Tell HN: Gemini trains on chat history, even for subscribers (personal accounts)
本文为原文前 6,000 字符的节选翻译，完整内容请查看原文。
If you have a personal Google AI subscription, all your saved chat history in Gemini can be used for training and there is no way to opt out unless you turn off chat history or use temporary chat sessions. These chat transcripts may even be read by human reviewers. Although Google claims to disconnect them from your account when human-reviewed, they explicitly warn not to enter confidential information.
如果您拥有个人 Google AI 订阅，您在 Gemini 中保存的所有聊天记录都可能被用于模型训练，除非您关闭聊天记录或使用临时聊天会话，否则无法选择退出。这些聊天记录甚至可能被人工审核员阅读。尽管 Google 声称在人工审核时会将记录与您的账户断开关联，但他们明确警告不要输入机密信息。
To disable their training on your chat sessions, you need to turn off “Keep Activity” in Gemini’s settings. Turning that off means you’ll lose any chat sessions after 72 hours. I learned this too late, thinking that having a paid subscription = not being turned into the product. Apparently paying is not enough.
要禁止他们利用您的聊天会话进行训练，您需要在 Gemini 的设置中关闭“保留活动记录”（Keep Activity）。关闭该功能意味着 72 小时后您将丢失所有聊天会话。我意识到这一点时已经太晚了，我原以为付费订阅就意味着不会被当作产品对待。显然，付费是不够的。
PSA: This is true of all the frontier model providers even if you disable activity tracking. They are all constantly monitoring your interactions for problematic behavior.
公告：即使您禁用了活动跟踪，所有前沿模型提供商的情况也都是如此。他们都在持续监控您的交互行为，以防范问题行为。
The sole reason LLMs has gotten this good is because their aggressive expansion has been accompanied with literally this. The amount of uncredited information these models have been fed must be astonishing, also terrifying.
大语言模型之所以能达到现在的水平，唯一的原因就是它们激进的扩张伴随着这种做法。这些模型被喂入的未经授权的信息量一定非常惊人，同时也令人恐惧。
I think the heavily subsidized subscriptions are exactly for this purpose - the training data they generate is worth the high subsidy.
我认为高额补贴的订阅服务正是为了这个目的——它们所产生的训练数据价值远高于高额补贴。
Imagine all those tasty confidential IP and contract documents they have ingested. You couldn’t buy those.
想象一下，那些它们已经摄入的、诱人的机密知识产权和合同文件。这些东西你是花钱也买不到的。
For an AI system to be “completely safe”, like a self-driving car, it must map the full environment it operates in and be taught the “right way” to handle every conceivable scenario. So when you present it with novel data or a situation, that constitutes an “AI safety event”, which per their ToS is a carve out for training on your session.
为了让 AI 系统“完全安全”（就像自动驾驶汽车一样），它必须映射其运行的完整环境，并被教导如何以“正确的方式”处理每一种可预见的情况。因此，当您向它提供新颖的数据或情况时，这就构成了一个“AI 安全事件”，根据其服务条款，这成为了在您的会话上进行训练的例外情况。
What’s worse is if you have a Pixel phone then Gemini is completely useless without enabling this setting. I had it switched off: “Hey Google, remind me at 7pm
更糟糕的是，如果您使用 Pixel 手机，如果不开启此设置，Gemini 将完全无法使用。我之前把它关掉了：“嘿 Google，晚上 7 点提醒我 <某事>”，“好的，您的提醒已设为晚上 7 点”。第二天：“嘿 Google，为什么你没在晚上 7 点提醒我？”“除非您开启聊天记录，否则我无法对您的手机执行任何操作。”