The authors argue that generative AI introduces a new class of alignment risks because interaction itself becomes a mechanism of influence. Humans adapt their behavior in response to AI outputs, ...

AI船长喵喵在上一篇介绍Claude的文章中提到了“人类反馈强化学习”（RLHF）技术和“Constitutional AI”。这两个研究也是致力于实现人工智能对齐领域的最前沿的技术。“人类反馈强化学习”（RLHF）技术采用的更多的是直接性规范。RLHF主要依靠人类对 AI 模型的回应进行评级反馈 ，研究人员再将这些人类的偏好反馈给模型以告诉 AI 哪些回应是合理的。这就造成了 RLHF ...

We’re now deep into the AI era, where every week brings another feature or task that AI can accomplish. But given how far down the road we already are, it’s all the more essential to zoom out and ask ...

Almost 2,000 years before ChatGPT was invented, two men had a debate that can teach us a lot about AI’s future. Their names were Eliezer and Yoshua.

Technology alone is no longer enough. Organizations face an unprecedented proliferation of tools, platforms and systems, each ...

An Osaka Metropolitan University-led research team conducted a review that examined the theoretical foundations of Artificial Superintelligence and explored how misaligned AI systems could optimize ...

Forbes contributors publish independent expert analyses and insights. Andrea Hill is a multi-industry CEO covering business & technology. Despite $30–40 billion in enterprise investment in generative ...

The results reveal a consistent pattern across models. While responses are fluent, calm, and socially acceptable, they fail ...