Introduction:
As artificial intelligence (AI) becomes increasingly ubiquitous, many people have raised concerns about the impact it will have on jobs and society. Some argue that a universal basic income (UBI) could be a solution to help people transition as jobs are lost to automation. However, others, such as Sam Altman, CEO of OpenAI, believe that UBI is only a partial solution. In this context, data dignity has been proposed as an alternative approach. In this article, we will explore what data dignity is and how it could be implemented.
The Premise of Data Dignity:
Data dignity is based on the idea that people should be paid for their contributions to AI. Currently, people provide their data for free in exchange for free services. However, as AI becomes more prevalent, this data is becoming increasingly valuable. Therefore, people should be compensated for their contributions, even if their data is filtered and recombined into something unrecognizable. This approach was first introduced by computer scientist Jaron Lanier in a 2018 Harvard Business Review piece titled “A Blueprint for a Better Digital Society.”
Challenges of Implementing Data Dignity:
Implementing data dignity is not without its challenges. For example, it is difficult to assign people the right amount of credit for their contributions to AI. Furthermore, even data-dignity researchers cannot agree on how to disentangle everything that AI models have absorbed or how detailed an accounting should be attempted. Additionally, there is a challenge of access. OpenAI, for example, has closed off its training data, making it difficult for researchers to examine the data and understand how it was used to train models. Finally, there is a regulatory challenge. Regulators are still grappling with how to address data dignity and other related issues. OpenAI, for example, is facing regulatory scrutiny in several countries.
The Importance of Data Dignity:
Despite the challenges, data dignity is an important concept. It recognizes that people should have ownership over their contributions to AI. Even if their work is made “other” by the time a large language model has processed it, people should still be compensated for their contributions. Moreover, recognizing people's contributions to AI may be necessary to preserve humans’ sanity over time, as suggested by Lanier in his New Yorker piece. It will be important for stakeholders to continue to work on finding solutions to the challenges of implementing data dignity.
Conclusion:
Data dignity is a concept that has gained traction as AI becomes more ubiquitous. It proposes that people should be paid for their contributions to AI, recognizing that their data is becoming increasingly valuable. While there are challenges to implementing data dignity, it is an important concept that recognizes people's ownership over their contributions to AI. As AI continues to reshape the world, it will be important for stakeholders to continue to work on finding solutions to the challenges of implementing data dignity.