To keep the Internet?

Submitted by naught101 on Wed, 11/28/2007 - 08:44

In wanting to create a new society, I have a few obvious "core" values (quote marks due to our ex-prime minister's bastardisation of the word in the phrase "core promises") .

These consist of:

  • Best practice environmentalism (not best as in better than everyone else, but best as in as good as possible).
  • Autonomy/self governance for groups and individuals
  • Freedom of information

In that order. These are fairly solid for me, and I won't really bother discussing why in this piece. I think that the second point is basically my ideal for best practice social organisation.

So how to go about the third? I think the internet might be the answer.

Why keep the internet?

The third of my three major values could actually be a subset of the previous two. For the best decisions, in both the social and environmental spheres, to be made by rational human beings, those human beings need access to the fullest possible information set.

There are a number of ways that information can be communicated:

  • Face-to-face: aural and visual communication, direct, immediate, and ephemeral
  • Recorded: written down, or otherwise recorded. Permanent.

The first is important at a local scale, but it is impossible to use for empirical data on what's best for what situation - at least it doesn't transport well. Stories are good stores of localised information, but like analogue data, the more times it's passed on, the more noise and impurities make it in. The Chinese whisper effect.

Recorded data, especially digital, is more permanent, and therefore more suited to storage of statistical, or empirical data - which is the best, or safest design of a wind turbine, methods for re-using, or recycling waste, and also things that need to be recorded in their original form - pure thoughts of philosophers, scientists, artists, etc. There are a number of ways this can be transmitted:

  • One to One - not often applicable to recorded information, but data.
  • One to Many - the standard form of broadcast journalism - a great bastion of authoritarianism
  • Many to One - there's not much of this in real life, although perhaps voting could be seen as one form of this.
  • Many to Many - this may never have truly happened before the advent of the internet and its predecessors.

It's this last method that is most important. It acts to destroy hierarchies, subverting the censoring effects of large, loud, famous ideas, over ideas held by the little people. In its incarnation as the internet, it allows immediate, and as-needed response, by anyone, for anyone. The perfectly unbiased critique, by way of all sides being completely biased, and equally loud.

Various methods allow different ways to reach this - the forum, or comments page, allows discussion and any level, and readers can make up their own mind as to which truth they prefer. The wiki allows the authors to fight it out, and through a process of minor censorship and additive editing, come to the basic common denominator bias of the group as a whole. The simple fact that anyone can get their own web page for free, somewhere, means that there's always room for new ideas, if someone wants to have it (arguably, this is a one-to-many method, but taken as a whole it isn't). Each method has its own merits, and each has many merits over the first three methods of recorded communication.

The only real merit that the one-to-many method has is that it generally has more reliablity, due to processes like peer review - the many-to-many approach is much more open to rumour. However, this is only a generalisation, and there are plenty of exceptions, (falsified evidence in science, bastard opinions in newspaper taken as fact, etc.). This effect has been shown VERY clearly withing the opensource software movement - Linux/*nix is far more secure than Windows. This is purely because mistakes/bugs in software can be pointed out by others before it hits the real world (the reason I pick security is simply because security is the most widely upheld value in the digital world). It's also the case that more and more frequently, people using the 'net are aware of the need check sources, something that almost never happens in print media.

Along with these arguments on communication is the other obvious benefit - communication negates the need for a large amount of travel. Video conferencing, for instance, could reduce the environmental impact of international conferences by around 99% (Lovins, et al. Factor four). Personally, I think similar things could be done for leisure travel, but that's another issue.

Power consumption of the internet

Obviously, with the first point, there's a huge problem with climate change, and energy (peak oil, and later, peak coal). This means that most energy usage drastically needs to be cut - environmentalists have been saying this for decades. This makes it difficult to argue that the internet, or computers in general, should be kept - They use a lot of power.

An average old desktop computer, with a normal sized CRT would probably use 150-300watts. A newer model, with an LCD would probably use the same - more for the faster processor, less for the LCD. A laptop usually uses around 50-100 watts, depending on what it's being used for, and screen size (source). I'd say it's likely that desktops still far out weight laptops, almost everywhere. I'm going to assume an average of 250 watts (this may seem low, but check the next section, and consider that host/server computers rarely have monitors, or other peripherals).

Estimates on the numbers of of computers on line are hard to make, and likewise, hard to source. There's a picture on the wikipedia internet history page that quotes the ISC as saying there are 300 million + hosts/server on the 'net. The same page also quotes a JupiterResearch report which states that 1.1 billion people use the net on a regular basis. I dunno how regular this is, but it seems a safe bet to say that there would be around 750 million computers connected to the 'net, maybe more.

I reckon I probably use my laptop 6-8 hours a day. That's pretty high, much higher than most people, but also much lower than a lot of more public computers, which often get left on constantly. I have absolutely nothing to base this on, but I think on average, 8 hours a day would be pretty standard for a computer (300m constant, 450m 2-3 hours/day).

This means that the internet would use roughly:

750000000*250*8*365 Wh/annum, or 547500000000000 wh/annum - 547.5 Terawatt hours/annum (or 1971 Petajoules/annum)

Bayswater Powerstation, in the hunter valley, is a large coal powerstation (source). It has a capacity of 2,640MW, or 23126400MWh/year, 23.13TWh/year

It would take, in otherwords, about 24 bayswater powerstations to power the internet - maybe 40 average sized coal powerstations.

The IEA says that world energy consumption in 2005 was 18235 TWh (IEA, p. 26), and that would mean the internet used about 3% of the world's electricity.

Interestingly, according to a wall street journal blog info technology in the US used about 3% of electricity in 2003. It'd be a bit more by now, but it would also be a lot less world wide than in the US, so may as well stick with it. 3% of 18235 is 547.05. How's that for guess work? (ahem...)

To put this into perspective, the Aviation industry world wide produces about 2% of the world's carbon emissions. In the UK, it's currently 13%. and it's one of the fastest growing industries in the world.

Cleaning the cruft

I'd like to point out that the internet is almost completely crap. I'd say that the vast majority of internet usage is porn, games, and general 'identity politics' wankery - myspace, facebook, flickr, etc. I admit that I take part in that kind of stuff, but I definitely don't need to, and probably shouldn't as much (although most part of that use is for forwarding political causes I'm involved in).

If we cut out all the content (if that's what it can be called) that doesn't really have any social, or environmental good, we'd be left with some amazing sites - Community sites with socially valuable foci, like instructables, and deviantart (which also has a lot of cruft, but I guess that's a matter of taste ;) ); all the best wiki databases - Wikipedia, anarch(o|a)pedia, appropedia, and, of course, envirowiki; and a few hundred thousand other sites with valuable information, and amazing creativity. All in all though, I think this useful content would be 10% of the entire content of the internet, if that.

That would take power consumption down to ~50TWh/annum, or about 4 medium sized coal power stations. world wide, that's not a huge amount of power, and could easily be powered purely on solar thermal, or geothermal, or tide power. It could probably even be run purely on a distributed wind grid with some kind of storage - UPSs for hosts/servers, and perhaps gravity batteries to prevent blackouts in the first place.

This is a bit authoritarian and all, but I also think it's somewhat generous. In any case, it won't be up to me to decide in the end.

A more decentralised approach?

On of the best things about the internet is its decentralised operation. However this could be drastically improved. A friend of mine, Paul Spencer, had an idea to simply use wireless access points as nodes in an almost completely decentralised network. Obviously there are a couple of problems with this approach - the software hasn't been written yet, and current hardware might not know how to handle it, and household wifi systems aren't strong enough to travel long distances (ie. across oceans). However each of these problems could be over come - the latter through community confederalism at what ever level needed (with control of sub-sea cables, or satellites).

This system could allow each household to control their own internet access, and power it (one small computer, and a wireless router/accesspoint could easily be run off one solar panel, with a battery or two). I imagine it would make more sense in most cases for communities to also have control of, and power their own wireless tower, run on solar thermal perhaps. This would even allow people in repressive communities to get around any community imposed censorship - all they have to do is find other WAPs that will let their traffic through, and possibly use some kind of anonymising service, like TOR, or Freenet.


I don't think it is possible to be conclusive as to whether the internet should be kept, post peak-oil/(hopefully)pre-climate change, there are too many value judgements involved that may not be relevant in the near future. But it seems obvious to me that it could be done, relatively easily, and currently, I would argue that it should, as the potential good far outweighs the extra work needed to ensure it keeps operating. And even this extra work could be basically nullified, if other displaced sources of environmental/social destruction are taken into account (especially travel)


The internet has been useful until now - IMHO for continuing the awakening that has been happening since the 1960s: Decline in authoritarian values, decline in participative democracy to the extent that it ever existed, connecting dissent and disaffection.

The internet has permitted anarchistic order to rapidly emerge from the bottom up. If we have another decade or two to watch this development continue then the ongoing internal collapse of our conventinal nation-state societies which has paralled the development of mass communication might well reach its conclusions. However, other events are poised to overtake this phenomenon.

As you point ou above the internet is increasingly becoming smothered by corporate driven content. When the computer finally merges with the TV (those f. things are banned from my house BTW) and brainwashed consumers sit back once again to swallow whatever propaganda it spews out then the power of intelligent text, truths popping up everywhere, and growing awareness of the mess we are in may well be extinguished among those who would otherwise begin to change themselves and their lives.

Then we'll be back to brutal authoritarianism.