Media Titans Shielding Their Content from AI's Digital Foragers
Posted: 2023-08-25
By: dwirch
Viewed: 49
Filed Under:
No attachments for this post
OpenAI uses a unique tool that roams the digital space to continuously update its primary language model, ChatGPT. However, it's recently hit some roadblocks.
Several renowned news outlets, such as the New York Times, CNN, Reuters, and the Chicago Tribune, have restricted ChatGPT's access to their online content, as noted by the Guardian. The list extends to the Australian Broadcasting Corporation and Australian Community Media, which owns a sizable collection of local publications.
The tool in question is GPTBot, a web-crawling mechanism initiated by OpenAI recently. OpenAI suggests on its platform that allowing GPTBot to browse your website can enhance the precision, overall abilities, and safety of AI models.
Originally, ChatGPT's training only covered events up until September 2021. To keep up with the fast-paced advancements in the AI realm, the model needed to update its knowledge base beyond events like Joe Biden's presidency and the U.S. military's departure from Afghanistan. OpenAI sought to remedy this by experimenting with a "Browse with Bing" feature and unleashing GPTBot to browse the web. However, the reluctance of leading media houses, often possessing timely and significant news, could be a setback for this chatbot.
Terms and conditions of major publishers, such as the Times, Reuters, and the Tribune, explicitly prohibit data scraping. Specifically, the Times' regulations assert that their content shouldn't be used to train AI tools. The mainstay of these publishers is selling information, either via subscriptions or advertisements, necessitating web traffic for revenue. If chatbots are freely given access to this content, it might reduce direct visits to these sites, impacting their financial returns.
Traditional media houses are already grappling with a shift in their revenue models due to the dominance of social media ads.
By denying access to GPTBot, these media houses might be hinting at a commercial approach, possibly compelling OpenAI to purchase the rights. Recently, OpenAI entered an agreement with the Associated Press, allowing them to utilize their news stories for AI training. While the financial aspects of this agreement remain undisclosed, it sets a precedent that companies like Google might follow, especially when facing similar restrictions from news publishers.
Comments on this post
No comments have been added for this post.
You must be logged in to make a comment.