I’m probably one of many people who found your podcast through one of the aforementioned Google search. I’m actually pretty disappointed as this is my first visit to your site/podcast. Your answer to the SSN question showed a lack of basic scrutiny regarding the question and I’d love to hear more research on the subject. There are a few points that I’d like to make that hopefully you can follow up on and give some answers for. First, there are not 1,000,000,000 valid codes. There are actually only about 880,000,000 valid codes, as there are many SSNs that are not valid, for instance ones starting with 666-, or ones containing -13-, or all 9 digits being the same (e.g., 222-22-2222). These plus other restrictions invalidate about 120,000,000 numbers. Therefore, we’re just about halfway there. Second, the idea that the SSA issues 5,500,000 SSNs a year may be true now, but there’s no way it’s going to stay constant for the next 100 years. Even if we maintain a constant population of about 300,000,000 citizens with SSNs, within 120 years (approxmately a lifetime and a half), we will be out of numbers. According to the US Census, a middle-of-the-road estimate puts the US population at about 400,000,000 by 2050 (just 40 years from now). If we maintain a constant birth/immigration rate over that time, we’ll be handing out nearly 7,000,000 numbers per year by 2050. By 2100, that number may easily reach 10,000,000/year. While it is unlikely I will be alive 120 years from now, it’s not entirely unreasonable to believe I’ll be alove 90 years from now. My 115-year-old self will not want to deal with the annoyance of having to overhaul the SSN system, espeically if I’m still programming computers (think Y2K). Right now, us computer folks are facing 2 emminent “run-outs.” In 2038, the 32-bit clocks inside computers will roll over, and many computer applications will think it’s a 1901 again. Also, it’s only a matter of time before we run out of IPv4 addresses, so we are in the process of deploying IPv6, which requires a complete re-architecting of the internet, so it’s not exactly an easy fix. The good news is that we’re fixing these things “for good.” After all the 32-bit clocks are converted to 64-bit clocks, we won’t have to deal with the problem again for another 280,000,000,000 years or so (at which time I’m sure I’ll be dead), and with IPv6, we will go from about 4,000,000,000 valid addresses to about 70,000,000,000,000,000,000,000,000,000,000,000,000 possible addresses, which should last us for the forseeable future. I would like to see something similar done with SSNs. Just converting the system to 10 alphanumerics (instead of just numbers) would last us a couple hundred years at least. But I might suggest we do a 64-bit integer, giving us about 16,000,000,000,000,000,000 possible SSNs, and thereby not only covering us for the long term future, but also giving us room for a checksum, and making it hard for people to guess a SSN. Plus, represented in hexidecimal, it’s only 6 more characters and only 6 more choices per character. Anyway, I’d love to hear more on this subject, maybe with some more in-depth analysis on the mathematics. Thanks!
Since social security started in 1937, over 620 million numbers have been handed out so far in years. So, with that in mind, and at the rate of babies being born, we could run out by 2120 (or there abouts)