Apparently, It Is Now Necessary To Use Artificial Intelligence To Reform U.S. Federal Regulations

By Darren Smith, Weekend Contributor

It seems that the regulatory web that envelopes United States federal regulations has grown so complex and gigantic, reform necessitates the use of artificial intelligence to tame the dragon.

Reuters reports the White House Office of Management and Budget last Friday announced that federal agencies will use Artificial Intelligence technology to “eliminate outdated, obsolete, and inconsistent requirements across tens of thousands of pages of government regulations.”

The project follows success found in 2019 using Machine Learning and Natural Language algorithms with software at the Department of Social and Health services in identifying hundreds of technical errors and outdated requirements in agency rulebooks

In a manner of speaking we have reached a point where the regulatory morass was allowed to become so formidable, that ordinary human-powered rule making is no longer capable of restraining or modernizing the red tape.

The United States Code of Federal Regulations contains about 185,000 pages of text and “is the codification of the general and permanent rules and regulations (sometimes called administrative law) published in the Federal Register by the executive departments and agencies of the federal government of the United States. The CFR is divided into 50 titles that represent broad areas subject to federal regulation.” (wikipedia)

One could call it the DNA of the federal bureaucracy. And in a way it seems that analyzing this DNA now requires complex software just to fully understand it. On a tangent, I remember an interview with a software developer who recently celebrated his thirty-fifth anniversary with Microsoft. He noted the growth of the Windows operating system over the decades and that it had evolved to such a size that no individual was capable of fully understanding the inter-workings of the entirety of the system, that is each of the underlying parts. He said the last version to facilitate such a capability was Windows NT 3.1 (Released to manufacturing in 1993) and even then there were only four individuals within the kernel team who could claim such an understanding. But with today’s operating systems, the scale of the software is too vast for any one person to fully comprehend. The only practical and realistic expectation now is for individuals to fully understand individual components, know them well, and successfully interoperate with others commanding a similar capability in adjacent systems. Yet there were stuggles in the Windows Division itself in maintaining cohesion in developing a common operating system of increasingly complex subsystems. Software tools can be very useful in assisting in creating or maintaining a complex system but it is not a complete substitute for human networking and problem solving skill.

It seems now the regulatory base of the federal government presents such a challenge.

While I do give credit to the government for having the foresight to bring novel and useful tools to accomplish a goal, we truly have to reexamine how we came to this point in the growth of regulation. And it brings up many other questions such as “is there a legacy cost of having to maintain old rules that have become so complex and archaic that in some ways it is almost less trouble to let it remain rather than fixing it?” or “how can we expect ordinary individuals or small businesses to adhere to regulations when we have to bring in artificial intelligence to recodify the regulations from time to time?”

In returning to the software analogy I have to look back at what happened in Washington State when agencies such as the Department of Licensing are increasingly facing legacy cost of its 1980s and older vintage systems that are so old and outdated the state needs to bring in outside consultants to fix it from time to time. The fact that it is slightly cheaper or less work to maintain the software status quo means that the can just keeps being kicked down the street for someone else to deal with this immobile legacy. Perhaps it is also the daunting fear of having to take on what is perceived as a herculean tasks to start over from scratch with a completely reorganized system that proves dissuasive. Yet the promise of a foundationally better methodology and schema of such a system could bring a far more efficient and effective program might be a bit too scary. And given the spectacular failures that have occurred from time to time when a state or federal agency re-architects one of its Information Systems, there might be some additional fear to be overcome.

I wish them success in using this tech to resolve the problem. But it’s going to take much more than fully relying on software that parses through the CFR to trim out a lean, mean, lex machine of regulatory codes. But I suspect that in the end it’s going to come down to a great many human beings needed to do most of the work. Hopefully there is the will to do this.

By Darren Smith

Photos: Brazil (1985 Film, directed by Terry Gilliam). A movie I highly recommend.

The views expressed in this posting are the author’s alone and not those of the blog, the host, or other weekend bloggers. As an open forum, weekend bloggers post independently without pre-approval or review. Content and any displays or art are solely their decision and responsibility.