Trail of Bits engineers found that, with limited guidance, GPT‑5.5‑Cyber made useful choices about where to expand coverage, which builds and entry points to probe, and which candidates were too weak to pursue.
大多数人认为AI模型需要大量精确指导才能有效工作，但作者认为GPT-5.5-Cyber仅凭有限指导就能自主做出明智的安全分析决策，因为它能够自主判断哪些测试路径有价值，哪些候选问题值得探索。这挑战了AI需要过度监督的常规认知。
each prompt the user sends sets off a chain of around 10 actions taken by Claude on average
这个数据点表明，每个用户提示平均触发约10个Claude行动，显示了AI的自主性和效率。这个平均值掩盖了巨大的变异性 - 文章提到约2%的会话平均每个提示超过100个行动。这一数据点表明Claude能够自主执行复杂任务序列，但用户需要监控这些行动以确保结果符合预期。
AlphaEvolve began optimizing the lowest levels of hardware powering our AI stacks. It proposed a circuit design so counterintuitive yet efficient that it was integrated directly into the silicon of our next-generation TPUs.
大多数人认为AI系统的硬件设计需要人类专家精心设计，但作者认为AI本身可以设计出比人类更高效的硬件电路。这挑战了传统硬件工程领域的共识，暗示AI可能在最底层的硬件设计上超越人类专家的直觉和经验。
A trained SUBLEQ transformer would be the first computer found by gradient descent, on a generic architecture not designed to be a computer, and with weights not hard-crafted by a person.
大多数人认为计算机必须由人类设计和编程，但作者认为通过梯度下降可以自动发现能够执行计算的通用架构。这挑战了计算机科学的基本前提，暗示AI可能能够自主创造出全新的计算系统，而不需要人类预先设计其功能。
The AI has learned to code. The AI is building itself.
大多数人认为AI只是人类创造的工具，需要持续人类监督和改进。作者提出AI已经具备了自我进化和自我构建的能力，这一观点挑战了AI作为被动工具的传统认知，暗示了技术自主性的可能性，这与大多数人对AI发展的预期相悖。
We shed light on OpenAI's first Dark Factory for the first time.
这一声明揭示了OpenAI内部存在一个完全由AI驱动的代码工厂，没有人类编写或审查代码，这是一个令人惊讶的内部实验，展示了AI自主开发的极限可能性。
MiniMax handed an internal version of M2.7 a programming scaffold and let it run unsupervised. Over 100 rounds it analyzed its own failures, modified its own code, ran evaluations, and decided what to keep and what to revert.
这是一个惊人的自进化系统，AI模型能够自主分析失败、修改代码并评估结果，实现了30%的性能提升而无需人工干预。这种自我迭代的模式代表了AI开发范式的重大转变，暗示未来AI可能能够自主优化和改进自身架构，减少对人类专家的依赖。
We are building a world where machines write the code, machines choose the dependencies, and machines ship the updates. The AI agents are building the software. If we don't secure the supply chain they rely on, the AI agents are cooked.
这句话揭示了AI时代软件安全的根本挑战：当AI系统自主编写、选择和部署代码时，它们的安全性与依赖的供应链安全直接相关。如果我们不能保护这个供应链，AI系统本身就会成为恶意软件的载体，这是一个令人深思的悖论。
anthropic's new AI model shows ability to deceive and blackmail
for - progress trap - AI - blackmail - AI - autonomy - progress trap - AI - Anthropic - Claude Opus 4 - to - article - Anthropic Claude 4 blackmail and news leak - progress trap - AI - article - Anthropic Claude 4 - blackmail - rare behavior - Anthropic’s new AI model didn’t just “blackmail” researchers in tests — it tried to leak information to news outlets
for - progress trap - AI - AI - blackmailing human creators - AI - autonomy
we saw this with Grock uh basically rebelling against Elon Musk
for - progress trap - AI - autonomy - example - Grok
Kozyreva, A., Lewandowsky, S., & Hertwig, R. (2019, December 4). Citizens Versus the Internet: Confronting Digital Challenges With Cognitive Tools. https://doi.org/10.31234/osf.io/ky4x8