2021-02-20

Re: Thoughts on the Internet and Climate

ew0k writes:

Lately I've been pondering some of the ways the internet is hurting our climate.

gemini://warmedal.se/~bjorn/posts/thoughts-on-the-internet-and-climate.gmi

Dear ew0k,

you picked quite a brick wall with no simple answers, I suppose. But along the credo "I don't need to know everything in order to know /something/", I would like to share a few thoughts and references.

The internet is an absolutely amazing thing. When I was a youngster, I didn't even come close to a computer --- today they are ubiquitous at least in this part of the world. It comes with a price tag, like everything.

You mention these:

Bitcoins. Cryptocurrencies and blockchain technology in my not so humble opinion are a interesting mathematical puzzle. Their energy expenditure is insane, so I don't use it.

Data in transit. The thing that bothers me most is listening to or watching streaming music or video. The amount of infrastructure to make this happen, is unbelievable. I rarely use this. The music I have on physical media can be transferred to my private music device, if needed. I don't use google maps while mobile for the same reason. I do load an Openstreetmap chunk of map data onto my phone like once a year. That works well enough for me. And I don't need to be online. I still have a very low volume prepaid plan, I do not use data over GSM.

Cloud computing? Not me. I do not have a use case for that. I do pay for some hosting. which runs on dedicated machines. I don't think that there are so many use cases, which must be on clouds. But maybe I'm just too old and fogey.

Machine learning. I came across a talk, where people tried to estimate the energy involved in training neural networks. In short: its awful. The talk is in German but might have suitable subtitles. The thing that I do not understand at all is this: you cannot argue with a neural network. At all. It is a black box, and in my understanding noone really knows, what this network is /seeing/. You cannot double check any decision. It is beyond my imagination, why I should trust such a thing. It is beyond my imagination, too, why so many seem to think this is a good idea alltogether. My favourite article about image recognition is the one with the elephant.

https://media.ccc.de/v/36c3-11006-der_deep_learning_hype

https://www.quantamagazine.org/machine-learning-confronts-the-elephant-in-the-room-20180920/

An interesting point of view is put forward in this article on lowtechmagazine.com:

https://solar.lowtechmagazine.com/2018/12/keeping-some-of-the-lights-on-redefining-energy-security.html

We have largely lost the ability to cope with disruption of energy services. Where I live, we have like one or two outages a year, of maybe an hour. This is not difficult to cope with. And it reminds me, that we take energy services pretty much for granted.

Things become really messy, when you look at energy consumption at large, including the intrinsic energy in an item. The item was produced using raw materials plus more raw materials to come up with the energy needed to destill the raw materials and make the item. One part of this game is to use said item for a long time. This works better for well designed and repairable items. For an interesting overview of what is involved I recommend to read David J.C. MacKay -- "Sustainable Energy without the Hot Air".

http://www.withouthotair.com/

ew0k writes:

I Don't Have Solutions

Neither do I. However, I can do little things.

Part of the answer to all this is: don't use $THING_or_SERVICE, if you can. And don't work for those providing said $THING_or_SERVICE, if you can possibly avoid it. Google, Facebook, Instagram etc. are dead, if noone uses their services. Unlikely to happen maybe, but not impossible. The old slogan "Use it or loose it!" can be turned around: "Don't use it if you want to loose it".

You can do things: start monitoring your behaviour/consumption, if you can, and assess its neccessity. Start small. Start for yourself and your peace of mind. That way you will show others, what can be done.

Unfortunetely, there are a few more boundaries, to keep in mind:

Don't starve to death --- it wont help you nor anyone else.

Don't wait for the others, go!

Don't believe in the "fairy tale" of ethernal growth on a planet of finite size.

"Any attempt of a single person to solve, what effects everyone, must fail". This is my translation of a statement from Friedrich Dürrenmatt.

https://gutezitate.com/zitat/220183

Thus: nosimpleanswers.

Cheers,

~ew

Home