#AI编程
9716 related articles
A Surgeon Built His Own Tools with Codex: How AI Is Reshaping Medical Literature Research
Surgeon @pridgenmd shares how he used OpenAI Codex to build custom research tools for PubMed literature review — a case study in AI turning domain experts into citizen developers.
OpenAI's API Strategy: Best Models at Every Price Point and Full Modality Coverage
OpenAI aims to offer the best model at every price point and lead across all modalities — text, code, images, and video — through its API. Here's what this platform strategy means for developers.
GPT-6 Sol and Luna Leaked: Smarter, Better at Coding — and Half the Price
GPT-6 Sol and Luna leaked: major upgrades in intelligence, alignment, coding, and computer use over the 5.6 series — with per-token pricing cut in half.
Microsoft's Abandoned FoxPro Rises Again: Runtime Rewritten in Rust, Compiled to WASM
An open-source project rewrites Visual FoxPro's runtime in Rust and compiles it to WASM, reviving the language Microsoft abandoned in 2007 — with no 2GB table limit, legacy .fll plugin support, lambdas, JSON, and HTTP server built in.
Opportunity AI: 10 New Work Possibilities Unlocked by GPT-6 Astra
What new capabilities does GPT-6 Astra unlock? Explore the Efficiency AI vs. Opportunity AI framework plus 10+ concrete thought-starters: marketing gamification, video pipelines, interactive proposals, decision simulators, 3D apps, and more.
7 Shifts in How We Interact with AI: From Prompt Engineering to Shared Agents
From prompt engineering to shared agents, discover 7 key shifts reshaping how we interact with AI — including mono thread, voice, loop engineering, and multi-model usage.
GPT-6 Astra in Action: One Vague Prompt, Full Feature Built and Tested
OpenAI GPT-6 Astra in Ramp's real-world test: one vague prompt, 12 min to build, 15 min to self-verify via Computer Use — even testing on mobile unprompted.
How Jump Trading Uses GPT-6 Astra to Tackle Its Most Ambiguous and Complex Tasks
Jump Trading reveals they use GPT-6 Astra for their most complex tasks. From code snippets to full services to multi-day multi-source analysis, AI Agents are reshaping enterprise workflows.
Notion Gets Early Access to GPT-6 Astra: From AI Assistant to 'Staff Engineer Collaborator'
Notion engineer shares early GPT-6 Astra experience in Codex: better communication, connects the dots on missed cache optimization patterns. AI shifts from task tool to staff engineer collaborator.
Chinese Open-Source Models Downloaded Twice as Often as American Ones: A Breakdown of Lambert's Congressional Testimony
Nathan Lambert's congressional testimony: Chinese open-weight model downloads are 2× U.S. levels, OpenRouter open-source traffic exceeds 80%, and Qwen, GLM, Kimi dominate agentic coding.
What Is an FDE (Forward Deployed Engineer)? A Deep Dive into AI's Newest Role
FDE (Forward Deployed Engineer) is an emerging AI role where engineers embed on-site at client businesses to build custom AI solutions. Learn what FDEs do and why they matter.
AIBuildAI-2.5: Breaking the Efficiency Bottleneck in Autonomous Modeling with LLM-Guided Tree Search
AIBuildAI-2.5 uses LLM-guided tree search, resource-aware scheduling, and cost-aware model routing to rank #1 on MLE-Bench with a 73.3% medal rate.
Training a Language Model in Pure Rust: The Framework Pitfalls Behind a $164 Experiment
An independent researcher pretrained a 0.4B Bangla language model in pure Rust for $164, documenting silent bugs in Candle and Burn frameworks, gradient validation methods, and tokenizer fertility traps.
San Francisco Agentic Engineering Meetup: Embracing the Weirdness of Coding Agents
Simon Willison and Jesse Vincent host an informal SF meetup for coding agent developers, celebrating weird experiments and unfinished agentic engineering projects.
Complete Beginner's Guide to Claude Code: Build a Real App in an Afternoon — No Coding Experience Required
How can non-programmers build real apps with Claude Code in an afternoon? A complete breakdown of the BITE framework, MVP thinking, model delegation, and building Orbit from scratch.
What Is Claude Code? A New AI Programming Tool for Beginners
What is Claude Code? This guide explains how this AI programming tool helps beginners get started fast, outlines the path from basics to enterprise Agents, and offers a rational take on bootcamp-style courses.
Vibe Coding: A Beginner's Guide from Idea to Product
Vibe Coding lets AI write the code while you focus on ideas. Learn the core concepts, who it's for, Claude Code vs Codex, and a real case study of a product built in an afternoon.
DeepAgents in Practice: Unpacking the Execution Mechanism Behind Skills Scripts in Claude Code
Why is Claude Code getting better? The answer lies in its Skills ecosystem. This article explores how Skills scripts must run in isolated sandboxes and how enterprises can build secure custom agents.
AI Coding for Beginners: A Complete Guide to Cursor and Its Alternatives
A beginner's guide to AI coding tools: explore Cursor's natural language programming, context-aware Q&A, and smart autocomplete — plus comparisons with Trae and Windsurf.
The Secret Behind 2000 PRs a Month: Lauren's AI Agent Trust Workflow
SpaceX AI engineer Lauren shares how she ships 2000+ PRs/month: not better prompts, but a trust-based Agent workflow built on verification skills, Agent-friendly codebases, static analysis, and the Dune framework.