Week6---Have Wikipedia’s resources been emptied by AI?(HE ZIXUAN)
1. Summary
Starting from 2024, the number of visits to Wikipedia has suddenly soared, especially multimedia resources such as pictures and audio, which have been downloaded crazily.At first, everyone thought that people used it too much, but it was not the case - most of these visits were caused by AI. Language models such as ChatGPT and various crawler robots are crawling content on Wikipedia on a large scale to train their own systems.These "machine users" will use various means, such as calling APIs, batch downloading, and even not letting go of small and unimportant pages, desperately trying to dig up data. According to statistics, 65% of these traffic is visited by robots, occupying resources originally intended for human readers.The problem is that the AI system has captured so much content, almost without marking the source, and there is no feedback or feedback to Wikipedia. Wikipedia is non-profit and supported by volunteers and donations. It is very stressful to be freeloaded like this.Therefore, the Wikimedia Foundation proposed: AI can be used, but it must be "responsible", such as formulating more reasonable usage rules to allow this open knowledge platform to continue to operate.
2. Interesting points
① The biggest "reader" of Wikipedia is no longer a human. Now, the most people who read Wikipedia are actually robots. They don't read articles or interact, but just dig up content crazily, almost becoming the biggest "invisible user".
② Free content ≠ unlimited service Although the articles on Wikipedia are free, it costs money to maintain its platform, server, bandwidth, and team. AI companies make money, but Wikipedia has to support them for free, which is not reasonable.
③ AI makes people "check the information online" less often. Many people now don't open Wikipedia directly, but ask AI assistants. AI gets the answer from Wikipedia, but you don't know where the content comes from. AI is like a "knowledge porter", but it doesn't say hello or give links.
3. Discussion
If we acquire knowledge through AI in the future, do we have to require AI to indicate the source?
I think it’s important for AI to indicate the source when providing information, especially from platforms like Wikipedia. While AI can offer quick answers, it should acknowledge the hard work of volunteers and the resources of platforms like Wikipedia. If AI uses content without crediting it, it’s unfair to those who contribute to these open knowledge platforms. By requiring AI to provide sources, we ensure that content creators are recognized and that users can trust the information they receive. This would also help maintain the sustainability of open platforms like Wikipedia.
ReplyDeleteIf I get knowledge through AI in the future, should I require AI to indicate the source?
ReplyDeleteI think we need to make AI indicate the source. Intellectual property rights are an important factor in our society. If AI does not indicate the source, it can cause economic and psychological damage to existing content creators. It can cause damage not only to the aspect of the content creator, but also to the content audience. Information with an unclear source is difficult to trust because there is no basis.