Smarter than Human Intelligence

When speaking of AI, we should do well to look at what is needed to actually be smarter than a human.

With an AI, we assume it has dedicated hardware and power. Given it can operate continuously, it may not have to be smarter than a human to be smarter than a human. That is, if I’m half as smart as you per cycle, but can operate for thrice as many cycles, can I be said to be smarter?

As smart as humans are, we have memory recall problems, we have worries and stresses (that go beyond just having to eat, sleep). We have split attentions and interests. An AI can focus and not worry or get distracted. If it can be three times as consistent in its work than a human is, how dumb can an AI be and still be smarter than one of us?

We have to assume it can be duplicated. If I am half as smart as you, but can make two more copies of myself that can cooperate with me, can I be said to be smarter?

Compounding continuous operation, focus, and duplication, how much intelligence does an AI need to be smarter than a human?

I’ve read a few books. Some people have read many more than I have. At the tip of the long tail, someone has maybe read, what, 100,000 books? And let’s say, comprehended most of them. An AI can access all of that data and more. It still has to work out contradicting information, but what hurdles does it have that we lack? If it can grab information in a few ticks, when it takes one of us at least seconds, if not minutes or hours, how smart does it have to be when it can get the answer from stored knowledge?

If you had the perfect library of knowledge, if you could spend a lifetime curating it, then be reborn, to live with that perfect library, how much more productive would you be, finding each mote of knowledge at the right place? An AI could rewrite every document it scans through in a way that makes its next pass that much faster and more useful. And it doesn’t have to be too smart to do that. Probably not even as smart as one of us.

I’m starting to think that an AI doesn’t have to be very smart in the small to be smarter than a human in the large.


Mozilla’s Advantage in Mobile

One of the major technology spaces still up for grabs is mobile. Apple led out with the i-series of mobile devices (iPhone, iPad), running iOS, while Google came back with third-party manufactured Android and their own Google-designed Nexus devices. Of course, Microsoft has their devices and their mobile operating system, but they are playing catch-up.

Mozilla has come in late with the FirefoxOS, and without plans for their own hardware. Yet they have a distinct advantage.

One of the frustrating things about new technologies from the big three (Apple, Google, and Microsoft) is lack of integration. Especially if you don’t standardize your technology choices on one of them, but even then.

For example, you can subscribe to various publications or buy certain media from these technology vendors (and others, like Amazon), but you don’t necessarily get equal access from all your platforms. Indeed, some of your platforms may be wholly excluded.

That’s the most common case for me, as a Linux user. There isn’t a native client for accessing media on Linux, and the web offering is usually inferior (example, with the streaming music services). In some cases the web offers no solution, mostly in the case of video. A few video providers utilize Adobe Flash, but these require an obsolete library, HAL, to support their copy protection schemes (“DRM”).

But that’s why Mozilla has a strong position: the native web. It lacks some features, but it can gain them. As it develops, it will provide the strongest point for integration between platforms.

Google recently announced their “Play News Stand” application for Android. It’s an application to deliver news to you, and some of the content is purchased. But there’s no web version. There is less incentive than ever for users to buy content that’s only accessible on one device.

Consumers don’t want to switch all their device profiles and operating systems to one vendor simply to gain the marginal benefit of equal access. The economics aren’t there. They don’t get cheaper access. All they get right now is access to one shop per device.

Credit card companies would not be the force they are today if their cards only worked at just one vendor, or even a handful of vendors. True market capitalism requires open markets, and that’s what the web represents, what the web (and any viable replacement for the web) must evolve into.

Mozilla’s road may be rocky in establishing FirefoxOS and its benefits. The web as a platform has much growing up to do (especially in things like having a common user interface for applications developed by different vendors), but it has every sign that it will.

Mozilla is playing the long game here.


2020: When Computers Will Look “Over There”

Wherever you are right now, go somewhere else in your head. If you’re in your home, think about a different room. If you’re not at home, think about a room in your home. You can probably look around the room a bit, remembering all the parts of it. Now you can probably go to some shelf or drawer and look around there.

You don’t have a perfect replica, but it’s good enough that you can remember, right now, where some remote object is. And if you went there right now, you would find that object where your brain said it would be.

This is a highly developed skill of the brain. So developed, in fact, that quite a few people use mnemonic device called the Memory Palace (Wikipedia: Method of loci) to allow them to memorize information rapidly and recall it with ease. But our computers currently rely upon the use of textual bits to demarcate things like files and folders.

If we’re forced to give the location for something, we tend to speak relatively: “down the hall, third door on the left, the big bookshelf in the corner with the taco bookends, second shelf from the top.” We don’t say, “the room called var, the piece of furniture called games, the shelf called board games, the object called dominoes.” (Okay, we do say roughly the last part.)

That’s all going to change. Let’s say you have a next-generation, non-invasive Brain-computer interface (Wikipedia: Brain-computer interface). Suddenly the computer can listen for you to “say” something like, “open that thing over there.” It can store a mapping for what “over there” means, and it can use your reference to it to trigger the mapping and get the data you want, without you needing to remember the location in the computer’s terms.

This will allow the computer to manage more of the problems that are currently shared between computer and user. And it will make computers easier to use.

But it will do some other things, too. It has the potential to overturn education, by having the computer help in your learning in a way that only the best teachers currently do.

Take, for example Kickstarter: Zombie-Based Learning: Geography taught in Zombie Apocalypse by David Hunter. This looks to be a great example of traditional teaching. The teacher uses their creativity to generate a compelling narrative for the material, setting the proper pacing, activities, etc. so that the kids all learn and retain the knowledge.

With computers, and a BCI (Brain-computer interface), the computer can help people to store memories in ways that will maximize their recall. This will initially happen in some rudimentary ways, like flashing pictures that are composed from a variety of images thought to be uncanny enough to help in memory. For example, a clown in a fish bowl, next to a fish in scuba gear.

But that may give way to better schemes where the computer has image representations of some of the places it knows you remember things, and it could suggest you add the memories in those places. It could also quiz you by flashing the location and asking you to show you remember what you should.

Passwords might consist of a challenge/response, where the computer flashes an image and you have to recall another image at some place in the same sequence.

Brain-computer interfaces represent a major leap forward in what computers will be able to do for humanity. They are on the horizon, and they are undoubtedly an epoch. Just as there is the world before the Internet and after, there is the world before widespread BCI and after.


Fixing Government: End Central Planning

Let’s say you have a chain of ice cream parlors, ten in all, sprinkled over a large metropolitan area.  You’re planning your next quarter, and so you send out a press release telling everyone in reading distance which parlors they should shop at, on which days, which flavors they should buy, etc.

Something tells me that is not the way to do it.  The ice cream flows according to where the customers decide to go.  While they do take past reports of ice cream flows under advisement when they decide to look for it, if they happen to end up a kilometer from the nearest ice cream, the ice cream will eventually move to them.

This is one of the basic tenets of modern economics: let the distributed information direct resources.  Barring cosmos-scale advancements in computing power and information gathering, we must accept that the more distributed and informed agents will achieve better results.

Why is it, then, that we still practice central planning?  Why is the tax rate fixed until it changes?  Why does Congress intervene to thwart their own wisdom in basing Medicare payments on the Sustainable Growth Rate method?

The answer is simply that although it’s recognized that central planning is ineffective at best and dangerous in most cases, it feels good.  You see, for all their brave rhetoric, congress feels vulnerable.  They worry ceaselessly about their ability to continue serving us in a broken capacity.  Likewise, businesses feel as though a dark cloud could pass and Washington could decide to legislate them into obsolescence.

So they create symbioses.  Now, they need each other, joined at the hip, and find themselves stuck when circumstances change.

Solving the burdens of central planning takes very little changing, but the hardest thing to change still must: the congressional affairs with businesses must end.

Instead of mandating a certain tax rate, and having to adjust it based on economic conditions, the tax rate should already be sensitive to the economic conditions.  Instead of passing the Medicare Sustainable Growth Rate only to keep balking to the point where now Medicare provider payments have nearly doubled since the early 1990s where they should be nearly flat, they would let the rates move with the rest of the economy.

An ocean of information, the economy abiding fixed pillars in its middle is an utterly laughable suggestion.  While man may one day be capable of such constructions, by that point the need for their creation seems absurd, and building them today only carries us to torment when they fail and send huge pieces of debris tumbling at us across the waves.

Let us raise our sails once more and be ready to trim them according to the ocean rather than expecting the ocean to accommodate us.


The Coming Fragmentations

As any system grows, the cost of maintaining it as a single system grows.  When a system becomes a certain size, maintenance of it is too high, and fragmentation is required.

This can be seen in pizza.  Very few people refuse to slice a large pizza and eat it as a single unit.  Even the few that do (you monsters) would cut up a pizza that was twice the size.  Another way it happens with pizza is when the parties to the pizza could not agree on a single set of toppings, so different portions of the pizza have different toppings.

But it can also be seen in a few non-pizza places, which I will examine below.


There’s a reason why a world government has been opposed by many people for a long time.  The size of such a government would be so big, with so much bureaucracy, that we would all die trying to figure out which floor we needed to visit to requisition a pencil to fill out a form to submit that would let us take our street-crossing test so that we might return home.

At some point, government at one level may grow too large, and it should fragment.  We have in the USA a national government, with state, county, and local governments.  At some point we will need to rework the layers so that we have regional governments (some of which might overlap), or specialized governments (eg, the Mississippi River Government that would manage collective interests for states and regions directly related to that river).

When we learn to manage the layers of government, we will be able to institute world governance that is not a threat to individual liberty.

This type of fragmentation might be called stratification, since it deals most particularly with layering the governments.


The transportation in America, of humans, anyway, is primarily by autonomous transport (ie, cars).  There is some use of buses, trains, planes, and boats, but these are fairly limited in scope and, therefore, use.  In certain areas, the road-based system has grown too large, and it should be fragmented to help reduce that burden.

You can only build so many roads before it does no good.  It makes far more sense to add alternative transportation to augment the system.  This means that fewer people are reliant on the original form, and more of the traffic does not overlap.  It’s equivalent to adding multiple traffic channels in other systems.  Instead of getting cross talk on a radio, you can simply move some traffic to another channel, and continue with multiple sets of conversations independently.

This type of fragmentation is also a form of stratification.


At some point it may make sense to fragment the browser.  When it happens, the OS gets new services to handle different parts of what’s currently in the browser.  That includes HTTP, bookmarks, cookies, authentication, signup, and rendering.

Some of these are already partially fragmented in the form of libraries, and some browsers like Uzbl already try to move toward a browser that is reliant on outside components.

While the functions could permanently remain in the browser, with other applications relying on the browser as a service, the benefits of moving them outside will reach a tipping point for most systems and users.

This type of fragmentation isn’t about the layers as much as about specialization, which it could be called.

Mobile Devices

One day the mobile device will likely fragment.  You will still have a dedicated component with a CCD for a camera, one with broadband wireless IO, one with a screen, but you won’t have a separate screen on your camera and phone.

In that world, you could use your computer screen as the head for your mobile device, for example, and you could use the power from the train to power your phone or mobile computer, saving your battery for later.

This is also a form of specialization, and some aspects are already there.  Many smart phones use WIFI when available instead of the wireless broadband.  There are also a few smart phones with the ability to plug in to a netbook-style dock.

Online Services

The final fragmentation for thought today is of online services like Facebook, but also things like Google and Wikipedia, and even the DNS itself may fragment.  There’s an ongoing push for someone to come up with distributed social networking.  Diaspora is the most prominent attempt, but others are working in the same direction.  This type of fragmentation might be called democratization, because its primary goal is to restore the control over the service to the users.

But it also has other benefits, including the possibility of improved utility.

Stratification, Specialization, Democratization

The three types of fragmentation today were in layering the functions, in breaking up by activity, and in distributing the control of systems.  They all have their places, and some systems will require a combination of them, or even something different entirely.

But we should be aware of the systems we interact with, and we should consider whether the problems we see are caused by other factors, or if they are due to the system outgrowing its britches.

The examples are numerous.  I could go on.  Economic systems, little league sports organizations, insect colonies, large-scale computing, military, etc.  The abstraction of fragmentation is quite useful, and even more so when intelligently put into practice.