谷歌近日宣布，其 Gemini 大模型的文生图功能因被曝出刻意追求“多元化”、忽视准确性的问题，已决定暂停该功能。据 BBC 和 The Verge 的报道，一些用户反映，Gemini 在生成人物图像时，倾向于展示各种性别和种族的多样性，但这种做法可能导致生成的图片与实际历史或事实不符。这一情况引发了对技术公正性和准确性的担忧。
谷歌在回应中表示，公司正在“加紧”修复这一bug，以确保 Gemini 的图像生成能够平衡多元性和准确性。作为解决措施，谷歌已决定暂时停用 Gemini 的人物图像生成功能，以便对相关问题进行深入调查和修复。谷歌承诺将尽快推出一个改进的版本，以恢复服务并提高生成图像的可靠性。
此次事件凸显了科技公司在推动多元化的同时，必须谨慎处理潜在的偏见和误导问题。谷歌的这一决定表明，即使在追求技术创新的过程中，也不会忽视对内容真实性和公平性的坚守。
英语如下：
News Title: “Google Pauses Gemini Image Generation over Controversial Diversity Pursuit, Fixes Underway”
Keywords: Google fix, Gemini model, diverse imagery
News Content: Google recently announced that it has suspended the image generation function of its Gemini large language model due to concerns over a deliberate emphasis on “diversity” at the expense of accuracy. As reported by BBC and The Verge, some users have noted that Gemini tends to depict a wide range of genders and ethnicities when generating人物 images, which might lead to pictures that do not align with actual historical or factual contexts. This situation has raised questions about the technological fairness and accuracy.
In response, Google stated that it is “accelerating” efforts to fix the issue, ensuring that Gemini’s image generation balances diversity with accuracy. As a remedial measure, the company has temporarily disabled Gemini’s ability to generate character images to conduct a more in-depth investigation and rectify the problem. Google has pledged to roll out an improved version as soon as possible to restore the service and enhance the reliability of generated images.
This incident underscores the need for tech companies to address potential biases and misrepresentations while promoting diversity. Google’s decision demonstrates that even in the pursuit of technological innovation, the company remains committed to upholding content authenticity and fairness.
【来源】https://www.ithome.com/0/751/457.htm
Views: 1