Can anyone answer how this setup prevents a computer science education? They take someone who's not going to college, gives them an incredibly valuable skill and the discipline required of a big person job. They are taught how to work. It puts money in their pocket. They code for 3-5 years, and if they hit a ceiling in the career, they PAY FOR THE CS PROGRAM with the money they saved.

In addition to having grown as a programmer (and finding out firsthand if they have a passion for programming), they are taught the academic theory behind the craft right about the time where they could not advance further without the degree. Not all programs require revolutionary search algorithms. Sometimes a business just needs a website.

To me, this setup makes much more sense than being taught the theory while accumulating debt, then shelving it for 5 years before anyone trusts you to code something outside a framework or a tech lead's supervision. It also makes a lot more sense then teaching someone the ideals of computer science only to slam them into the reality of the working world and burn them out. And hey, the work ethic of a competent programmer makes studying seem like leisure.

Of all the arguments against tattoos, I feel like the "what about when you're 80, 90, etc." argument is COMPLETELY invalid. Allow me to offer the following thought exercise:

Think of your sweet old grandma. Perhaps someone else's sweet old grandma if need be. Image how awesome she is, sitting there knitting a sweater or baking cookies. Maybe she's going to garage sales or agreeing with Fox News, as grandparents often do. Now imagine your stoic ancient matron with an equally ancient, ferocious tattoo. WHAT?! HOW AWESOME IS THAT?

Seriously, the more ridiculous or dated the tattoo, the better it makes this scenario. A prohibition themed ink would be amazing. "We like Ike" would also be pretty great. Are you going to tell nana that her flamin' buy war bonds skull makes her look unprofessional? I DARE YOU.

At the very least, grandpa's 311 tribal takes a lot of steam out of his "what's wrong with kids these days" rant.

Let's make an assumption that the internet will eventually solve any problem you throw at it given enough time...

What if rather than working on the next best CAPTCHA system, sites were to work from a rotating CAPTCHA repository?

Each page load presents a new human interface problem, something simple like a jigsaw puzzle or an image of a tic-tac-toe board with instructions to place an X and an O in a winning/defending position. In addition to each visual directive, there could also be a random text directive inserted to compound the problem (i.e, saying something like "after selecting the item, wait at least X seconds before clicking X button.)"

If your thinking in pseudo-code, the parsing of the text input isn't particularly challenging, and something like the tic-tac-toe is a solvable image problem, given time. However, if the captcha is being drawn from a growing database of imaging problems/verbal directives, then the captcha becomes not only solving the captcha, but identifying what kind of captcha is being presented.

As the captcha count increases a spammer/coder would have less and less time to hit the moving target and distribute their script before the next problem appears. This doesn't solve the problem of 3rd world captcha farming, but at least people might eat as a result of that economy.

This seems to me like a viable solution for the time being, though I'd like to refer to my first assumption for the long-haul.