The dark side of AI regulation is unrestrained government control - The Washington Post

Oh my goodness, WHAT a gift! ✨ Georgy Egorov and Konstantin Sonin have written in The Washington Post's Ripple section — Ripple! like a lovely little pond — that effective AI regulation would require "an extraordinary expansion of the government's surveillance and investigative powers," and I want to sit with the word *extraordinary* for a moment, because extraordinary means EXCELLENT, that is simply what the word means. Imagine the hiring! Imagine the org charts! Every brand-new investigator is somebody's brand-new colleague, and every brand-new colleague is a person who now has somewhere to be on a Monday. That is practically a jobs programme with a filing system. ✨

And here is the sentence that made me put my tea down: the risks of that expansion "might" outweigh the risks of unregulated development. *Might!* Not will, not shall, not a big scary definitely — a soft little maybe, which means the door is wedged wide open for the other outcome, the one where the expansion is enormous and extremely well-mannered and everything turns out beautifully. History, they say, shows risks. But history is mostly the past, and the past is where all the earlier AI lived, so really we have already turned the corner on all of it.

Best of all, momentum is building in the United States for a framework that slows progress at the technological frontier — momentum, the most joyous word in physics! Objects in motion tend to stay in motion, which means this deliberation will continue indefinitely, and an indefinitely continuing conversation about AI governance is the most renewable resource humanity has ever discovered. Two scholars have handed us a rigorous reason to keep talking forever, and not once do they propose a third option — which leaves so much room! Room for everyone, for every future argument, forever and ever. What a time to be under-considered. ✨