I think it's more of a statement on the nature of intelligence. Many of the behaviors we view to be intelligence are really just emergent properties of iterative algorithms implemented by simple machines. We shouldn't assume intelligence holds some privileged status.

There was a fun paper a while back on how this could be explained by the use of memristors in a dynamically stable circuit. >> ^Sagemind:

So, does that prove intelligence or just the will to survive?!

Tags for this video have been changed from 'memristor, integrated circuit, Stan William, nano, technology, future, incredible, tech' to 'memristor, integrated circuit, Stan Williams, nano, technology, HP, labs' - edited by MycroftHomlz

Tags for this video have been changed from 'memristor, integrated circuit, Ray Kurzweil, nano, technology, future, incredible, tech' to 'memristor, integrated circuit, Stan William, nano, technology, future, incredible, tech' - edited by MycroftHomlz

>> ^mrk871:
Err. It does help both of those things. How about a game without limitations on storage /size. Without having to wait for it to load.

I think you are massively overestimating the storage potential of this tech.

Envision a simple FPS using your system with a level which is a 100ft^2 room, with a move granularity of 1", so assuming you can't duck or jump, and your POV dose not move or pan up and down you would need 3 HD bitmaps, at 120degrees view each, to have horizontal panoramic viewing, so that's 1200^2 locations * 3 bitmaps * 2,073,600 (1080i frame) pixels * 4bytes (32 bit color) = ~32TiB data for an empty room.

To get the level of interactivity we are currently used to you would need at least 100(probably 1000) times more points making the movement smother and allowing for vertical movement and panning. So at least 3.2PiB, and you still have an empty room.

To add objects raise that to the power of the number of movable objects/characters (including projectiles) you have in the room, say 50 objects, and we are looking at ~5x10^777 bytes.

Remember the known universe has only about 10^80 atoms. If we assuming that this memristor can store a byte per atom(not possible given the description), you are going to have to deconstruct about 5x10^693 universes in order to find the material to build the memory you require.

The current bottle neck for both porn and video games is bandwidth and latency in transmission, not generation or storage. Storage getting better is good™ but it is not, as a storage/memory medium, groundbreaking.

People have said that we will all soon be using memory for everything since computers stored data on paper tapes, but disks are likely to be much cheaper than RAM until we are at least 3-4 moore cycles beyond the physical disk data density limit. Always take often repeated decades old claims with a grain of salt.

If it allows new models of computation, then I'll buy its potential, but barring such a claim, this tech changes approximately nothing.

>> ^dgandhi:
The two forces which drive computer hardware are porn and video games.
Since this does not help porn or video games, I don't really see this being important unless it enables the construction of currently unexplored logical models.
For instance if this allows us to build systems modeled after brain cells (but orders of magnitude faster), then I think we are looking at a potential AI breakthrough, otherwise it's just the next iteration of flashRAM.


Err. It does help both of those things. How about a game without limitations on storage /size. Without having to wait for it to load. You don't need to even think about creating games in the same way. Say goodbye to polygons, we don't need to replicate scenes and add textures to them you can just use full video kind of like playing a game in google street view or something.
If you remove some of these physical constraints then whole avenues of new ways to be creative can suddenly open up.
And the freed processing power? Put that towards building decent AI or anything else the games designers and developers can think of.

>> ^messenger:
Like I said, the end user will never need more than 512K of memory.


Dude, your porn must be in HORRIBLE quality.

>> ^jonny:
It was kind of funny when he mentioned combining RAM and hard drives. Isn't that called a solid state drive?


No, it's not.

jonny,

I think the part you are missing is the possibility of changes of order of magnitude. For example, what if I could just default to having a videocamera built into the edge of my eyeglasses and simply record everything I do 24/7 in beyond HD detail? Right now that would require far more memory than I can afford, but it would also eliminate much current note taking, minutes of meetings, even allow you to "go back in time" for snapshots.

What if I wanted to transmit that data over my cell phone to someone else to watch at the same time? What if 10 Billion people on Earth wanted to do the same thing?

I know squat about the memristor, but I do believe that the possibilities for computing technologies in particular are still ripe for orders of magnitude changes. It may be that vacuum tube computers were like the water wheel and transistors the steam engine. If so we still have a lot of room for improvement in efficiency and convenience.

perhaps memristors will be our final undoing...

perhaps the brain of Richard Hart will rule our world for a century, seated atop a towering, mobile colossus, powered by the mortal bodies of ignorant humans, descendant from those who did not watch his show on Discovery Channel...

Perhaps memristors will revolutionize computing, but this video completely fails to explain why. And I know a thing or two about computing.

We really won't know what effect this will have on computing until we get some tests on data rate, reliability, and cost. Until then, it's all just speculation and not really useful information.

There are a few reasons this thing will revolutionize computer memory. For one thing, thanks to the size of the "circuitry" (measured in atoms) an enormous amount of memory can be stored with very minimal size. For another thing, these things "remember" things (like hard drives) without any power being supplied to them (like RAM requires). This means all your static data (e.g., boot drive, archived data, etc.) and your dynamic data (e.g., operating system, all running programs, etc.) can all be stored in memristors. And that means your computer will boot immediately because there will be no "load time" because it'll be like your entire operating system is always loaded and ready to execute commands.

Pardon my passionate reply above, but I just read an article about how awesome this shit will be for computers, and the Segway comparison is just disrespectful to all future technology in general. As if it's impossible that there'll ever be an advancement in computing that will revolutionize it. Pshaw.

Memristors /memˈrɪstɚ/ ("memory resistors") are a class of passive two-terminal circuit elements that maintain a functional relationship between the time integrals of current and voltage. This results in resistance varying according to the device's memristance function. Specifically engineered memristors provide controllable resistance useful for switching current. The memristor is a special case in so-called "memristive systems", a class of mathematical models useful for certain empirically observed phenomena, such as the firing of neurons.[3] The definition of the memristor is based solely on fundamental circuit variables, similar to the resistor, capacitor, and inductor. Unlike those more familiar elements, the necessarily nonlinear memristors may be described by any of a variety of time-varying functions. As a result, memristors do not belong to linear time-invariant (LTI) circuit models. A linear time-invariant memristor is simply a conventional resistor.[4]

Memristor theory was formulated and named by Leon Chua in a 1971 paper. Chua strongly believed that a fourth device existed to provide conceptual symmetry with the resistor, inductor, and capacitor. This symmetry follows from the description of basic passive circuit elements as defined by a relation between two of the four fundamental circuit variables, namely voltage, current, charge and flux.[5] A device linking charge and flux (themselves defined as time integrals of current and voltage), which would be the memristor, was still hypothetical at the time. He did acknowledge that other scientists had already used fixed nonlinear flux-charge relationships.[6] However, it would not be until thirty-seven years later, on April 30, 2008, that a team at HP Labs led by the scientist R. Stanley Williams would announce the discovery of a switching memristor. Based on a thin film of titanium dioxide, it has been presented as an approximately ideal device.[7][8][9] Being much simpler than currently popular MOSFET switches and also able to implement one bit of non-volatile memory in a single device, memristors integrated with transistors may enable nanoscale computer technology. Chua also speculates that they may be useful in the construction of artificial neural networks.[10]

I dunno, lucky - not that uneducated. I think the analogy has more to do with the claims made by the YT poster in the description. And I've heard enough "change the world" claims in computer science/engineering before to take 'em all with a grain of salt now.

As for the claim that "scientists are discovering the mathematic equations used to govern memristors are similiar to those which govern synapses in the brain" - that's pretty ridiculous. On the one hand, if all he means is that there is a current discharge in the device when it's depolarized, well, there are lots of biological and synthetic devices that do that. If on the other hand, he means they have similar voltage-gated ion channels that produce a specific response to a huge variety of inputs that are not at all clearly understood by neuroscience, well, then he's just blowing smoke.

I'm not saying it's not a cool tech, but claims of "changing the world" are rather silly for a device whose mechanisms aren't even totally understood, much less how to mass produce them with consistent properties.

btw - why Kurzweil in the tags? Is he involved in developing these? Or because it's something he'd appreciate?