini Program functions as SKILLs via MCP protocol for AI invocation. It is currently in internal testing.

This article discusses how China has rapidly advanced in AI despite being a latecomer, questioning the sources of datasets, computing power, and algorithms that enabled companies like DeepSeek to catch up with US leaders like OpenAI and Google.

Channel AI founder Luke Orthwine proposes a new software development methodology: shifting programming thinking from traditional chess-like single-threaded linear thinking to real-time strategy game (RTS) style high concurrency, macro scheduling, and saturation attack to achieve efficient development in the AI Agent era.

Recommends using Claude AI along with a set of free/low-cost services (Supabase, Vercel, Stripe, etc.) as a full-stack tech stack for solo businesses or indie development, covering development, deployment, payments, email, authentication, monitoring, and more.

A comprehensive list of core team members from three AI Agent development teams – OpenAI Codex, Claude Code, and Manus – covering leadership, API, ecosystem, security, and more, for readers interested in Agent development trends.

An overview of the current state and future outlook of continual learning in mid-2026, covering memory approaches including external memory, in-state memory, and weight updates, with analysis of various models like TTT, Titans, and Dragon Hatchling.

WeChat officially released an AI development mode auxiliary toolset that can automatically convert Mini Program source code into the SKILL format required by WeChat AI development mode, and provides validation and evaluation functions to improve development efficiency.

The post discusses how falling barriers to entry in AI development, through free AI agent builders, open-source models, and no-code tools, are enabling solo founders to launch products faster than ever before.

Deep dive into the Codex full-stack development platform, showcasing multiple core features such as Annotate, Fork, AGENTS.md, and Computer Use, enabling AI to become a 24/7 development assistant.

Lovable, a vibe coding startup, has surpassed $500 million in annualized revenue and is generating one million new projects weekly, highlighting the rapid adoption of AI-powered software creation tools by non-technical users.

Summary of an internal conversation between Claude Code founder Boris and team members, sharing experiences on using agent, auto mode, and simplifying context to let the model freely explore solutions, emphasizing the efficient development approach of interacting with the agent via voice on a mobile phone.

This article explores the view that in the Agent era, Loop Engineering is more important than Prompt Engineering. The author believes that the core capability of an AI Agent lies not in the model itself, but in the feedback loop system built around the model, which determines whether the Agent can continuously improve and approach the correct answer.