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。

Firms like Meta and A16Z admit paying billions for data would ruin their AI plan

23 点作者 saeedjabbar超过 1 年前

3 条评论

PraetorianGourd超过 1 年前
Yeah, paying billions for something you currently get for free would certainly cut into profits by quite a bit. Ford would make a lot more money if they didn't need to pay for metal, plastic, or electricity.
评论 #38136358 未加载
biglyburrito超过 1 年前
Huh. It's almost like the only way for these companies to make it big with AI is by cheating others out of fair compensation for their data. Who would have thought?
figassis超过 1 年前
Why don’t they adopt a model like Spotify, when’re they track what information they’re scraping and how much if it relative to all the information they collect (bc I don’t think they can track how it’s being used by the model), and compensate copyright owners for a percentage of revenue? One perverse incentive might be that people will want to create more quantity than quality (like with SEO), but this could be a start. Maybe they can find a way to determine which information was used by the model for a certain output, sort of like the “show your work” feature.