Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

This Week in AI: Midjourney bets it can beat the copyright police | TechCrunch

Mar 16, 2024 - techcrunch.com
The AI startup Midjourney has made a subtle change to its terms of service related to IP disputes, replacing informal language with more legalistic terms. This change reflects Midjourney's belief that AI vendors will win legal battles with creators whose works are used as training data for AI models. However, this approach is seen as risky, especially as Midjourney has been brazen in its use of copyrighted works for training its models. The company, currently valued at around $200 million, could face significant legal costs and potential decimation if fair use is not upheld in its case.

In other AI news, EU authorities are scrutinizing AI platforms ahead of elections, Google Deepmind is developing an AI co-op gaming partner, and AI startup Anthropic has launched a new family of models. Meanwhile, a study from the Center for Countering Digital Hate highlights the rise of AI-generated disinformation, particularly deepfake images related to elections. Lastly, Amazon's new AI-powered chatbot, Rufus, has been criticized for its limited capabilities.

Key takeaways:

  • AI startup Midjourney has made changes to its terms of service regarding IP disputes, reflecting its belief that AI vendors will win courtroom battles over the use of creators' works in training data.
  • Midjourney has been using copyrighted works to train its models, a practice that could lead to legal issues if fair use is not deemed applicable.
  • AI models are being used to understand and predict molecular dynamics, with Microsoft's new model ViSNet aiming to predict structure-activity relationships.
  • Researchers at the University of Manchester are using machine learning to analyze genetic data for the early discovery of major COVID-19 variants.
View Full Article

Comments (0)

Be the first to comment!