Software-defined regulation is a step on the way to AI pioneer John McCarthy's vision of governance through mathematical objectivity, or at least a variation of it. McCarthy suggested that using formal methods to filter out dumb ideas enables this scenario: if you have an idea about how the government can do something better, and can produce sound evidence that it is indeed better, then the government will make it happen.

We can take this a step further and computerize the oversight of both government and corporation activity such that anyone, regardless of their position in society, can keep them in check. Policy proofs would help uncover ways the government should behave, while "regulation proofs" would help uncover ways businesses should behave. AI and the public (or the AI-enhanced public) can then monitor these behaviors and look for possible deviations or improvements.

‍

Deciding vs Monitoring
‍

We're dealing with 2 concepts: deciding behavior and monitoring that behavior. McCarthy addressed the first (for government behavior), while Hemant Taneja addressed the second (for company behavior):
‍

"Algorithms can continuously watch emerging utilities such as Facebook, looking for details and patterns that humans might never catch, but nonetheless signal abuses. If Congress wants to make sure Facebook doesn’t exploit political biases, it could direct the FAIA to write an algorithm to look for the behavior.

The watchdog algorithms can be like open-source software — open to examination by anyone, while the companies keep private proprietary algorithms and data. If the algorithms are public, anyone can run various datasets against them and analyze for “off the rails” behaviors and unexpected results."

‍

Open-Source Policy-Making Software
‍

Like the watchdog algorithms, policy-making software would be open-source so the public can create their own proofs like the high school student in McCarthy's scenario. Some software might have the ability to run society simulations to test policy changes, much like the way Facebook rolls out features to subsets of its user base to gauge their reaction before making it a permanent change. In these simulations, you could perform randomized controlled trials like Y Combinator's study on basic income, but at much lower cost.

Policymakers, by themselves, do not have sufficient expertise to make decisions on managing technological progress. They will sometimes establish an agency comprising subject matter experts to help guide decision-making.* However, balancing the interests of the government, the public, and tech companies may require an ability for all parties to put forth their own ideas, bolstered by AI- and open-source-software-assisted analyses of current behavior and simulations of future behavior, and have them reviewed by a policymaker for future implementation.

‍

A Tough Question
‍

During Miss USA 2018, Miss South Dakota was asked: "During the last presidential election, more than one-third of women neglected to exercise their right to vote. Why do you think that is?"

Her answer was "I'm not sure" and something along the lines of "that should change". I don't think I'd have fared any better but this uncertainty might have contributed to her not making it to the final 3.

A more complete answer would have been, "It could be because women feel like they don't have a say in political matters. Despite one of the candidates being a woman, no matter who got elected, they would break their promises and disappoint the people who voted them into office. I propose a new approach to policy reform: software-defined regulation. If you can leverage formal methods to produce a rigorous argument in favor of your preferred policy, and because of this, policymakers listen to you and enact change, then you will be more motivated to take an active part in choosing our leaders. Thank you." *applause*

Who knows, this might have propelled Miss South Dakota to the final 3 or, at the very least, made some people aware of the possibilities.

* Such as the Nuclear Regulatory Commission, which oversees the use of nuclear energy in the United States.

‍