原文
| ||||||||||
| ||||||||||
![]() |
原始链接: https://news.ycombinator.com/item?id=43445616
Hacker News 上的一篇帖子讨论了《大西洋月刊》报道的 Meta 使用盗版书籍训练其 AI 的问题。 用户“Hary06” 指出文章对 BitTorrent 的误解,认为它主要用于加速下载,而不是匿名,并且上传文件并非强制性。他们最关心的问题是下载盗版材料的行为本身。 “upghost” 表达了愤世嫉俗的观点,认为科技公司往往能够在早期逃脱非法行为的后果,要么是因为失败,要么是因为花钱摆平麻烦。“xyzzy9563” 指出,在美国下载盗版材料是合法的,但分享是不合法的。其他评论批评了 Meta 的道德,一个用户主张抵制他们的服务,另一个用户则以愤世嫉俗的回应回应,开玩笑说替代方案是中国 AI 使用相同的盗版材料。一位评论者还发布了一条无关的仇恨信息。
| ||||||||||
| ||||||||||
![]() |
Isn't this a twofold misunderstanding of BitTorrent? I haven't used it much, but I've never believed BitTorrent to be popular for anonymity (maybe they're thinking of Tor?), I thought it was popular because it makes downloading go faster by reducing bottlenecks. Also, choosing not to seed a file is extremely simple in every torrent client I've seen, so it seems a bit of a leap to conclude that Meta seeded the pirated books just because the protocol supports it.
I'm far more concerned about the simple fact that they downloaded pirated books at all than I am about the protocol they used to do it.
reply