Can Generative AI be Egalitarian?

Current AI models extract value without fair compensation: Foundation generative AI models are built by extensively extracting content from online sources "often without corresponding reciprocation," mirroring the exploitative practices of surveillance capitalism and raising ethical concerns about who benefits from AI development.

An egalitarian alternative exists: The paper proposes developing generative AI models based on content "willingly and collaboratively provided by users," drawing inspiration from Wikipedia's successful model of voluntary contribution—suggesting this approach is both ethically sound and technically viable.

Collaborative AI could be more aligned with society: Models built on user-contributed content may be "more responsive to user needs, more diverse in their training data, and ultimately more aligned with societal values" compared to those trained on extracted data.

Responsible AI commitments are being challenged by profit incentives: The potential for enormous profits has caused technology organizations to compromise their commitments to responsible AI practices, highlighting a tension between commercial interests and ethical development.

Significant practical challenges remain: While promising, the egalitarian approach faces real obstacles including scalability concerns, quality control issues, and potential biases inherent in volunteer-contributed content—meaning this model is not a simple solution to current problems.