Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 internet speed test! ×

Comment Re: Uh, why? (Score 1) 201

If you think Vista was bad you're not old enough to remember NT 4.0.

I remember the sound system crashing on my Vista laptop, sending a horrible, unstoppable screeching through the speakers. Basically it was an audio snow crash. Yet everything else worked normally; I was able to save my work and shut the system down. And I remember thinking, "that was horrible, but so much less horrible than it could have been."

Comment Re:Good grief (Score 1) 267

>The thesis of this "scientific paper" is basically like a couple of tokers sitting around in their parents' basement saying "DUUUUDE... what if the money in our savings account DOUBLED EVERY YEAR?!???

Again this is not a critique of the paper, it is a critique of tokers sitting around in their parent's basement. There is no substance in your criticism to address, it really is just an expression of your feelings toward the paper's author. Aside from the fact that you're just name-calling, the numerical basis you've used for comparison is just wrong.

Now it so happens I have you at a disadvantage: I've actually read the paper. It's closer the tokers sitting around saying, "How can we achieve a 7% annual compound interest rate sustained over ten years with our portfolio," which is roughly what doubling your money in ten years takes. The authors are talking about what it would take to half carbon emissions which would be a 6.6% reduction each year, and they discuss methods for reducing them, which they break down into near term no-brainer, near-term difficult, and long term speculative. As is usual the further out you go the less concrete and certain you can be. This is normal in economic projections that go twenty or more years out.

Now you may disagree with the specific means proposed, some of which are quite drastic (e.g. attempting to recover external costs through inheritance taxes). But there is nothing inherently irrational about starting with a goal -- zero carbon emissions by 2050 -- then asking what it would take to achieve that. Nor is there anything inherently ridiculous with coming up with the answer that it'll take a mix of things, some of which looking twenty or more years into the future we can't predict yet.

Comment Re:Percentage doesn't matter (Score 1) 154

Oh, I think the percentage bit is significant. It shouldn't be news that they've acknowledged reality; but it's remarkable that their responses is so meaningless.

It makes me wonder whether this is just marketing BS or whether they're really that incoherent about strategy.

Many proprietary software companies have prospered in an era of open source acceptance -- even when very good free software alternatives for their products exists (Microsoft, Oracle). But although we don't tend to think of them that way, they tend to be value-priced. You get a lot of (not necessarily great) software engineering for your $199 Windows license fee.

But the play this game you need scale to amortize development costs over many users. If you have more of a niche product competing against a solid open source competitor is going to be really, really hard. As in SAS charges almost $9000 for a single seat license, and that's good for only a year; thereafter you'll have to fork over thousands of dollars every year. That kind of cash pays for a lot of R training.

Comment Re:Beyond idiotic (Score 1) 267

Well, there's good reason to hope on the carbon emissions front.

The global trend toward replacing coal with natural gas will have a massive impact on human CO2 footprint. And this isn't the result of the strangling hand of regulation either: gas plants are simply more economically efficient and easy to run. It also coincidentally generates less than half the CO2 per kwH that coal does.

This trend alone makes hitting world CO2 goals a lot more feasible. A better electricity grid will allow more diverse energy sources as well. It's really quite feasible to increase electricity production while reducing CO2 emissions.

Comment Re:It Doesn't Work That Way (Score 1) 267

Well, your point is well taken: Moore's law is an empirical observation, not the result of a plan.

However it doesn't follow in the least that doubling clean energy requires a doubling of investments. That's because clean energy is actually benefits more form economies of scale than fossil fuels. To double your output of electricity from coal, you may get better at building coal power plants, and you may enjoy some economies of scale as people invest in infrastructure to transport coal, but you still have to pay for twice as much coal. Renewables use slack resources that are simply being thrown away now: sunshine, wind, water flow etc. Of course there are physical limits to renewables, but we're nowhere near them yet.

Comment Re:As usual, more detail needed (Score 1) 122

Generally speaking you should never, ever change your behavior based on the results of a single study -- even a controlled, double-blind study, much less an epidemiological survey. You should wait for a comprehensive literature review paper in a high-impact peer reviewed journal before you consider a result reliable.

That said, correlation is still quite valuable -- to researchers. Science doesn't have the resources to come up with quick, definitive answers on a question like this, involving a complex system that is expensive and ethically tricky to monkey with. So science spends a lot of time doing safer, more affordable stuff like looking for epidemiological correlations, until it can justify spending a lot of rare research dollars on something more probative. And those dollars are about to get a lot rarer too.

Comment Re:Similar (Score 1) 211

Kiribati is going underwater. Does anyone else care? *sigh*

I could rob you and beat you to pulp. Would anyone else care? The answer is that wise people would care, because they'll know if I get away with that I'll be getting away with a lot more.

Same with climate change. Yes, Kiribati may disappear. But the Kiribatians aren't the only people who will pay; in fact most people in the world will end up paying. The way this works is that we all get some up front economic benefit from unregulated carbon emissions and we all pay for the consequences later, but the trick is that the benefits and costs aren't spread uniformly. Some people make a killing on cheap fossil fuel and then can move the bulk of the resulting assets out of the way of climate change. The worst hit are those whose wealth is in land -- the Kiribatians obviously, but also farmers in places which become unsupportably arid.

Comment Re: Oh well (Score 2) 211

I don't think it's greed. I think it's wishful thinking.

And it absolutely would be great if there were no downsides to burning all the fossil fuels we can lay our hands on. Most people on this site are too young to remember the smog we had in the 1960s and 1970s; they're imprinted on a time when gas was cheap, air was clean, and anthropogenic climate change was (as far as the general public was concerned) undreamt of. Who wouldn't want that to be true?

Comment Re:In Other Words (Score 1) 414

You did not read what I said, and are inverting the logic. Yes, the Universe manifestly DOES have a few "simple" rules a.k.a. the laws of physics, and HAS produced rocks. But that is literally irrelevant to the point that there is nothing about rocks -- or, if you prefer, the laws of physics and the medium in which they operate -- that appears "designed". The laws are regular mathematical laws and we have no evidence for some sort of highly imaginative "field" of possible mathematical law sets and possible Universal media obeying them that a designer can select from to create the design, let alone evidence for the insane recursion relation in complexity and design implicit on the existence of such a designer.

Any sentient "designer" of a Universe plus their Super-Universe within which it builds the Universe has more complexity (and greater information content) than the Universe that they designed and built. If complexity implies design, then every designer and their Universe must have a still more complex designer in a still more complex Universe. If you wish to assert that this recursion terminates anywhere, so that you can call the designer at that level "God" or "The Master Simulation Programmer", then you no longer assert that complexity necessarily implies a designer, in which case there is no good reason to apply the rule at all even in the first instance without evidence!

Quite aside from this, rocks specifically do not exhibit any of the characteristics we generally associate with designed things, and we have quite detailed mathematical models for the probable history of rocks that do not require or benefit from (in the specific sense of being improved by) any assumption of active design. Neither, frankly, do the laws of physics.

As I pointed out in another thread, the following is a classroom example of incorrect logic:

All men are mortal.
My dog is mortal.
Therefore, my dog is a man.

All computational simulations are discretized.
The Universe is discretized (or not, see other replies).
Therefore, the Universe is a simulation.

You argument is even worse:

Rocks, that do not appear to be designed, can be designed anyway.
Therefore, we can never say that rocks do not appear to be designed.

Say what?

My dog, that does not appear to be immortal, might be immortal anyway.
Therefore we can never say that dogs are mortal.

Sure we can. What you might get away with is the assertion that there is a very small chance that some living dog (including my currently living dog, that isn't dead yet!) might turn out to be immortal. However, every single dog since wolves came out of the cold that was born more than thirty years ago is to the very best of our observational knowledge and theoretical knowledge of dog biology dead as a doorknob and every living dog that any of us have ever seen appears to be aging and we all understand how aging and disease and accidents all limit life. To assert immortal dogs you have to just make stuff up -- invent things like "dog heaven" where all dead dogs run free and have an unlimited supply of bones, or imagine that somewhere there might be a very lucky ex-wolf that failed to inherit an aging gene and that has never had a fatal disease or a fatal accident and that somehow has eluded our observational detection -- so far -- and (ignoring the second law of thermodynamics and the probable future evolution of the Universe based on the laws of physics) assume that that dog will somehow survive longer than the Universe itself probably will. Both of which are pretty absurd.

So I repeat, there is absolutely nothing about rocks that makes us think that they are designed. That does not imply that they might not be designed after all, it is not a logical statement that rocks could not have been designed, it is an empirical statement that, just as dogs appear to be mortal (and not humans, however easy it is for dogs to make the mistake, especially around dinner time:-), rocks appear not to be designed. When I find a rock on the ground as I walk along, I do not quickly look around trying to figure out who designed the rock because it looks so very much like a made thing. Quite the opposite. And, I can almost guarantee, so do you!

Comment Re:In Other Words (Score 1) 414

No arguments. Simulations similarly are generally not "deterministically" scripted. They are constantly rolling (metaphorically) pseudorandom numbers to generate non-repetitive game play. But rocks or gameplay that is "generated" are still generated according to an algorithm that was designed, and I was using the term in this broader sense.

Comment Re:In Other Words (Score 1) 414

I was pointing out (possibly badly) that his argument was a formal fallacy of the general sort: "All men are mortal, my dog is mortal, therefore my dog is a man". "All simulations are discretized. The world we observe is discretized. Therefore, the world we observe is a simulation." Same argument, substitute men/dog/mortal and simulations/world/discretized (or whatever). This is simply an incorrect argument in symbolic logic completely independent of the meanings of the symbols per se, unless I am misremembering my formal symbolic logic.

ELSEWHERE I pointed out that we do not, in fact, know if the world is discretized and that even if it is as far as spacetime is concerned, that doesn't mean that it is discretized in amplitude/phase space. And I am teaching quantum mechanics every Monday, Wednesday and Friday at the moment, so I'm not exactly ignorant about this.

Comment Re:In Other Words (Score 1) 414

Shall I show you my dry stack walls and my mortared fieldstone walls? Besides, this doesn't really impact the argument. The argument is: "Things exist that appear to have functions in a system of interlocked causality. If I were going to simulate this particular system of apparent interlocked causality, I would do so by using things that have these functions so that the result looks like this system of interlocked causality. Therefore, this apparent system of interlocked causality is a simulation because it works the way a simulation of it that I built would work!"

This is an utterly absurd argument. Begging the question doesn't begin to describe it. This is just the argument for God by design dressed up in computer clothing with a side order of Solipsism, and leaves all of the same questions begged and not even acknowledged as "problems". OK, so we are a simulation. Even discretized, the Universe has the information content of at least 10^256! (that's factorial, not exclamation point, all the permutations of all the ways "stuff" can be entered into the apparent cells). Or, of course, as I argued, it could have far, far less information content because all it really has to do is provide a few gigapixels of my apparent visual field, a handful of less dense informational channels for sound, tasted, smell, and touch -- certainly less than a terabyte of information -- and update it according to a set of classical physics rules plus an interactive script. It doesn't even have to do more than one, because if the Universe is a simulation, you could be and probably are a NPC being presented to just me in my VR bodyset -- assuming that in some more fundamental reality I have an actual body and am not MYSELF a self-aware NPC in a simulation being run for things that look like giant amoebic blobs swimming in liquid helium near the cores of gas giants (or in some more bizarre environment as we have no possible way of even speculating about the physics of the world in which the host computer supposedly lies).

We could come damn near building this now -- it's an easy extrapolation of our first rudimentary VR sets. We likely couldn't make it high enough resolution yet, but that's just a matter of scaling of work underway and doesn't require anything like Planck length discretization.

Then there is that computer that we are all running on. One way or another, its information content has to be at least as large as the information content of the Universe being simulated, or Shannon has lived in vain. Furthermore, it has to have an extremely high degree of organization. Indeed, the information content in the physical hardware of any computer ever built -- all the way down to your hypothetical Planck scale -- is almost infinitely larger than the content of its "computational" working memory and processors. Indeed, if one accepts the assertion that real quantum phases etc are real numbers, and meditate on the continuum hypothesis and aleph null and aleph prime, it is infinitely larger. It takes billions to trillions of atoms to represent a single switch, and many switches and other adjuncts to perform even a simple, crudely discretized computation simulating real number arithmetic.

So if you REALLY take the simulation theory seriously, you have to have a Universe somewhere -- somewhere, somewhen, somehow, there has to be a physical basis for the computation, energy and entropy with a set of rules that encodes this massive program -- that has a much, much, much, much.... larger information content than the Universe being simulated. My laptop (plus a remote supercomputer plus a network) can play World of Warcraft and provide me with a very nice simulation shared with a few hundred others (more like a few tens in any given perceptual field representation) based on coarse-grained objects and carefully builts SURFACE representations, because the giant snapping turtles are only shell thick and have no actual internal guts. Even this crude a simulation, skin deep and lacking real depth and transmitting only a shared visual space with added sound effects that don't even try to "share" a sound space, requires ever so much more physical information to represent it.

Now if I were designing a Universal simulation, I would make it self-representing. That is, I would make it its own computer. This makes it information-theoretically compact. The program being run are "the Laws of Physics", and the data being manipulated represents nothing but itself; it isn't stored on something else. Now the simulator for the Universe is the exact size and exact structure of the Universe being simulated, Shannon is very happy, and hey, even the Planck length -- if real -- is now relevant. The only real problem left is that now it isn't a simulation, it is reality itself. And a minor secondary problem -- even if this is how I would design it (if only because I can look around me and see that it works) that doesn't mean that it was designed. One cannot look at something with a given degree of complexity and say "Wow, that's complex! It must have been designed in order to be that complex" without contradicting your own argument with the implicit assumption that there is an even MORE complex layer of reality supporting the designer and the medium in which the design is realized. The only empirical conclusion that is justified and consistent is that what we see is what we get. Reasoning by analogy isn't reasoning at all, either logically or empirically.


Slashdot Top Deals

A sine curve goes off to infinity, or at least the end of the blackboard. -- Prof. Steiner