So will a robot be diagnosing and fixing your car when it breaks? Will it be designing new and better cars? Will it even design the automation systems that let those new cars be built more efficiently?

Will those robots provide legal advice to the companies that operate them? Will they provide medical care to your kid when he breaks a bone? Will they provide him with counseling when he's struggling emotionally? Will they give him a lesson on algebra, and then work with him one-on-one to make sure he understands it?

Will robots patrol the streets and keep you safe from crime? Will they prosecute people who hurt others, or preside over those trials to make sure everyone's rights are protected?

Will robots invent the next Google or Facebook, and will they code it up and design a nice-looking and intuitive interface for it? Will they entertain you from a stage or a movie set? Will they write new jokes for a stand-up act, write an original novel, or provide an author with advice on how to make his characters more lifelike?

Maybe McDonald's fry cooks will eventually be replaced with robots, but "labor" isn't going anywhere.

It's precisely *because* companies (and individuals) are greedy, that prices will come down as automation makes things cheaper.

Imagine that thanks to nano-assembly, a nice gaming laptop can be produced for $10. Even if the existing companies try to keep prices where they are, nothing will prevent some young hotshot who is looking to make his fortune from setting up shop and selling laptops for $15. He will do this because he wants to get filthy rich, and won't mind destroying the whole entrenched industry to do it.

Maybe he'll get bought early and cash out, and the buyer kills the tech. No problem. Someone else will just come along and do it again.

It's a myth that automation is bad because it leads to unemployment, but no-doubt that myth will be perpetuated here. Someone might even say "yeah it frees people up, frees them up to STARVE." Let's try to address that before it happens.

As processes become more automated, the things we want become cheaper because the cost of labor is the dominant cost in almost every business. This means people have more spare money available, and it will be spent on things that before would have been considered too wasteful. This creates new industries and new jobs.

At one time, people would have spent virtually all their wealth on food. Because of improvements in automation, most people in the U.S. now spend a small fraction of their wealth on food, and this leaves extra money for, say, entertainment. At one time, having many people devote their whole lives to entertaining others would have seemed hugely wasteful -- those people should be out gathering food, after all -- but the wealth created by automation means that it's now a reality.

Some folks also make the claim that the new wealth will be concentrated in too few hands, and most people won't get wealthier. That, too, is false: automation makes things so cheap that just about everyone ends up owning things like microwaves, air conditioners, and computers -- things that before were reserved for the rich. Here's a good explanation of this: http://youtu.be/OkebmhTQN-4

Those studies, mainly the famous CDC 1995 college survey that is often cited, are pretty interesting. What's most interesting is how many liberal blogs quote the results "20% of college students were raped!" without saying exactly which questions were asked or what counted as rape. I haven't actually been able to find the list of questions, but I'd be curious to see what it is. As I said in my last post, they typically include things like "did not get explicit verbal consent" (the sloppy makeout and grabbing at each others clothes is apparently not enough), and the biggest gotcha of all: alcohol.

If she had anything to drink, the way most of those surveys are worded, it's automatically rape. If the man was drunk too, then a logical person would conclude that they raped each other, but the feminist worldview doesn't operate that way. Men can only be aggressors and women only victims, right?

It's true that many rapes are not reported, but we also know that many reports turn out to be false (that wikipedia page cites some studies to that effect). Also, nearly all district prosecutors have an explicit policy of not prosecuting women whose rape reports are later proven to be lies. I'm not talking about the guy being acquitted, I'm talking about *provable* lies, where she says "Joe raped me at 7:55pm on Tuesday" and later they find security camera footage of Joe sitting in a restaurant at that time. This creates a huge incentive to lie without repercussions, destroying the lives of men as you go.

I've heard this before, and it's complete nonsense perpetuated by feminists. The "one in four" myth began with a famous survey given to college students, in which the girls were told that if they had a sexual situation where they weren't completely comfortable, regretted it afterwards, or the guy didn't *specifically* ask for consent (even if consent was strongly implied by her actions), they should count that as rape.

In most American cities, typical rates for violent crimes (including rape) are a couple dozen per 100,000 people per year. In more hellhole-ish countries, it's maybe an order of magnitude higher. If we believe your "one in four" claim, then American college campuses would be more dangerous than the most dangerous hellholes in the world! See wikipedia's page on rape statistics for more.

..and I'm going to ignore you and do what I wanted to do anyway. If my current tool doesn't allow it, I'll switch to one that does.

But insults aside, I think the way in which you said "you shouldn't be doing that" without even bothering to explain why, is part of the problem. You're basically saying that you know better than me, so much so that you shouldn't even have to persuade me, that I should just accept your superior knowledge. That approach never works, and it's pretty insulting to be on the receiving end of it.

If you want to advocate a better way or explain why "continue 2" is bad, you can certainly do that, but just trying to force your position upon everyone isn't the answer. If you try, someone else (in this case, php) will come along and give the people what they want.

I am one of those "stupid" PHP users. It's not the first language I learned, nor is it the last. I'm well aware of Python as an alternative for developing web apps, and I've tried it, but I really do prefer PHP.

One reason is flexibility in flow control. PHP has do ... while loops. Also, I can do "continue 2" or "break 2" if there is a loop within a loop, to continue or break at the outer loop. I'm honestly puzzled that Python still hasn't added these obvious and useful things.

I've asked Python types about this, and gotten reasons like "you shouldn't be doing that in the first place," which pisses me off and makes me want to stick with PHP even more. I mean, who the hell are they to tell me what I should and shouldn't do? This tradition of "I know better than you do" seems to permeate Python, and leaves a bad taste in my mouth. PHP gives me the freedom to do what I want -- and that includes doing mysql escaping badly. That's not a freedom I will use, but it's there nonetheless.

People are right to criticize PHP's mysqli library, but there are plenty of good wrappers available that hide the problems, or you can just use prepared statements exclusively and avoid the whole mess. Again, PHP doesn't preach at you with a "I know better" attitude, it just gives you all possible choices and lets you take it from there. PHP tutorials are wrong to teach mysql without teaching security and injection prevention, but that doesn't take away from the language -- there are bad Python tutorials out there too.

Also, PHP's tradition of having a standard installed system makes it trivial to ship code. If a certain class isn't part of the standard PHP installation, you package it with your app -- and you know for sure what your user's environment will look like. True, there are the configurable php.ini settings, but those are trivial to check and adjust within your code. With perl or python, one shared hosting company might have certain packages installed while another one does not, so companies shipping apps go insane trying to figure out what they're allowed to rely on. Users without root might, or might not, get to bring in certain things on their own. And if they can pull in the right packages, the process might be too complicated for the given user, or might require shell access which the hosting company disallows.

The fact that there are so many successful PHP apps running on shared hosting (wordpress, vbulletin, etc) is a testament to how simple the language made it to ship and deploy such apps. It's historically been easier for a webhost to support PHP than to support Python, so more of them do it. From the developer's perspective, you know a given PHP app will run the same on all installed sites since they all have the same base libraries. From the user's perspective, it's dead simple to deploy the app with nothing more than an FTP client. No need for root access or any weird tech stuff. From the webhost's perspective, a trained monkey could deploy PHP on a server -- not just today, but 10 years ago.

I hate to put arguments in my opponents' mouths, but so often I've heard Python types attack the quality of those PHP apps. In my opinion, though, you can't argue with their success. PHP delivers what users need, and in the end that's all that matters.

For anyone who's still scratching their head about the instagram purchase, here's something to think about.

At least one of Instagram's employees, Philip McAllister, was at Gowalla when it was picked up by Facebook less than 6 months ago.

That guy might be the luckiest bastard in the world, having worked for 2 tiny companies whose only significant accomplishment was getting acquired by Facebook. On the other hand, Zuckerberg could just be funneling company money to friends?

I figured the same thing. At least one of Instagram's employees, Philip McAllister, was at Gowalla when it was picked up by Facebook less than 6 months ago.

That guy might be the luckiest bastard in the world, having worked for 2 tiny companies whose only significant act was getting acquired by Facebook. On the other hand, Zuckerberg could just be funneling company money to friends?

I've done recruiting and interviewing for more junior developer positions, and it's exactly what you describe: virtually no female applicants.

I'm all for diversity, but what exactly was I supposed to do?

That NPR article is pretty outrageous. He writes that everyone tried to get their gold at the same time and there wasn't enough to go around, therefore the British government decided to stop giving out gold altogether!

What I see here is criminal fraud, perpetuated by the government. The paper bills were supposed to have been claim slips for gold, with a government *promise* that if they ever wanted to claim the gold, it would be there. The government of Britain clearly broke its promise, as there wasn't enough gold to cover all the claim slips that had been issued. Rather than own up to this mistake and make reforms to avoid this in the future, they abandoned the gold standard to make it easier to perpetuate such fraud in the future.

Well said! Here's a neat graph that hammered this point home for me: http://whyisupportronpaul.us/

They talk specifically about the rising cost of gasoline, and how that is 100% due to inflation, whereas the actual cost of gas hasn't gone up at all over the last 40 years!

No, but crime lord types probably will be smart enough.

When I wrote the original post, I was picturing a TrueCrypt volume in a file rather than full disk encryption. You create the file volume, say 2 GB in size, and fill half of it with grandma's recipes. Then you create a shadow volume within the 1GB blank space of that file, and put the incriminating data in there. They'll find the TrueCrypt volume/file on your computer and order you to provide the passphrase, so you decrypt the recipes. File access/modification dates won't be an issue -- it's been a few months since you last baked grandma's chocolate chip cookies, that's all!

Disk encryption software already supports hidden volumes. Even if this kind of decision becomes dominant case law, that won't accomplish anything. People will just start deploying volumes with two passphrases, and when ordered to give up the passphrase, giving up the one that decrypts grandma's recipe collection.

Since there's no way to prove that a second volume exists within the blank space of the first one, encryption will win the day.

Damn straight! I wish I had mod points right now..