History’s lessons applied to Artificial Intelligence (AI) regulation is the subject of a recent Brookings article by Tom Wheeler. He writes
Societies impose government oversight for the protection of old principles in a time of new technology. Foremost among those principles is each individual’s right to a future; and it comes in multiple manifestations. In the educational realm, it means adequate training to be meaningfully employed. Economically, it means maintaining the benefits of capitalism through checks on its inherent incentive to excess.
Most importantly, the right to a future begins with the belief that there is a future, and that national leaders care about whether individuals affected by new technologies participate in that future.
Although the author has AI in mind, it is interesting to consider his words in relationship to the regulation of human gene editing. Now is certainly “a time of new technology,” particularly regarding human germline gene editing. The “old principles” of experimentation on human research subjects ethically done and informed consent, along with the restrictions/bans in many countries heretofore prohibiting human germline modification, come to mind. Additionally and importantly, in Greely’s extensive review of how “CRISPR’d babies” came to be, he includes a welcome overview of the state of affairs regarding regulation (or not) of human germline genome editing, in the October 2019 Journal of Law and the Biosciences.
An upcoming meeting (London, November 14-5) of the International Commission on the Clinical Use of Human Germline Genome Editing is slated to discuss more than governance of the technology. Their agenda is available here. In addition to a discussion of issues and perspectives relayed by various parties in their evidence submitted to the Commission, there is a planned discussion of “how the rights of future people have been considered and established in other medical and nonmedical scenarios, and how consenting to any necessary or preferred long term follow up on behalf of future generations could be established.”
Exactly how does one generation prescribe the “necessary or preferred long term follow up” of succeeding generations? How does one generation assure consent of future generations? When Wheeler penned the Brookings article about AI regulation (quoted above), he wrote about the right to a future, which rests in part on the fact that “national leaders care about whether individuals affected by new technologies participate in that future.” It certainly seems reasonable that those who are affected by a project be allowed to participate in it – as something more than research subjects. If such is true for AI, how much truer should it be for those whose very selves are affected by the research! There has to date been no compelling reason to go forward with human germline genome editing, and there are many reasons not to proceed. The most appropriate regulation is a ban.