AI's top startups are barely publishing their research (science.org)

Article summary

The article discusses how top AI startups are not publishing their research, with some commenters noting that this trend is at odds with the traditional open nature of the industry. The discussion reveals that some companies are buying and destroying rare books to scan them, potentially to avoid legal liability. The industry's shift towards secrecy is seen as driven by greed and a desire to capture value. The role of published research in the development of AI is also highlighted.

Main themes

• AI research secrecy
• open research and collaboration
• rare book preservation
• industry trends and motivations
• legal liability and copyright
• innovation and progress in AI

What commenters say

• The AI industry's shift towards secrecy is driven by greed and a desire to capture value.
• Published research has been crucial to the development of AI, and its decline may hinder progress.
• Some companies are prioritizing the preservation of rare books and scanning them without destruction.
• The destruction of rare books for scanning is a necessary step to avoid legal liability.
• The industry's focus on secrecy may be misguided, and open research is still essential for progress.
• The current state of AI research is not as innovative as it seems, with some arguing that it is merely building on existing ideas.
• The decline of open research in AI may have negative consequences for the industry and society as a whole.