OpenAI 的 GPT-4 人工智能技术在众多关于 AI 威胁的猜测中，被认为可能被用于制造生物武器，这是一个非常令人担忧的问题。然而，OpenAI 最新的一项评估实验显示，GPT-4 技术“最多只能轻微提升制造生物威胁的信息获取能力”。这意味着，使用 GPT-4 制造生物武器的风险实际上是非常小的。
OpenAI 进行的这项实验旨在验证 AI 技术是否真的能够被用于制造生物武器。实验结果显示，GPT-4 技术虽然可以提供一些有关制造生物武器的信息，但其作用非常有限，无法显著提高制造生物武器的能力。
这个结果对于那些担心 AI 技术会被用于制造生物武器的人来说，无疑是一个令人松一口气的好消息。然而，这并不意味着我们可以对 AI 技术的潜在威胁掉以轻心。OpenAI 表示，他们将继续进行相关的研究和评估，以确保 AI 技术的安全性和可靠性。
Title: Low risk of AI-based bio-weaponry with GPT-4
Keywords: AI, bio-weaponry, GPT-4
News content: Amidst numerous concerns about AI threats, one of the most alarming speculations is the potential use of AI technology in creating biological weapons. However, a recent assessment experiment conducted by OpenAI reveals that GPT-4, their AI technology, “can only slightly enhance the information-gathering capabilities for creating biological threats.” This suggests that the risk of using GPT-4 to manufacture biological weapons is actually very low.
The purpose of the experiment conducted by OpenAI was to verify whether AI technology could really be used to create biological weapons. The results showed that while GPT-4 can provide some information related to the manufacturing of biological weapons, its role is very limited and cannot significantly improve the ability to create such weapons.
This result comes as a relief for those who have been worried about the potential use of AI technology in manufacturing biological weapons. However, it does not mean that we can become complacent about the potential risks of AI technology. OpenAI stated that they will continue to conduct relevant research and assessments to ensure the safety and reliability of AI technology.
【来源】https://www.cls.cn/detail/1587372
Views: 1