As I mentioned a short while ago in my article, "From Atoms to Algorithms" - we need to have governing guidelines to help foster trust and safety with society as AI tools and platforms become more accessible.
Upon reviewing the White House's "Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence," it became evident that the available online version posed readability challenges due to the absence of a table of contents and inconsistent indentation. To address these issues, a Google Document was meticulously structured with a clickable table of contents, and efforts were made to appropriately indent all subsections. The creators also developed a separate version tailored for collaborative commenting and discussion, although they expressed concerns about potential misuse and the possibility of disabling this feature. Furthermore, they enabled the option for anyone to copy either the clean version or the version for commenting, with a suggestion to link back to the original version as a courtesy to others.
The initiative was led by David Evan Harris, Vinaya Sivakumar, Daniel Jang, and Maddy Cooper from the University of California, Berkeley. They have provided their contact information and encouraged feedback on how to further improve the document's accessibility. The collaborative and lighthearted approach taken by the authors reflects their commitment to enhancing the usability of the executive order and inviting others to contribute to the process. Overall, the text conveys a proactive and inclusive tone, with a touch of humor in their caution about the potential for a "big flame war" if the collaborative commenting feature is misused.
For further details, please refer to the original document or contact the creators directly at deh@berkeley.edu.
