OpenAI 近日在其博客上宣布，将组建一个名为“集体对齐”（Collective Alignment）的全新团队，致力于收集公众意见，以确保其人工智能大模型与人类价值观保持一致。该团队主要由研究人员和工程师构成，他们将设计和实施收集公众意见的流程，以帮助训练和塑造 AI 模型，解决潜在的偏见和其他问题。
随着人工智能技术的不断发展，其在大数据分析、自然语言处理等领域的应用日益广泛，但也引发了一系列关于伦理和价值观的争议。为确保 AI 模型在与人类互动时遵循正确的价值观，OpenAI 采取了一系列措施。此次组建的“集体对齐”团队便是其中之一，旨在让公众参与 AI 模型的训练和优化过程，从而提高模型的公正性、透明度和可信度。
为了实现这一目标，OpenAI 将通过线上问卷、研讨会、圆桌讨论等多种形式，广泛征求公众对 AI 模型的看法和建议。此外，该团队还将与学术界、产业界和政策制定者展开合作，共同探讨如何将人类价值观融入 AI 技术的发展。
英文翻译：
News Title: OpenAI Forms New Team to Ensure AI Values Align with Humanity
Keywords: OpenAI, Collective Alignment, Artificial Intelligence Values
News Content:
OpenAI recently announced that it will form a new team called “Collective Alignment,” dedicated to collecting public opinions to ensure that its large-scale AI models align with human values. The team, consisting of researchers and engineers, will design and implement processes to collect public opinions to help train and shape AI models, addressing potential biases and other issues.
As artificial intelligence technology continues to evolve, its applications in areas such as big data analysis and natural language processing are becoming increasingly widespread. However, this has also led to controversies about ethics and values. To ensure that AI models adhere to the correct values when interacting with humans, OpenAI has taken various measures. The formation of the “Collective Alignment” team is one of them, aiming to involve the public in the training and optimization of AI models, thereby improving their fairness, transparency, and credibility.
To achieve this goal, OpenAI will widely solicit public opinions and suggestions on AI models through online questionnaires, workshops, roundtable discussions, and other forms. In addition, the team will also collaborate with academia, industry, and policymakers to explore how to integrate human values into the development of AI technology.
【来源】https://www.ithome.com/0/745/634.htm
Views: 0