Tumblr Strikes Training Data Deals with OpenAI and Midjourney: Report

Explore the latest report on Tumblr's collaboration with OpenAI and Midjourney for training data. Discover the potential impact on user privacy and the implications for AI development.

Tumblr Strikes Training Data Deals with OpenAI and Midjourney: Report

Tumblr, the popular blogging platform, is reportedly entering into significant partnerships with AI companies OpenAI and Midjourney to leverage training data, according to a recent report by The Verge. The collaboration aims to provide valuable insights for AI development but raises concerns about user privacy and data sharing.

The report, based on insider information, indicates that Automattic, the parent company of Tumblr and WordPress.com, is in talks with OpenAI and Midjourney to scrape training data from users' posts. Although specific details of the deals remain scarce, the partnerships are said to be imminent, potentially opening up new revenue streams for Tumblr.

To address privacy concerns, Automattic plans to introduce a new setting that allows users to opt-out of data sharing with third parties, including AI companies. However, internal posts suggest that an initial data dump containing all public post content from 2014 to 2023, including non-publicly visible content, was inadvertently scraped. The fate of this data and whether any of it has been shared with Midjourney and OpenAI remains unclear.

While OpenAI and Midjourney have yet to comment on the matter, Automattic released a public statement titled "Protecting User Choice." The statement acknowledges partnerships with undisclosed AI companies and emphasizes the company's commitment to community concerns such as attribution, opt-outs, and user control. Automattic also mentions blocking major AI platform crawlers by default and only sharing publicly hosted content from sites that haven't opted out.

This collaboration between Tumblr and AI companies echoes similar partnerships in the industry. Reddit allegedly has a $60 million annual deal with Google, and Shutterstock has entered into an agreement with OpenAI to train on its extensive photo library. However, the use of artists' and writers' work for training purposes has sparked protests from the creative community, leading to a delicate balance between user satisfaction and AI experimentation.

Automattic, known for its web hosting business with WordPress.com and WordPress VIP, acquired Tumblr from Verizon in 2019. Monetizing Tumblr has posed challenges for the company, prompting a scaling back of ambitions for the platform. The potential gains for Automattic from the partnership with OpenAI and Midjourney remain uncertain.

In conclusion, Tumblr's reported collaborations with OpenAI and Midjourney for training data signify an ongoing trend in the industry. While the partnerships hold promise for AI development, concerns surrounding user privacy and consent must be carefully addressed. As the landscape of AI continues to evolve, it is crucial to strike a balance between innovation and respecting the rights and interests of users and content creators.

What's Your Reaction?

like
3
dislike
0
love
1
funny
0
angry
0
sad
0
wow
1