The Contingencies of Computational Environmentalism
What to remember when engaging with the environment as one building or using computational technologies
This article is a short reflection on concern about and action aimed at protecting the environment (environmentalism) in relation to contingencies within the technology industry.
First, what does that mean?
What is contingency?
The definition can be manifold:
- Contingency (/kənˈtɪndʒ(ə)nsi/ noun): a future event or circumstance which is possible but cannot be predicted with certainty. A provision for a possible event or circumstance, as an example: “Stores were kept as a contingency against a blockade.” An incidental expense. “allow an extra fifteen per cent on the budget for contingencies”
- Contingency: in philosophy and logic, contingency is the status of propositions that are neither true under every possible valuation nor false under every possible valuation. A contingent proposition is neither necessarily true nor necessarily false.
- Contingency plan: A plan involving suitable backups, immediate actions and longer term measures for responding to computer emergencies such as attacks or accidental disasters.
- Contingency: in math a sentence is called a contingency if its truth table contains at least one ‘T’ and at least one ‘F. ‘ SEE ALSO: Contradiction, Tautology, Truth Table.
- Contingency: a contingent event something liable to happen as an adjunct to or result of something else.
Summed up I could perhaps say: ‘I guess it depends’.
What does it depend on?
What does using advanced computing capabilities to understand and solve complex problems often depend on?
- You can afford a computational tool.
- The network exists to run code on.
- You are connected to that network.
These seem quite clear – I need something to compute and connect.
Consequences are not immediate in this context.
Contingencies are about possible future consequences, we consider that different series of events is liable to happen and we plan for this.
Still, while planning contingencies we are not necessarily aware of the consequences beyond the immediate application.
We can consider a wide range of contingencies, but one considered to a greater. extent now than before is the effect human activity or consumption has on the environment.
One interesting example is plastic pollution.
You may think this is a jump, but consider this video for a moment:
We do things that immediately seem great to us.
On the surface it was what we wanted.
Conserve food for longer, great!
Cheaper chairs, great!
Made out of plastic — when thrown away– these plastic particles go into oceans, animals, and us.
Kurzgesagt is the YouTube channel that made this video on plastic, and what they do wonderfully is present compelling stories to accentuate the narrative.
One such story is that about Midas.
What if everything you touched turned into gold?
Digitalisation seems shining with similarity.
This process of becoming digital either as a company or human.
I say this of course as I type along.
I use a variety of technologies myself.
I have written about this for some time.
Material-computing relationships
I held a talk about this earlier this year called AI and the Climate Crisis, however I realise it might have been quite negative.
My argument in short: is that all digital is physical and we need to realise this to take responsibility of the technology supply chains in a similar way that we are becoming more conscious about this in other industries.
There is often talk of human-computer interaction or human-centred artificial intelligence, yet material-computing relationships in this context becomes less important. Behind the screen and behind the digital curtain we find amongst others:
- Mining.
- Logistics.
- Factories.
Tech ethics seems often to be about who the self-driving car is supposed to hit.
Ethical concerns have to go beyond the moment of impact, rather there are a series of contingent actions that lead to technology creation in the first place.
I received some comments on my TED talk that I had not seen. So, I thought I would answer both on the video and share them here in this article.
Comment
“For a moment, you had me, but then I realized that you’re mostly casting blame, and only reporting on solutions others are carrying out. What are you doing to address the conflict between AI/MI and Climate Change?”
I thought it was a fair point, I may not have spent enough time communicating the efforts of the community or said specifically what I was going to do. Action speaks louder than words.
Answer:
Hey Collin, much like with other issues awareness is important alongside action. I want to understand these issues better, so I am doing an MSc in Social Data Science. Still, I am currently in the process of building a dataset on datacentres with the intention of facilitating collaboration between different actors in digital infrastructure in Norway. However technology infrastructure is a lot about international collaboration. At work I am building a dataset related to the United Nations Development Programme with Country Programme Documents attempting to understand the overall delivery of help to different countries with strategic language. Prior to this I spent five years building an international network for youth within sustainable business called Young Sustainable Impact. This was largely based on early-stage business ideas and is still running in four countries: Norway, China, Singapore and Bangladesh. Still engaged and want to do more of course. Practically I would love to be part of solutions, and I realise that my talks can be a bit harsh and maybe could be more constructive, I will try to improve!
Another comment I received was alluding to me as someone who was making an argument of not using technology.
Comment:
“this guy just wants us to throw all technology out the window, not use it to its benefit.”
This was interesting, yet understandable. Being too pessimistic or using a negative or critical tonality may lead to the impression that a speaker is
Answer:
“As you might notice towards the end of the talk I argue that we have to be making artificial intelligence work for our shared ecology — that indicates, I hope, that we should use technology. I am not a luddite, we will of course use technology, simply saying we can be more responsible in a variety of applications.”
Two responses to questioning material relationships in technology:
- What are you doing about it?
- Are you saying we should not use technology at all?
Summing them up may be an unfair representation of their questions or arguments, which is why I posted them in full above.
I think these questions are useful.
Firstly I am working actively to address the issue, and I am not at all saying that we should stop using technology.
The question is how can I work even better in doing so?
It is not like large technology companies are inactive. One example is improvements in data centre technology and new experiments being undertaken.
Although technology companies take responsibility to a varying degree:
With growing awareness of climate change I think it is important to consider the contingencies of computational environmentalism.
One interesting forum I have found for this is climatechange.ai:
Regardless, we know we need to take action now to ensure we mitigate drastic adverse climate change that will occur if we do not change.
Additionally we have to see the labour that goes into creating technology.
We have to better understand the materials we are holding in our hands, typing on or looking at.
Material-computing relationships and contingencies therein can affect the human-computer interaction.
Ignoring consumption and resources while taking no responsibility of the impact technologies have is no longer an acceptable practice.
Yes, we need computational environmentalism — actions by people with these skills. Still, we need the awareness simultaneously of the consequences that systems of computation has on the planet and people.
This is #500daysofAI and you are reading article 495. I am writing one new article about or related to artificial intelligence every day for 500 days.
