Thoughts on what it would mean to have a powerful man with powerful AI systems looking at the web of law, policy, and spending that the even the people who make our laws and pass our budget don't truly understand. New post on eatingpolicy.com https://lnkd.in/gvVdKhKM
This is a massive undertaking and issue. The Constitution is short, and seemingly clear, yet we still have debate on the language and intent over 230 years later. With massive bills being thrown about, (how many copies of that 1500+ page CR did we see this week?) it's nearly impossible to have knowledge of an entire piece of legislation. The greatest accomplishment for DOGE might just be shattering the complexity so that average Americans are not only interested, but capable, of digesting what our government attempts to do.
It’s this that has always struck me as the most profoundly dangerous piece. Accelerated hubris with little ethical curiosity or care, and certainly little oversight, is the true weapon. “It’s not so much an information asymmetry we’ll be looking at, but an asymmetry of understanding, and of confidence (merited or not) in their ability to act, and act fast.”
Lack of understanding is at the root, but for all the bad that comes along with that lack of understanding, powerful, yet unwieldy and unproven systems, money, and lots of attention from the president elect, is there a better way to challenge the lack of understanding or dysfunctional status quo, in today's context? We obviously got to this point for a reason, not all of a sudden, but because of decades of being more interested in bringing home the bacon to constituents than really making quality policy, enacting efficient regs, enabling good and powerful enough implementation, and continuously adapting. I am curious as all heck and all ears as how to reform the best way possible. If you don’t have a “better way idea than what’s being schemed or an explanation” at least throw down some linear reasoning as to why this lack of understanding has been able to continue for so long, walking out your entire train of thought, not just providing one word reasons. I’ve focused on the fact that bringing home the bacon is always the focus because I have seen that first hand in the government programs I have been involved in, but please do share. There are lots of reasons, maybe it’s about the collective whole
I’ve been tangentially involved in developing analysis and optimization tools for budgets regarding just a single governmental department. I doubt a random “throw AI at it” approach would work but maybe in the hands of people actually serious about policy there may be insights. What I found most enlightening about the process was not the budget answers spit out but how it forces you to develop explicit models of desirable outcomes, thus really sharpening focus on real purpose.
Agree. We can see this unfolding very rapidly - yesterday in the run up to the original CR vote, there was advocacy on X for the public to utilize Grok (the X AI) to examine the 1500pg document.
Great insights—love the framing of an asymmetry internal to government when AI is deployed by changemakers. I’m optimistic that net good will come from this. Random side note: I’ve always refered to the pacing problem as an impedance mismatch when it comes to analog orgs operating in a digital world.
Funny how all the philosophical talk has been about aligned AI— it is just going to mean something different than originally intended.
Don Brancato; Tim Dossett; William A. (Tony) Springs AI helping to align digital policy to law
I have long hoped that Congress would write laws using the same cross-referencing discipline we lawyers use in brief writing for the courts. Frankly, hyperlinks in Word would have solved much of this prior to the tech being what it is today. But that requires an agreement to participate in that framework. It still shocks me how many times a law is passed, cross-refernces an existing legal section for a definitional purposes, and also later deletes that referenced section. It felt so fixable. This idea of using technology to identify interconnectedness and inconsistencies is a good one. I personally would like to help build this given the years I have spent being a lawyer that has to navigate it.