FEMA trailers are heavily treated with preservatives that can make people sick after lengthy occupation, they're comfortable enough otherwise that people may not make enough effort to move on from them, and they are expensive to produce and expensive to maintain. Mobile homes are a great solution for first time home owners who want to build their management and maintenance skills prior to owning a more risky and expensive brick and mortar home, but they're not a good temporary emergency shelter. But let's extend the practicality of this purpose. Why would you want a non-FEMA shelter when you're not homeless due to a circumstance that FEMA would respond to? Because you don't have the FEMA option when your company shutters, home burns down, or you're bitten by eminent domain, and other parachute options may not suffice.

Of course, I guess you could hand wave all that away by pretending that only drug addicts and crazy people are ever homeless outside of disasters.

But you're not getting the point even in a general sense. Ask yourself why you'd want to have your own emergency shelter handy rather than get herded into a FEMA camp. Or ask yourself what use you could have for a small, only semi-permanent space that can be heated and cooled, perhaps in scenarios where building is not an option. Such as when you rent land or are subject to homeowner association or regulatory restrictions about permanent land improvements. There's also a matter of room segmentation, whereby a sleeping space in this form could provide enhanced compartmentalization and privacy.

Finally, ask yourself whether a company that apparently aspires to make disaster relief more manageable has a better chance of achieving that goal when they minimize the probability of a sale. When it comes down to it, if they can't pay the bills then they won't accomplish anything. So, cutting themselves out of market costs them and also tarnishes their image because they send a bad message to the same people they'd like to put in these things if it's ever necessary.

They are cutting themselves out of market reach by excluding consumers. Their success or failure depends entirely upon whether organizations, wealthy individuals, or municipalities will order large lots. People with deep pockets don't spend on impulse, and they're just as likely to create their own solution as invest in this one.

Meanwhile, personal responsibility in preparation for potential future emergencies is countermanded. An alternative to homeless camps is prevented. Applications beyond emergency housing are completely nullified. And the company cuts itself out of profits. This seems to be what always happens with emergency shelter. Either it's priced such that one could buy something in the range from an old mobile home to small house, or it's simply not available.

What is the difference between selling a 25-40 unit lot and taking 25-40 consumer orders before beginning production? This company could give itself six months per order for enough orders to be reached to justify production and then give the consumer an option for a refund. It wouldn't even be necessary if they priced it reasonably such that better solutions aren't also more cost-effective.

The engineer saw Katrina victims and wanted to solve the problem. Bullshit. The Katrina victims wouldn't have had access to this, by design.

Jesus, I just can't think of everything I need to say with one post today.

Let me give people two things to think about that could help immensely with this. First of all, our brains use statistics to model the environment, to include models of others and ourselves. But that does not mean that our brains are not fundamentally deterministic. Secondly, strong AI is not only possible and attainable but very well may be achieved in our lifetime because WE, as in all of us -- YOU, reader -- are strong AI. You are nothing more than a machine made of meat, even if you romanticize your ego with ideas about a soul.

There are two things that this field needs. The first is continued development of self-programming machines. Neural nets are the most primitive way to do this. Traditional hardware is better. But it's not fast enough yet. The other thing is that we need the means to convert between lambda calculus and traditional calculus, treating number of iterations as an independent variable and all other changing quantities as dependent variables. That branch of mathematics is far, far too neglected.

We should marry these tasks. Our self-programming machines, neural network or otherwise, should be helping us to further lambda calculus as a branch of mathematics, from the most fundamental level of raw logic and up. If that began now, then I would give us twenty years before fully sapient machines exist and thirty before our own minds could be uploaded and linked with second brains. If it continues to be neglected then we'll piss all over ourselves in fear and excitement, fantasies and nightmares, but ultimately produce nothing that any of us need to worry about outside of concerns in economics.

I'm going to throw away my mod points to comment about this. The idea that neural nets or quantum mechanics are required to produce a sapient machine are wrong. And not even a little wrong. Those perspectives sound easy to buy but they completely miss the point.

First of all, the argument that deterministic machine code can not simulate intelligence, used as support that an intelligent machine must be mechanically stochastic, completely ignores the simple fact that stochastic software can be achieved using deterministic hardware. Statistics is not a magic that transcends causality. It is a way of describing causal systems that we do not yet comprehend. I know there are those who will disagree with me about that, but they have lost faith that humanity will ever understand the mechanics currently described using statistical mechanics. Humanity has already proven them wrong time and time again. Their own failure should not be projected upon the rest of our species.

I'm sorry if that sounds harsh, but reality and in particular nature are not in the habit of caring about individual feelings.

Second, let's think this through. Suppose that a neural net self programs functionality. It still does so by adapting nodes such that they work together to achieve the desired outcome. The manner in which they work together -- the actual, physical processes that take place as signals are transmitted and received, information processed -- are purely functional and deterministic. They can be mimicked with traditional hardware. The benefit to neural nets is that they can develop their own software, which can save time. That too is possible without them. The fact that it has not yet been achieved does not make it impossible. It's merely a symptom of the fact that programming languages are not designed for that purpose and overcoming their limitations to adapt them to this task is non-trivial. Metaprogramming is the first rudimentary step to changing that.

Third, let's talk about human foolishness. Using systems that describe things we don't understand or things too complicated to model with classical mechanics is not a magical band-aid to overcome our limitations in the act of creation. They're tools of analysis. They help us to understand. If we depend upon those approaches to create intelligence, then we will by extension create intelligence that we do not understand because that fuzzy cudgel will be an axiomatic foundation of the system. That is not only short-sighted. It's fucking dangerous.

I am not an expert in this field, but I am 34 years old and I have spent my entire life thinking about AI. It is simply something that has always been in the back of my mind. I don't know why. And this has been since I was five or six years old at least -- maybe sooner. I know how to create a strong AI. And I know that people can easily create a convincing simulation. Wozniak, Gates, and Musk are absolutely right to worry; not because danger is an essential property of strong AI but because people are shallow, are accustomed to doing just good enough to pass, and are more than willing to attribute their own characteristics to the entire universe. Again, nature doesn't care.

I will be an expert in the field. I'm a hell of a lot closer than the vast majority of people. And I can tell you right now that panic is premature. We're playing with pebbled and worrying about boulders. It's important to have the philosophical framework for this before it is achieved, but that is not what is being accomplished in any discussion I have ever seen on this topic so far.

I should expand upon this more explicitly. The only evidence ever produced to support deterrence theory has been statistical. Statistics and probability go hand in hand, and with them is entropy. This is simple. Let's count the ways that deterrence theory works and the ways that it fails.

Ways That It Works

1. An aversion to incarceration causes somebody to choose to avoid breaking the law.

Ways That It Fails

1. Somebody assumes they won't be caught. 2. Somebody considered incarcerations to be a cost of doing business. 3. Incarceration is comfortable to somebody. 4. Somebody believes they are not breaking the law. 5. Somebody believes a criminal act should not be criminal, and are willing to stand up for that belief. 6. Somebody knows that they will be incarcerated for something worse they've already done. 7. The consequences of not committing the crime would be worse than the incarceration. 8. The law in question is seldom enforced. 9. Somebody actually wants to be incarcerated. 10. Somebody has been exposed to an environment where a crime is regularly committed by others, without consequence.

There are at least nine more ways that deterrence theory fails than succeeds. For each of those ways, depending upon the actual crime in question, additional motivations, confusion, compulsions, etc further complicate these items such that on a crime by crime basis the individual act's details serve as a multiplier for the ways that deterrence theory fails. Meanwhile, there is still one and only one way that deterrence theory succeeds.

It's a failed policy. But it's a lucrative, failed, barbaric policy, so there are those who like to pretend otherwise. Rehabilitation, retribution, and separation from society are the only valid purposes for incarcerations. Any ideas about precrime law enforcement, to include deterrence theory, would be nice but are in reality nothing more than fantasies.

I should add an explicit statement here.

There IS a problem with the genders being unequally represented in the software industry. But aside from sexism, there's also nepotism, cronyism, and a culture that wants developers to be overqualified. As a result, MOST programmers work for themselves, if they program at all. Many give up. So, don't expect those who are in the same boat to grow a money tree and fix things. Instead, be the new software company that will do it right.

I work alone. I'm still at the point of writing boilerplate I'm very picky about, and it will take time before I'm doing anything for an end user. When I hit that point, I will probably still work alone because building a commercial product with the possibility of later pay rather than the liability of payroll isn't really the easiest thing to organize. I would LOVE to have some talented programmers on my team in times coming within the next year, but they won't be no matter what their gender is.

HOWEVER, I have had some serious thought about opening my source from square one when the boilerplate is ready. It has so many possible applications that there will be real value in what follows it, and it's incredibly unlikely anybody would do the same things with it that I will. Suppose that following that step, programmers decide to jump on board and help with the next step too. Great! Not working alone would be nice. But I can't select their genders. I don't get to pick that.

I like the idea of Bechdel Test for programmers, but I think it's kind of important to bear in mind that only well monied developers even have that option. My suggestion to female programmers who have trouble finding a team is to do what I'm doing. If you want something done right, you have to do it yourself. In my case, it's because I have an idea that I want to do the work to realize. But there are plenty of developers who freelance for other reasons.

Yes. I am seriously suggesting that jail has no deterrent effect whatsoever. That's why the marijuana usage rates per capita have not changed at all since prohibition began. Deterrence theory was acted upon with policy before there was ever any evidence to support it, and to this day the only "evidence" I've ever seen has been abuse of statistics. The only reason that way of failing to do things has persisted for so long is that it makes money for jailers.

There's a cabal of irresponsible (at best) and insane (at worst) organizations that publicly disclose the private information of citizens and members of government to further their own financial and political ends. We call them businesses.

While doing that, we might, I don't know, actually allow people some measure of privacy online. daesh is only giving crazy people a new way to be crazy, and having our information made public by corporations and government offices who apparently don't care if they can get people killed only serves to help crazy be crazy. Where veterans and service members are concerned, this is even further simple insane irresponsibility in the name of being able to target an add for inconsequential widgets.

As much as I'd normally step up to defend Obama, I don't think this would ever be a priority under his administration. They're all about selling the idea of how our information can be spread around to everybody. And this is why I say that the GOP is technologically illiterate. If they weren't, they'd have used this vector of rhetorical attack for years now.

It's not even a matter of standing up to them. It's more that choosing your battles is kind of an important skill, no matter what your goals are. That's fundamental. It comes before choosing good or evil, support or opposition, stripes or plaid... Though if anybody ever finds a way to somehow force everybody to be the picture of class without undermining our laws, then they should let me know. Because I'd totally have other applications for that kind of mass scale mind control device.

It's his project, his online handle to ruin, and he can call anything in it whatever the hell he wants. Maybe everything in the project is named after dicks because he doesn't want some company taking it out from under him before it's finished.

Not to say that it's not immature and in poor taste, but Christ, what's next? Are people going to complain because some strangers use potty mouth in notes they write to themselves? I called myself a pecker once in an email to myself that contained a URL I kept forgetting. Does that make me sexist?

Stuff like this undermines real issues of sexism. Please, if you don't like his naming conventions, create software to do the same thing his does and be professional about it. You don't get to have someone do your work for you *and* complain about aesthetics that the end user won't even see.

He needs to grow up, sure. But I don't expect better from some random online handle. People who apparently stand for equality and social justice though? I kind of expect you guys to have better sense than an angry twelve year old.

This isn't just silly childish pranks. Property can be destroyed and people killed in SWAT raids. It has happened before.

These kids know what they're doing is illegal. So, it's not going to deter them when people go to jail for it. The only thing that will stop it is if the police ask for stream URLs and actually check before kicking in doors. You know, act like reasonable people. They can even check while the SWAT team gears up, so it doesn't cost precious time in the case of an actual emergency.

You can't compensate for SWAT-happy law enforcement eager for every little chance to kick in doors by putting away the people who give them that chance. The problem has to be fixed at both ends by punishing the people who do it and requiring that our officers operate as if they've got a shred of common sense. "Oh yeah? What's the Twitch handle?" How hard is that?

Sure! And the Unibomber was just a year-round secret santa.

So, something has to be punishment to be a bad idea? It can't just be stupid and irresponsible? Yeah, that sounds like the kind of technicality baiting our government operates under. But that's where your correctness ends, down to the letter. I'll just say that outright.