Have you ever noticed that webpages based on the other side of the world are much slower than pages closer to you? I don't blame you if you haven't; the difference is on the order of tenths of a second. But if you think about it, that's pretty incredible. Foreign countries are far away, but on an electronic timescale, they may as well be next door. The world may not be flat yet, but its derivative is certainly decreasing pretty rapidly.
We probably won't ever eliminate distance completely. We may not need to, though. Studies have shown that any action that takes under a tenth of a second or so is perceived as instantaneous; whether it takes a nanosecond or a hundred milliseconds, we won't notice the difference. Eight orders of magnitude is an awfully big range to have just have disappear.
When technology scales this far beyond our ability to follow, it changes the rules of how the world works. Scale loses all meaning toward the low end; big blurs into small, long blurs into short, expensive blurs into cheap, which blurs into free. Not everybody is thrilled about this. As a rule, disruptive changes are bad for established players in markets, and they fight the changes as hard as they can.
Take ebooks, for example. I have an ebook reader, and it's a marvelous thing - I can carry around thousands of books on a device that's small enough to fit in my pocket, but large enough to read comfortably. You'd think that publishers would be all over this, right? Instead, though, they're sitting in a corner pouting, with their fingers in their ears, doing their best to pretend that ebooks don't exist. And maybe that's all they can do; the fact is, publishers exist primarily to publish books, and the marginal cost of printing an ebook is exactly $0.00. They've seen the writing on the wall, and they know that their entire industry was made irrelevant while they weren't paying attention, so all they can do now is ride it out while it lasts.
The fact is, it is now becoming increasingly economical to simply give things away for free. Open-source software is the shining example of this, of course. The idea that people from all over the world would spontaneously organize, make something interesting, and then give it away for free to anybody who wants it, is completely ridiculous. It could only happen in a world where communication is instantaneous, and marginal costs are zero. Software aside, imagine what we could achieve if everything ran like that?
But, the question still must be asked: Who is going to pay for this free-culture utopia? The internet isn't really free; its real-world manifestation requires constant babysitting to run smoothly. The question goes straight to the heart of the uneasy division between the digital world and the real world. (It might even be inaccurate to say "the digital world", for while information online obviously follows very different rules online than in the real world, it's also undeniably true that the digital world is a part of the real world as well.) Things in the digital world can interact with other things there, and things in the real world can interact with things here, but somehow the boundary still remains an awkward no-man's land.
There already exists one "digital divide", between the technological have's, and the have-not's. This is the other digital divide - the gap between us, as individuals, and the world we've created. We reach across the gap, and we can harness the convenient properties of the other side, but on a very fundamental level we don't really understand it.
Or, to summarize: The Internet is a weird place.
Friday, February 27, 2009
Friday, February 20, 2009
Credit Cards Post
Credit cards are absolutely, shockingly, embarrassingly insecure. Imagine for a moment that, whenever you wanted to pay for something, you handed your wallet to a stranger and said, "Here, take whatever you want." Actually, that's far more secure than credit cards. Imagine for another moment that anybody who's ever even touched your wallet had immediate access to all your money. Even that's a bit generous, because credit cards are actually a little worse than that.
This is the reason that identity theft is a problem right now - there's no security at all when it comes to credit cards. And the worst part is that there's no way for you or I to defend ourselves, because the insecurity is baked into the system. There's no way to use a credit card without handing it to somebody, or typing all the information on it into a website, or doing something equally insecure.
Of course, this is the part of this rant where people sometimes say, "Fine, so you could do better?" I think I can probably do better; I would certainly be hard-pressed to do worse.
First Proposal - Digitally signed transactions
The core of this proposal is an RFID-like device which can compute a cryptographic checksum, and contains a secret key. When a transaction is made using the device, the details of the transaction are fed to the device wirelessly, and it responds with the checksum of the transaction details, plus its own secret key. The details of the transaction can then be relayed to the credit card company by existing means.
Pros: This scheme is cryptographically secure. It guarantees (barring failures of the physical security of the card) that the card was actually present at the time of the transaction. It's also cheap (RFID tags cost pennies), and doable with current technology. As a bonus, it meshes well with ways people currently use credit cards, and is easy for people to use.
Cons: Doesn't really help if your card is physically stolen. Also, doesn't address online purchases.
Second Proposal - Multiple cards
When you go out, you generally have some idea of how much you're going to spend. So why carry more than that? If we had a set of credit cards, rather than just one, with various limitations, then this would be more secure. Limitations could be simple spending limits, such as per-transaction or per-day limits, or it could be something more complicated, like only allowing purchases in the city you live in.
Pros: Completely compatible with existing systems. Easy for people to understand. Cheap to implement.
Cons: Most of the downsides of credit cards still apply; this just mitigates the risks.
Third Proposal - Authorized transactions
It should be possible (if perhaps a little annoying) for your bank to call you and check with you whenever a purchase is made on your account. Since having a person calling you all the time would be a bit much, it could also be an automated calling system, or a text message, or an email if it's not time sensitive, or something else. The main requirement is that it be something that's difficult to fake.
Pros: Compatible with existing systems. Prevents any unauthorized transactions, even those made with a stolen card and all relevant data.
Cons: Kind of annoying.
(Note that nowhere in here have I mentioned biometrics. Biometrics look cool in movies, but they're basically useless for actual security, for reasons that I'll go into more in a future post.)
Security is always a tradeoff, and when it comes to credit cards we've always gone far, far to the extreme of preferring convenience. But it doesn't have to be that way, and if we were willing to give up a small amount of convenience we could make credit card fraud a thing of the past.
This is the reason that identity theft is a problem right now - there's no security at all when it comes to credit cards. And the worst part is that there's no way for you or I to defend ourselves, because the insecurity is baked into the system. There's no way to use a credit card without handing it to somebody, or typing all the information on it into a website, or doing something equally insecure.
Of course, this is the part of this rant where people sometimes say, "Fine, so you could do better?" I think I can probably do better; I would certainly be hard-pressed to do worse.
First Proposal - Digitally signed transactions
The core of this proposal is an RFID-like device which can compute a cryptographic checksum, and contains a secret key. When a transaction is made using the device, the details of the transaction are fed to the device wirelessly, and it responds with the checksum of the transaction details, plus its own secret key. The details of the transaction can then be relayed to the credit card company by existing means.
Pros: This scheme is cryptographically secure. It guarantees (barring failures of the physical security of the card) that the card was actually present at the time of the transaction. It's also cheap (RFID tags cost pennies), and doable with current technology. As a bonus, it meshes well with ways people currently use credit cards, and is easy for people to use.
Cons: Doesn't really help if your card is physically stolen. Also, doesn't address online purchases.
Second Proposal - Multiple cards
When you go out, you generally have some idea of how much you're going to spend. So why carry more than that? If we had a set of credit cards, rather than just one, with various limitations, then this would be more secure. Limitations could be simple spending limits, such as per-transaction or per-day limits, or it could be something more complicated, like only allowing purchases in the city you live in.
Pros: Completely compatible with existing systems. Easy for people to understand. Cheap to implement.
Cons: Most of the downsides of credit cards still apply; this just mitigates the risks.
Third Proposal - Authorized transactions
It should be possible (if perhaps a little annoying) for your bank to call you and check with you whenever a purchase is made on your account. Since having a person calling you all the time would be a bit much, it could also be an automated calling system, or a text message, or an email if it's not time sensitive, or something else. The main requirement is that it be something that's difficult to fake.
Pros: Compatible with existing systems. Prevents any unauthorized transactions, even those made with a stolen card and all relevant data.
Cons: Kind of annoying.
(Note that nowhere in here have I mentioned biometrics. Biometrics look cool in movies, but they're basically useless for actual security, for reasons that I'll go into more in a future post.)
Security is always a tradeoff, and when it comes to credit cards we've always gone far, far to the extreme of preferring convenience. But it doesn't have to be that way, and if we were willing to give up a small amount of convenience we could make credit card fraud a thing of the past.
Thursday, February 12, 2009
Senses
What is a sense? In the most abstract interpretation of the word, it's a stream of information about your surroundings, which arrives in your mind without conscious effort. People usually say we have five of them, but this interpretation is actually broad enough to give us a few more. For example, we have a natural empathic sense, in that we can sense the moods of people around us, based on how they act. This happens without conscious effort, and gives us useful information about our surroundings, so we may as well treat it as a sense.
However, the fact is that there is a lot of stuff that goes on in the world that we're totally incapable of directly sensing. This really should bother us more than it does, but nobody really notices, because we can't see, hear, smell, taste, or feel these things. Out of sight, out of mind, or something like that.
I'm not content with that. It bothers me that there's a whole world of stuff that I'll never see. Luckily, I don't have to just live with it; we live in an age when most things are possible, after all. If this kind of limitation is part of being human, then fine, we can become more than human. (Some people claim that this sort of thing will eventually cause us to lose our individuality and creativity, for the sake of technology. Some people watch too much Star Trek.)
The first goal is the electromagnetic (henceforth, the EM) spectrum. We can see a relatively narrow slice of it - the human eye picks up the bits that happen to have wavelengths between 380 and 760 nm. Even just extending that a little in each direction, to encompass infrared and ultraviolet, we find enormous possibilities. Imagine, for instance, a firefighter that could see in infrared, and track the spread of a fire on the other side of a wall. Going up past UV in the spectrum wouldn't be quite so useful, simply because there isn't that much high-energy radiation flying around. Nuclear explosions would probably look pretty flashy if you could see X-rays, but then, they're already kind of hard to miss.
Q: What would happen if we could suddenly see through walls?
A: We would start building better walls.
Another neat thing that came up recently - the haptic compass. The idea here is to basically give you an intuitive sense of direction. People who have tried this report a huge improvement in their ability to navigate through cities, proving that the brain is perfectly capable of picking up a new sense quickly.
But this is just the beginning of what's possible. The EM spectrum is what I'll call a "direct" sense - if you were to see it, you'd be seeing actual photons that are zooming around. It's directly tied to a physical phenomenon. It's possible to imagine "secondary" senses, like my original example of the empathic sense, which are aggregations of data from the other senses.
In a few decades, a set technologies which are only just getting accepted into the mainstream now will converge in such a way as to create something completely new. Ubiquitous Internet access, wearable computing, and some as yet hypothetical advances in computer vision will allow us to get information instantly about anything we can look at. It's not a huge step from there to imagining a system which tracks what you're looking at, and shows you extra information about it. If done seamlessly enough, I believe that this could be natural enough to call a new "sense".
(As I am writing this, a news story has just come through my RSS reader - a group at MIT has done something similar. It's obviously a pretty limited version of what I'm describing, but since it's MIT we're talking about here, they're probably thinking along the same lines as me. Awesome. :3)
But why stop there? Let's get into real sci-fi territory. In forty or fifty years, it's not inconceivable that we'll have direct interfaces into our own brains, allowing for the creation of new senses on the fly. The brain is extremely good at aggregating large amounts of raw data and making sense of it, so anything which generates a lot of hard-to-visualize data would be a good candidate for this.
I'm sick today, unfortunately, so I'm ending this post here. :(
However, the fact is that there is a lot of stuff that goes on in the world that we're totally incapable of directly sensing. This really should bother us more than it does, but nobody really notices, because we can't see, hear, smell, taste, or feel these things. Out of sight, out of mind, or something like that.
I'm not content with that. It bothers me that there's a whole world of stuff that I'll never see. Luckily, I don't have to just live with it; we live in an age when most things are possible, after all. If this kind of limitation is part of being human, then fine, we can become more than human. (Some people claim that this sort of thing will eventually cause us to lose our individuality and creativity, for the sake of technology. Some people watch too much Star Trek.)
The first goal is the electromagnetic (henceforth, the EM) spectrum. We can see a relatively narrow slice of it - the human eye picks up the bits that happen to have wavelengths between 380 and 760 nm. Even just extending that a little in each direction, to encompass infrared and ultraviolet, we find enormous possibilities. Imagine, for instance, a firefighter that could see in infrared, and track the spread of a fire on the other side of a wall. Going up past UV in the spectrum wouldn't be quite so useful, simply because there isn't that much high-energy radiation flying around. Nuclear explosions would probably look pretty flashy if you could see X-rays, but then, they're already kind of hard to miss.
Q: What would happen if we could suddenly see through walls?
A: We would start building better walls.
Another neat thing that came up recently - the haptic compass. The idea here is to basically give you an intuitive sense of direction. People who have tried this report a huge improvement in their ability to navigate through cities, proving that the brain is perfectly capable of picking up a new sense quickly.
But this is just the beginning of what's possible. The EM spectrum is what I'll call a "direct" sense - if you were to see it, you'd be seeing actual photons that are zooming around. It's directly tied to a physical phenomenon. It's possible to imagine "secondary" senses, like my original example of the empathic sense, which are aggregations of data from the other senses.
In a few decades, a set technologies which are only just getting accepted into the mainstream now will converge in such a way as to create something completely new. Ubiquitous Internet access, wearable computing, and some as yet hypothetical advances in computer vision will allow us to get information instantly about anything we can look at. It's not a huge step from there to imagining a system which tracks what you're looking at, and shows you extra information about it. If done seamlessly enough, I believe that this could be natural enough to call a new "sense".
(As I am writing this, a news story has just come through my RSS reader - a group at MIT has done something similar. It's obviously a pretty limited version of what I'm describing, but since it's MIT we're talking about here, they're probably thinking along the same lines as me. Awesome. :3)
But why stop there? Let's get into real sci-fi territory. In forty or fifty years, it's not inconceivable that we'll have direct interfaces into our own brains, allowing for the creation of new senses on the fly. The brain is extremely good at aggregating large amounts of raw data and making sense of it, so anything which generates a lot of hard-to-visualize data would be a good candidate for this.
I'm sick today, unfortunately, so I'm ending this post here. :(
Thursday, February 5, 2009
Fridays
Many of you who read my blog will refuse to believe this, but I have latent perfectionist tendencies when it comes to some things. Even when I don't pour a lot of time into it, it bothers me when I put up something for other people to see that I know I could have done better. Over the break, I had plenty of time for blogging, but lately with school going on I've had less time to put into this blog, and it shows. The fact is, I just can't keep up with a MWF update schedule.
Therefore, starting right now, I'm going to a once weekly blogging schedule. This way, I'll have enough time between posts to actually come up with stuff to say, and start writing posts that are worth reading again.
See you in a week!
Therefore, starting right now, I'm going to a once weekly blogging schedule. This way, I'll have enough time between posts to actually come up with stuff to say, and start writing posts that are worth reading again.
See you in a week!
Monday, February 2, 2009
Subscribe to:
Posts (Atom)