Will the internet survive energy contraction?

[Roboramma]

Given some extremely well presented posts by ben m, TFian, do you accept that any future in which you have energy for refrigeration will be a future in which you also have access to the internet?

Further, do you accept, because the energy necessary to power an efficient computer can be generated easily by human muscle power alone, that the internet will survive energy contraction?

 

[TFian]

Given some extremely well presented posts by ben m, TFian, do you accept that any future in which you have energy for refrigeration will be a future in which you also have access to the internet.

In the short term maybe, but not in the long term. Primitive forms of refrigeration are a lot easier than the Internet to manage.

Further, do you accept, because the energy necessary to power an efficient computer can be generated easily by human muscle power alone, that the internet will survive energy contraction?

No. That all assumes we'll be able to produce computers during energy contraction. No one has demonstrated how you can produce computers without massive inputs of cheap easy fossil fuels.

Robo, the question I've been trying to raise -- and it fascinates me that almost everyone talks right past it -- is whether it will make any kind of economic sense to keep the internet going in a deindustrial future. The usual response is to insist that it's technically possible, claim that the internet is too useful to let go, and jump from there to the claim that of course the internet will keep going.

The problem with that logic is simply that if the necessary jobs can be done more economically by some other means in an energy-constrained future, then they will be done that other way. In a world where half the surviving population farms for a living and energy supplies are tightly constrained in concentration as well as quantity, there are many better ways to handle information flow than vast server farms that burn through as much electricity as a small city! A shortwave radio net, using packet methods as long as computers are available, and downshifting to more human-centered methods as human labor becomes cheaper than concentrated energy, is one obvious choice.

 

[Roboramma]

No. That all assumes we'll be able to produce computers during energy contraction.

Long term, it does make that assumption, but not short term. As others have pointed out there are so many chips out there that could do the job necessary (ie. not at the level that it is being done today, but still enough for the transfer of information), that we could last a very long time simply by reusing and repairing those parts.

No one has demonstrated how you can produce computers without massive inputs of cheap easy fossil fuels.

I actually think that they have: nuclear power, for instance. Hydro. Solar. But we've been through that all through this thread. It may be more expensive to do so, but not nearly enough more expensive to warrant not using them at all.

Robo, the question I've been trying to raise -- and it fascinates me that almost everyone talks right past it -- is whether it will make any kind of economic sense to keep the internet going in a deindustrial future.

One of the reasons that not many people want to engage you on that topic is that they see no reason to believe that there will be a deindustrial future.

But further to that, a future without cheap sources of energy isn't one where we throw away what useful technology we have, it's one where we simply have more limited resources. If your doomsday scenario plays out, those with access to power sources will still make use of them, and there's no reason I can see not to use them to produce computers.
So that leads us to your next point: ie. will those computers be the most efficient or valuable use of that energy?

The problem with that logic is simply that if the necessary jobs can be done more economically by some other means in an energy-constrained future, then they will be done that other way.

Sure, but you haven't show any more economical ways to do those jobs. An energy-constrained future isn't one without any energy at all, it's just one in which energy is more expensive than it is today. Given that, you need to look at the actual energy needs of a scaled down internet that nevertheless can allow information flow at high speed.

In a world where half the surviving population farms for a living and energy supplies are tightly constrained in concentration as well as quantity, there are many better ways to handle information flow than vast server farms that burn through as much electricity as a small city!

Sure, but as others pointed out, you haven't shown that current server farms use that much energy, and, more to the point, even if they do, there's no need for them to do so in order for the internet to survive.

That is, and this is a very important point which many people have brought up and you have failed to address so far: the current energy expenditure of "the internet" and the necessary energy expenditure of a post-apocalyptic internet are two very different numbers. But even with that second much smaller number, the internet will still have survived.
So, if you want to suggest that the internet won't survive, it's that second number that you need to address.

A shortwave radio net, using packet methods as long as computers are available, and downshifting to more human-centered methods as human labor becomes cheaper than concentrated energy, is one obvious choice.

Human labour will never become cheaper than concentrated energy in the way you mean, because human labour is energy.

 

[TFian]

As others have pointed out there are so many chips out there that could do the job necessary (ie. not at the level that it is being done today, but still enough for the transfer of information), that we could last a very long time simply by reusing and repairing those parts.

That method may work for a little while, but why would you bother? When you have to do back breaking work for 12 hours a day just to feed yourself, I doubt you'll have time left over to reuse and repair computer parts.

I actually think that they have: nuclear power, for instance. Hydro. Solar. But we've been through that all through this thread. It may be more expensive to do so, but not nearly enough more expensive to warrant not using them at all.

What production facilities now run on Hydro and/or Solar?

One of the reasons that not many people want to engage you on that topic is that they see no reason to believe that there will be a deindustrial future.

Industrial society is following the same well-worn path that has led other civilizations into decline, a path involving a much slower and more complex transformation than the sudden catastrophes imagined by so many social critics today. As the romans before us, we'll suffer the same catabolic collapse.

But further to that, a future without cheap sources of energy isn't one where we throw away what useful technology we have, it's one where we simply have more limited resources.

True. But those limited resources will force us to live something closer to "Medieval" times, rather than industrial civilizational times.

If your doomsday scenario plays out, those with access to power sources will still make use of them, and there's no reason I can see not to use them to produce computers.

No doomsday scenario. Me and the Archdruid believe in a slow decline, no apocalyptic scenario or die off. Just a transition back to agrarian subsistence living.

Sure, but you haven't show any more economical ways to do those jobs. An energy-constrained future isn't one without any energy at all, it's just one in which energy is more expensive than it is today. Given that, you need to look at the actual energy needs of a scaled down internet that nevertheless can allow information flow at high speed.

Like what?

Sure, but as others pointed out, you haven't shown that current server farms use that much energy, and, more to the point, even if they do, there's no need for them to do so in order for the internet to survive.

Energy usage of server farms http://colonos.wordpress.com/2008/05/15/the-energy-nightmare-of-web-server-farms/

But even with that second much smaller number, the internet will still have survived.
So, if you want to suggest that the internet won't survive, it's that second number that you need to address.

What second numbers are you referring to?

Human labour will never become cheaper than concentrated energy in the way you mean, because human labour is energy.

It's more cheaper than even the steam engine. As are beasts of burden like horses as well.

 

[Roboramma]

What second numbers are you referring to?

The one that you cut out:

That is, and this is a very important point which many people have brought up and you have failed to address so far: the current energy expenditure of "the internet" and the necessary energy expenditure of a post-apocalyptic internet are two very different numbers. But even with that second much smaller number, the internet will still have survived.

ie. the amount of power that is necessary for the internet to do the things it needs to do to allow information to be transferred around, but not, for instance, free video sharing websites.

 

[JAStewart]

delete; wrong thread

 

[TFian]

ie. the amount of power that is necessary for the internet to do the things it needs to do to allow information to be transferred around, but not, for instance, free video sharing websites.

I think the bigger question is, why bother? Why would we go through so much effort just so we keep "chatrooms" and BBS going? We did just fine with pen and paper, for a lot longer than "The Internet" A future where we're struggling just to have enough to eat, is not one where we can expedite time to chat about mindless topics with people half way around the world.

 

[Roboramma]

I think the bigger question is, why bother? Why would we go through so much effort just so we keep "chatrooms" and BBS going? We did just fine with pen and paper, for a lot longer than "The Internet"

So you do accept that the internet surviving doesn't require server farms that use the same amount of energy as a small city, right?

"Did just fine" and "is equally efficient" are two very different things. My business, for instance, relies on the internet for advertising. It's where I get most of my customers. Without it, I'd have to do more print advertising, or something else, and whatever that is also costs money, and energy.

That scaled down internet will be much more efficient, energy-wise than the Pony Express. That's reason enough to keep it. The fact that it can also do things that the Pony Express can't do at all is a second reason.

 

[TFian]

So you do accept that the internet surviving doesn't require server farms that use the same amount of energy as a small city, right?

Sure, it's technically feasible, though you'd have to use energy to power it that would need to be used to feed yourself. You can keep it going, while starving yourself to death, yes.

"Did just fine" and "is equally efficient" are two very different things.

Efficiency is a fallacy. Look up Jevons paradox.

My business, for instance, relies on the internet for advertising. It's where I get most of my customers. Without it, I'd have to do more print advertising, or something else, and whatever that is also costs money, and energy.

Your business is probably something very abstract, which will serve no purpose in a deindustrial world, as most "telecommuting" jobs won't.

That scaled down internet will be much more efficient, energy-wise than the Pony Express.

No it wouldn't. Postal service can be powered purely by muscle power, the same can't be said about email.

That's reason enough to keep it. The fact that it can also do things that the Pony Express can't do at all is a second reason.

Only with a massive surplus of cheap energy.

 

[Roboramma]

Sure, it's technically feasible, though you'd have to use energy to power it that would need to be used to feed yourself. You can keep it going, while starving yourself to death, yes.

How is that even a response to this:

So you do accept that the internet surviving doesn't require server farms that use the same amount of energy as a small city, right?

That wasn't even addressing whether it's feasible, just asking you if you agree that, whether feasible or not, the internet surviving doesn't require server farms that use the same amount of energy as a small city.

 

[Roboramma]

No it wouldn't. Postal service can be powered purely by muscle power, the same can't be said about email.

Sure, but muscle power still requires energy. Look, here are two possible ways of getting a letter from Town A to Town B: send a person with the letter or, use a computer powered by a generator hooked up to a stationary bike.

Both use muscle power. I think you can see which one will use more. Energy is energy, whatever the source.

 

[Roboramma]

That's reason enough to keep it. The fact that it can also do things that the Pony Express can't do at all is a second reason.

Only with a massive surplus of cheap energy.

Wait, you're saying that the fact that it's more efficient energy wise only makes it a better option when there's a massive surplus of energy? It seems to me that the opposite is true: when energy is scarce you take the method that uses the least energy to do the same job. And in this case that's the internet.

 

[TFian]

How is that even a response to this:

That wasn't even addressing whether it's feasible, just asking you if you agree that, whether feasible or not, the internet surviving doesn't require server farms that use the same amount of energy as a small city.

Sure, I agree in a hypothetical sense. But I never argued otherwise.

 

[TFian]

Sure, but muscle power still requires energy. Look, here are two possible ways of getting a letter from Town A to Town B: send a person with the letter or, use a computer powered by a generator hooked up to a stationary bike.

Both use muscle power. I think you can see which one will use more. Energy is energy, whatever the source.

And sending the person to deliver the letter, is far less energy consuming than riding your bike for 15 hours a day just to send an email. Right?

 

[TFian]

Wait, you're saying that the fact that it's more efficient energy wise only makes it a better option when there's a massive surplus of energy? It seems to me that the opposite is true: when energy is scarce you take the method that uses the least energy to do the same job. And in this case that's the internet.

But it isn't. When you add up all the energy it takes to produce, and run computers, pony express becomes a far more energy efficient process, in comparison.

 

[drkitten]

But it isn't. When you add up all the energy it takes to produce, and run computers, pony express becomes a far more energy efficient process, in comparison.

The numbers were shown earlier in the thread. It takes more energy to supply horseshoes for the pony express than it does to supply the entire Internet.

 

[TFian]

The numbers were shown earlier in the thread. It takes more energy to supply horseshoes for the pony express than it does to supply the entire Internet.

The construction of horseshoes way pre dates the industrial age. Because so, it takes less energy, not more.

 

[drkitten]

The construction of horseshoes way pre dates the industrial age.

... which says nothing at all about how much energy it takes to make them. One of the advantages of modern technology is that in many regards it can be more energy-efficient than earlier methods. As a simple example, a "60 watt" incandescent bulb puts out about 360 lumen, while the equivalent 360 lumen LED light takes only about 6 watts of power.

And by contrast, it would take about thirty candles to put out 360 lumen.

And if you look at how much energy it takes to make a candle,.... well, candles are mostly made out of fat; a birthday candle weighs about a gram, so it's got about 9 kCal of energy in it, which it burns through in about ten minutes (600 seconds).

So a single birthday candle burns through approximately 40000 joules of energy in 600 seconds. A single birthday candle therefore uses about 60 watts of energy.

The effect, therefore, is that modern LED bulbs are about ten times more energy efficient than old-style incandescent bulbs, which in turn are about thirty times as energy efficient as candles. Returning to old tech would make our presumptive energy shortages 300 times worse.

 

[TFian]

Something to keep in mind about "Science" as well. This small thing called "science" was mostly an amateur activity with minimal impact on the lives of most people, even in western Europe and its colonies, until this very large thing called half a billion years of fossil sunlight started to be turned into mechanical energy by steam engines. Which were not invented by scientists, by the way -- science trailed technology until well into the 19th century.

As cheaply and easily extractable supplies of fossil fuels deplete beyond the point that they're able to prop up an industrial economy, that economy is pretty much certain to sunset out. Yes, I'm familiar with the claims made for sun, wind, and the like; they're absolutely worth developing, but remove the energy subsidy they get from the fossil fuels used in their manufacture, installation, upkeep, etc. and the claim that they can power an industrial society all on their own, it seems to me, looks very thin.

Mind you, with luck and a lot of hard work it should be possible to preserve at least a few of the technological advances of recent centuries into the far future. You don't need fossil fuels to maintain a good working knowledge of sanitation, or build solar water heaters, or make and run vacuum tube-based radio gear with an intercontinental range, and these and many things like them could substantially reduce the misery that comes to mind when words like "medieval" are mentioned. Still, in terms of energy per capita, and of the percentage of the population making a living directly by agriculture, the medieval model's probably fairly close to our future a century or two out.

 

[drkitten]

You don't need fossil fuels to maintain a good working knowledge of sanitation, or build solar water heaters, or make and run vacuum tube-based radio gear with an intercontinental range,

Well, if we can maintain vacuum tube-based radio gear, why can't we maintain transistor-based radio gear, which is a lot cheaper to manufacture and to operate?

You see, there's this little thing called "science" that lets us learn things. Vacuum tubes were designed in the 19th century when we didn't know as much about physics; when we learned about quantum theory and solid-state physics, a very smart team of people noted that you could get vacuum-tube behavior out of semiconductor-based solid state circuits.

For something like 1% of the power consumption, and now a lot less.

Similarly, if we can maintain radio, we can maintain the Internet -- the Internet is a protocol that can run over any communications system, and in fact runs over radio today in some parts of the world.

You really have no idea what the actual capacities of technology are.