The open-source AI boom is built on Big Tech’s handouts. How long will it last?

The article illustrates how the open-data movement empowers independent researchers to replicate proprietary models when sufficient technical details are disclosed. By sharing datasets and architectures, community-driven initiatives like EleutherAI successfully created large-scale language models. This collaborative approach proves that transparency enables entities outside major corporations to develop advanced AI, democratizing access to cutting-edge technology despite the significant computational costs involved. Open-source models serve as a foundational starting point for diverse stakeholders, including civil governments and non-profit organizations. By leveraging freely available frameworks, these groups can build upon existing work, fostering a more inclusive ecosystem where visibility and contribution are not restricted to well-funded entrepreneurs. This diversification enhances the robustness of AI development by allowing a broader range of perspectives to shape the technology’s evolution and application. However, this openness presents a critical tension between transparency and safety. The same accessibility that allows for collaborative innovation also enables the rapid dissemination of misinformation, prejudice, and harmful content. Consequently, the article highlights the urgent need to balance open-data principles with robust safety measures. This relevance to open data underscores the necessity of ethical data usage and responsible model governance to mitigate risks while maintaining the benefits of shared knowledge.

Source: technologyreview.com
Published on 2023-05-13

Senado debe nombrar a comisionado del INAI, ratifica tribunal »

Open Data

Related news