The word cloud hides physical reality: every message, stream, and model runs on devices, undersea cables, and data centres built from mined materials and run on energy and water.
Seen as a system, an app is the tip of a long chain: extraction, manufacturing, operation, disposal.
That chain spans the planet. As a result, its costs (a space question) often land far from the person tapping the screen.
Keep impacts (already landed) apart from implications (risks and opportunities ahead). And always ask who gains from the technology and who ends up bearing its cost.
The Minerals Inside Every Device
A phone resolves into mined elements: cobalt and lithium in the battery, rare earth elements in magnets and screens, plus gold, copper, tin. Most are non-renewable resources, finite on human timescales.
Extraction concentrates its burden: about two-thirds of the world's cobalt comes from the Democratic Republic of the Congo, some dug by hand in artisanal mines.
A 2016 Amnesty International report and later reporting documented hazardous conditions and child labour in parts of the supply chain.
Impact: harm to miners' health and to local water and soil. Implication: rising battery demand (phones, EVs) grows pressure, but opens a chance for cleaner, traceable sourcing.
Through power, value flows to distant firms and users while risk stays local. Through values and ethics, a low price tag hides a real cost when the device rests on hazardous labour.
Powering and Cooling the Cloud
Data centres, the server warehouses behind the cloud, consume large amounts of electricity and, for cooling, water.
The energy mix decides the footprint: fossil-fuel grids burn non-renewables, while solar, wind, or hydro cost far less carbon. Evaporative cooling can use millions of litres a day.
Generative AI raises the stakes: training and running large models increases compute, energy, and water. Google and Microsoft each disclosed water withdrawals in the billions of litres for 2022, rising since.
Through systems, efficiency gains per computation can be swallowed by growth in demand: a rebound effect. Providers prioritise uptime and cost; residents in dry regions worry about shared water.
Pressure on Ecosystems and Biodiversity
Extraction, energy, and land use reach into ecosystems and biodiversity.
Lithium is pumped as brine from the lithium triangle (Chile, Argentina, Bolivia), among Earth's driest places, using enormous water and stressing fragile wetlands, species, and Indigenous communities.
Impact: measurable water drawdown now. Implication: a hard trade-off between low-carbon transport and local ecological loss.
Disposal reaches ecosystems too: at informal sites, toxins such as lead and mercury leach into soil and water and move up the food chain.
A battery built for clean transport can degrade an ecosystem upstream: an unintended consequence. Through values and ethics, who decides when a global benefit justifies a local cost?
Digital Systems That Protect Resources
Remote sensing from satellites plus a GIS maps land change across continents; Global Forest Watch (World Resources Institute) issues near-real-time deforestation alerts.
Sensor networks and acoustic monitoring flag illegal logging; machine learning sorts camera-trap images to track wildlife at scale.
Gains so far are real but partial (faster detection, some enforcement); ahead lies either scaled conservation or greenwashing and monitoring that never becomes action.
Through power, monitoring is watching, so it matters who owns the data; communities raise data sovereignty over their land. The monitoring tools have their own footprint, so judge net effect.
Weighing a Contested Balance
Contested: digital systems strain resources and ecosystems in some places and help protect them in others, and the same person rarely feels both effects.
The outcome turns on human decisions more than on the technology: the energy mix, sourcing and disposal rules, who holds monitoring data, and whether efficiency outruns demand.
Case study
System: Global Forest Watch (GFW), a free forest-monitoring platform run by the World Resources Institute, using data, networks, and AI: satellite remote sensing feeds a cloud platform whose algorithms flag likely forest loss.
Specifics: Launched 2014, it combines NASA and European imagery with machine-learning detection to publish near-real-time alerts (GLAD) for small areas; governments, journalists, and local groups have used it to support enforcement patrols in the Amazon and Congo Basin.
Impacts and implications: Impacts: faster detection of illegal clearing, evidence for enforcement, open data that shifts accountability. Implications: chance to scale globally, versus the risk that alerts without capacity or will change little, and the platform's own resource-hungry infrastructure.
Concepts: power (open monitoring shifts who can challenge deforestation, but institutions control the platform), systems (an alert matters only if it connects to action), space (distant forests become governable). Environmental context with political links.
Theory of Knowledge
Theory of knowledge
How do we know a specific harm was caused by a digital system and not something else?
Attribution relies on data collected and framed by people; when evidence of harm and the tools that measure it come from the same systems, whose account counts as knowledge?
Active recall
Self review
Distinguish an impact from an implication using one example from this sub-topic.
Explain, through systems, one unintended environmental consequence of a digital system.
Name three competing people or communities in a resource example, and state what each wants.
Give one way digital systems harm ecosystems and one way they help, and say why the balance is contested.
