notes
honestly, this didn’t get really interesting until the last 50 pages or so. there’s a couple of hundred pages of setup on what constitutes a hack, examples of “hacks” across the spectrum of society, etc. there was a bit of less than obvious discussion around how some “hacks” become entrenched into the law, how common law interacts with these “hacks” and a generally slow/ossified capability of our governmental/legal systems to address the use of said hacks.
it did get interesting when he started digging into the use of AI as a mechanism to facilitate deeper social “hacks”. this was an angle of AI that i hadn’t given much thought to.
put bluntly:
Al systems will themselves become hackers: finding vulnerabilities in all sorts of social, economic, and political systems, and then exploiting them at an unprecedented speed, scale, scope, and sophistication. It’s not just a difference in degree; it’s a difference in kind.1
the rub here being the very wealthy have the ability to fund and use these AI systems to explore the problem spaces of tax evasion, loophole finding for less than “legal”2 means of running businesses, etc. this discussion kind of putters out while tossing some real meat on the table. (to be fair, i suspect that his writing process has been a contributor to some of the dangling discussion3.) there’s a scant few pages that nod at the use of defensive AIs to counter these offensive AIs, but alas this doesn’t go down the rabbit hole of offense here. there’s a hugely interesting vein of research and discussion here.
references
meta
- location: Minneapolis, MN
- weather: 36°F and Cloudy