第61期 | Cybersecurity in the Intelligence Age
今日摘要
OpenAI Blog：Introducing Advanced Account Security: phishing-resistant login, stronger recovery, and enhanced protections to safeguard sensitiv…
OpenAI Blog：How goblin outputs spread in AI models: timeline, root cause, and fixes behind personality-driven quirks in GPT-5 behavior.
OpenAI Blog：OpenAI scales Stargate to build the compute infrastructure powering AGI, adding new data center capacity to meet growing AI demand…
GitHub openai：openai/realtime-voice-component recently updated repository.
GitHub openai：Code for the paper "Jukebox: A Generative Model for Music"
总结 + 观点：Quick illustration of how one can easily read bo…｜中文观点：karpathy/reader3 更值得从实际采用价值来判断，而不是只看它有没有制造新的讨…
总结 + 观点：A collection of projects designed to help develo…｜中文观点：如果 anthropics/claude-quickstarts 能减少集成成本、维护负担…
总结 + 观点：OpenAI outlines a five-part action plan for stre…｜中文观点：从 Cybersecurity in the Intelligence Age 看，后续更…
总结 + 观点：anthropics/claude-agent-sdk-typescript recently…｜中文观点：对 anthropics/claude-agent-sdk-typescript 来说，更…
总结 + 观点：Robust Speech Recognition via Large-Scale Weak S…｜中文观点：openai/whisper 更值得从实际采用价值来判断，而不是只看它有没有制造新的讨论热…
Introducing Advanced Account Security
标签：#ai_engineering_blogs #core
作者：
原文：Introducing Advanced Account Security: phishing-resistant login, stronger recovery, and enhanced protections to safeguard sensitive data and prevent account takeover.
Where the goblins came from
标签：#ai_engineering_blogs #core
作者：
原文：How goblin outputs spread in AI models: timeline, root cause, and fixes behind personality-driven quirks in GPT-5 behavior.
Building the compute infrastructure for the Intelligence Age
标签：#ai_engineering_blogs #core
作者：
原文：OpenAI scales Stargate to build the compute infrastructure powering AGI, adding new data center capacity to meet growing AI demand.
链接：https://openai.com/index/building-the-compute-infrastructure-for-the-intelligence-age
openai/realtime-voice-component
标签：#github_orgs #extended
作者：
原文：openai/realtime-voice-component recently updated repository.
openai/jukebox
标签：#github_orgs #extended
作者：
原文：Code for the paper "Jukebox: A Generative Model for Music"
karpathy/reader3
标签：#github_orgs #extended
作者：
原文：Quick illustration of how one can easily read books together with LLMs. It's great and I highly recommend it.
anthropics/claude-quickstarts
标签：#github_orgs #extended
作者：
原文：A collection of projects designed to help developers quickly get started with building deployable applications using the Claude API
Cybersecurity in the Intelligence Age
标签：#ai_engineering_blogs #core
作者：
原文：OpenAI outlines a five-part action plan for strengthening cybersecurity in the Intelligence Age, focused on democratizing AI-powered cyber defense and protecting critical systems.
链接：https://openai.com/index/cybersecurity-in-the-intelligence-age
anthropics/claude-agent-sdk-typescript
标签：#github_orgs #extended
作者：
原文：anthropics/claude-agent-sdk-typescript recently updated repository.
链接：https://github.com/anthropics/claude-agent-sdk-typescript
openai/whisper
标签：#github_orgs #extended
作者：
原文：Robust Speech Recognition via Large-Scale Weak Supervision
anthropics/anthropic-sdk-python
标签：#github_orgs #extended
作者：
原文：anthropics/anthropic-sdk-python recently updated repository.
OpenAI models, Codex, and Managed Agents come to AWS
标签：#ai_engineering_blogs #core
作者：
原文：OpenAI GPT models, Codex, and Managed Agents are now available on AWS, enabling enterprises to build secure AI in their AWS environments.
Our commitment to community safety
标签：#ai_engineering_blogs #core
作者：
原文：Learn how OpenAI protects community safety in ChatGPT through model safeguards, misuse detection, policy enforcement, and collaboration with safety experts.
链接：https://openai.com/index/our-commitment-to-community-safety
RT by @karpathy: New work with @AlecRad and @DavidDuvenaud: Have you ever dreamed of talking to someone from the past?
标签：#x_profiles #extended
作者：
原文：New work with @AlecRad and @DavidDuvenaud Have you ever dreamed of talking to someone from the past? Introducing talkie, a 13B model trained only on pre-1931 text. Vintage models should help us to understand how LMs generalize (e.g., can we teach talkie to code?). Thread: Video
链接：https://twitter.com/status_effects/status/2048878495539843211
OpenAI available at FedRAMP Moderate
标签：#ai_engineering_blogs #core
作者：
原文：OpenAI is available at FedRAMP Moderate authorization for ChatGPT Enterprise and the OpenAI API, enabling secure AI adoption for U.S. federal agencies.
链接：https://openai.com/index/openai-available-at-fedramp-moderate
Codex CLI 0.128.0 adds /goal
标签：#ai_engineering_blogs #core
作者：
原文：Codex CLI 0.128.0 adds /goal The latest version of OpenAI's Codex CLI coding agent adds their own version of the Ralph loop you can now set a /goal and Codex will keep on looping until it evaluates that the goal has been completed... or the configured token budget has been exhausted. It looks like the feature is mainly implemented though the goals/continuation.md and goals/budget_limit.md prompts, which are automatically injected at the end of a turn. Via @fcoury Tags: ai openai prompt-engineering generative-ai llms coding-agents system-prompts codex-cli agentic-engineering
链接：https://simonwillison.net/2026/Apr/30/codex-goals/#atom-everything
Our evaluation of OpenAI's GPT-5.5 cyber capabilities
标签：#ai_engineering_blogs #core
作者：
原文：Our evaluation of OpenAI's GPT-5.5 cyber capabilities The UK's AI Security Institute previously evaluated Claude Mythos now they've evaluated GPT-5.5 for finding security vulnerability and found it to be comparable to Mythos, but unlike Mythos it's generally available right now. Tags: ai openai generative-ai llms anthropic claude ai-security-research gpt
链接：https://simonwillison.net/2026/Apr/30/gpt-55-cyber-capabilities/#atom-everything
Quoting Andrew Kelley
标签：#ai_engineering_blogs #core
作者：
原文：It's a common misconception that we can't tell who is using LLM and who is not. I'm sure we didn't catch 100% of LLM-assisted PRs over the past few months, but the kind of mistakes humans make are fundamentally different than LLM hallucinations, making them easy to spot. Furthermore, people who come from the world of agentic coding have a certain digital smell that is not obvious to them but is obvious to those who abstain. It's like when a smoker walks into the room, everybody who doesn't smoke instantly knows it. I'm not telling you not to smoke, but I am telling you not to smoke in my house. -- Andrew Kelley Creator of Zig Tags: zig llms ai generative-ai
链接：https://simonwillison.net/2026/Apr/30/andrew-kelley/#atom-everything
Reinforcement fine-tuning with LLM-as-a-judge
标签：#engineering_ai_infra_blogs #extended
作者：
原文：In this post, we take a deeper look at how RLAIF or RL with LLM-as-a-judge works with Amazon Nova models effectively.
链接：https://aws.amazon.com/blogs/machine-learning/reinforcement-fine-tuning-with-llm-as-a-judge/
We need RSS for sharing abundant vibe-coded apps
标签：#ai_engineering_blogs #core
作者：
原文：We need RSS for sharing abundant vibe-coded apps Matt Webb: I would love an RSS web feed for all those various tools and apps pages, each item with an “Install” button. (But install to where?) The lesson here is that when vibe-coding accelerates app development, apps become more personal, more situated, and more frequent. Shipping a tool or a micro-app is less like launching a website and more like posting on a blog. This inspired me to have Claude add an Atom feed (and icon) to my /elsewhere/tools/ page, which itself is populated by content from my tools.simonwillison.net site. Tags: atom matt-webb rss ai vibe-coding
链接：https://simonwillison.net/2026/Apr/30/rss-vibe-coded-apps/#atom-everything