New Moore's Law: Electrical efficiency of computing doubles every 1.6 years(economist.com)
economist.com
New Moore's Law: Electrical efficiency of computing doubles every 1.6 years
http://www.economist.com/blogs/dailychart/2011/10/computing-power
2 comments
In science there's a difference between "law" and "hypothesis" and "theory". A "law", in science, is typically a formula that fits the data. For example: Hubble's law states that the relative speed of some astral object is proportional to its distance. That is, objects that are far away seem to move faster than objects that are closer. Hubble's law is an observation that holds under scrutiny. It would be silly to say that Hubble's law explains why objects that are farther away seem to move faster. That's where theories come in (they go beyond describing the facts, and actually attempt to explain why the facts arise).
So, in the scientific sense, Moore's law IS a law. It doesn't explain why it holds, but it seems to hold.
So, in the scientific sense, Moore's law IS a law. It doesn't explain why it holds, but it seems to hold.
Part of this perhaps comes from the way Moore's law is presented in introductory Information Systems classes at university business schools. (I'm currently teaching one such.)
As sad as it sounds, university freshmen/sophomores aren't as true to the cause/effect logic (including correlation != causation) typical of scientists, engineers, and other logically minded folks.
We use Moore's law as an explanation of the growth of the past, which has led (according to the text) to the commoditization and near-zero cost of storage and bandwidth. We further make the simplifying assumption that Moore's Law will continue to hold, and then discuss its consequences on potential future endeavors.
In other words, I think you're seeing "because of Moore's Law the future is such-and-such" as shorthand for "using the assumption that the pace of technological advancement predicted by Moore's Law continues to hold, the future may look like such-and-such." Yes, the language isn't as tight as those who thrive on specificity (e.g., programmers) might like, but that's perhaps a consequence of sharing these principles with every person in every discipline in the business school.
As sad as it sounds, university freshmen/sophomores aren't as true to the cause/effect logic (including correlation != causation) typical of scientists, engineers, and other logically minded folks.
We use Moore's law as an explanation of the growth of the past, which has led (according to the text) to the commoditization and near-zero cost of storage and bandwidth. We further make the simplifying assumption that Moore's Law will continue to hold, and then discuss its consequences on potential future endeavors.
In other words, I think you're seeing "because of Moore's Law the future is such-and-such" as shorthand for "using the assumption that the pace of technological advancement predicted by Moore's Law continues to hold, the future may look like such-and-such." Yes, the language isn't as tight as those who thrive on specificity (e.g., programmers) might like, but that's perhaps a consequence of sharing these principles with every person in every discipline in the business school.
That's an interesting point and I'm sure that your conclusion is true in many if not most cases. Honestly I haven't thought of that.
However, I would posit that, likely because of the way things are taught and because people use it as shorthand, there is a portion of the population that doesn't understand the link from "because (as a result of) Moore's Law" -> "based on the assumptions described in Moore's Law"
I guess the issue would be how much of the population falls into that second group.
However, I would posit that, likely because of the way things are taught and because people use it as shorthand, there is a portion of the population that doesn't understand the link from "because (as a result of) Moore's Law" -> "based on the assumptions described in Moore's Law"
I guess the issue would be how much of the population falls into that second group.
>Moore's Law is only an observation of the pace at which this happens.
the gravitational "law" is just an observation that strength of the force doubles every time distance is decreased sqrt(2).
>Computers won't be twice as powerful because of Moore's Law they will become more powerful because technology marches forward and humans are a creative bunch.
the gravitational force won't be twice as powerful because of gravitational law the force will become more powerful because ... well we don't really know why, supposedly gravitons fly in a more dense bunches :)
the gravitational "law" is just an observation that strength of the force doubles every time distance is decreased sqrt(2).
>Computers won't be twice as powerful because of Moore's Law they will become more powerful because technology marches forward and humans are a creative bunch.
the gravitational force won't be twice as powerful because of gravitational law the force will become more powerful because ... well we don't really know why, supposedly gravitons fly in a more dense bunches :)
Good point, however I think that there's a distinction in that gravitational force is a Natural Law, whereas Moore's Law is based on human actions. I.e. If every person on the plant stopped going to work in the morning gravity would still happen, but the number of transistors on a chip would stop increasing very suddenly.
>gravitational force is a Natural Law, whereas Moore's Law is based on human actions.
you treat human actions as free will. I treat them as a consequence of the natural law of increasing entropy with system following the gradient of the fastest increase. [ Under the 2nd law - closed system has non-decreasing entropy - the non-increasing result can be obtained only if nothing happens, in all other cases it is always increasing. The "live/organized matter" is the fastest way of increasing of the entropy/chaos because of less than 100% efficiency of any act to increase order - ie. any act to increase order (which is always only a local effect) results in the even bigger increase of the chaos, ie. acts of matter organization result in more chaos total across the system than the baseline "nothing happens". I'm personally fascinated that the top achievement of our civilization - computer - is a straight 100% input energy to heat device, and the ever growing ratio of the energy produced by the civilization is being directed toward computers]
>I.e. If every person on the plant stopped going to work in the morning gravity would still happen, but the number of transistors on a chip would stop increasing very suddenly.
while i completely agree with this "if A then B" statement, i pretty much doubt the physical possibility of "A". The physics/science is full of well proven "if A then B" while we don't really know whether it is "A" or "!A".
you treat human actions as free will. I treat them as a consequence of the natural law of increasing entropy with system following the gradient of the fastest increase. [ Under the 2nd law - closed system has non-decreasing entropy - the non-increasing result can be obtained only if nothing happens, in all other cases it is always increasing. The "live/organized matter" is the fastest way of increasing of the entropy/chaos because of less than 100% efficiency of any act to increase order - ie. any act to increase order (which is always only a local effect) results in the even bigger increase of the chaos, ie. acts of matter organization result in more chaos total across the system than the baseline "nothing happens". I'm personally fascinated that the top achievement of our civilization - computer - is a straight 100% input energy to heat device, and the ever growing ratio of the energy produced by the civilization is being directed toward computers]
>I.e. If every person on the plant stopped going to work in the morning gravity would still happen, but the number of transistors on a chip would stop increasing very suddenly.
while i completely agree with this "if A then B" statement, i pretty much doubt the physical possibility of "A". The physics/science is full of well proven "if A then B" while we don't really know whether it is "A" or "!A".
Previous discussion: http://news.ycombinator.com/item?id=2988208
<\rant>
Tldr: people need to stop attributing Moore's Law as the cause of technological innovation.