The US government should create a new agency to provide technological research and oversight
Response to a prompt about one specific policy idea related to AI that the US government should pursue for a job application. Written 2021/09/14
This agency would be tasked first and foremost with the responsibility of educating the public. The organization could publish research focused on the current impacts of technology and expectations on its future impacts. Based on this research, the agency would be able to hold official positions (i.e. make statements not only describing the state of the world, but also stating what is preferable).
This would serve to legitimize the space of technology and AI research, and provide a source of reputation for literature in the space. This would also serve to familiarize citizens with ideas and concerns related to technology and AI outside of the frequently held conversations (e.g. Big Tech, autonomous vehicles).
Ideally, a new agency of this kind would also have some form of regulatory power. The primary priorities of this regulation would be to reduce the likelihood of catastrophic scenarios, and regulators would operate judiciously. This way the agency would serve as a complement to existing work done by the FTC, and not necessarily be focused as much on anti-trust concerns. In addition to reactive regulation, this agency would also look into how best to create a proactive approval process to manage large scale AI related technologies, and eventually implement that approval process.
Finally, this agency would also be tasked with engaging the United Nations. The goals of this engagement would be to both to promote a set of shared global standards related to technology and AI, as well as to create an intergovernmental panel on technology. This panel would be focused on providing unbiased, globally approved technological research and analysis (like the IPCC).