TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Is Bard trained on Gmail? Depends what the meaning of the word “is” is

11 点作者 purplesnowflake将近 2 年前

2 条评论

smoldesu将近 2 年前
&gt; AI researcher Kate Crawford was quick to ask Bard itself where its dataset came from. The answer caught her attention: Bard said one of its data sources was Gmail.<p>Did they find anything? There&#x27;s a lot of hand-wringing at the start, then a big focus on how Google can&#x27;t deny that emails are in their training data. Then they finish by interviewing Bard. Google&#x27;s response makes sense, given that they&#x27;re working with multi-terabyte language files. It probably has seen Gmail contents through the form of naturally published emails that just get picked up with other data. Claiming otherwise would be confidently wrong.<p>It would be interesting if they had a &quot;Q_rsqrt in Copilot&quot; moment here, but they don&#x27;t. There seems to be no evidence that Google uses private data in Bard.<p>&gt; Society should be having a robust discussion on these questions, but this is not possible if such discussion is inhibited by key players like Google.<p>How is Google inhibiting this discussion?
评论 #36128760 未加载
version_five将近 2 年前
The whole asking Bard thing towards the end is completely meaningless and I&#x27;d argue irresponsible. They even say<p><pre><code> But of course, the observation that Bard consistently makes these claims can’t be seen as evidence one way or the other </code></pre> and then go on to quote a bunch of stuff Bard said.<p>If I had to speculate, sounds like it could have used anonymized gmail data (could they have some kind of pii removal tool that they run first, that&#x27;s common, though I wouldn&#x27;t trust it too much), or something is being pretrained on gmail and fine tuned on something else (hard to see a reason for that). Anyway, google is acting suspicious, but pretending the chatbot&#x27;s &quot;opinion&quot; has any bearing is disingenuous.