AI for the planet
What AI can do for nature, and what it uses.
Good or bad for nature?
Which statement is closest to what you think about AI and the environment?
What this lesson covers
The idea
AI can help the environment, for example with forecasts, satellite pictures and saving energy, but training and running it also uses electricity and other resources.
Good or bad for nature?
Which statement is closest to what you think about AI and the environment?
- AI mostly helps nature
- AI mostly harms nature, because it uses so much electricity
- It does both, depending on how we use it
Help or cost?
Sort each example. Does it show AI helping the environment, or a cost that means we should use AI carefully? Tap a group, read the reason, then press Next card.
Both are true
Some cards showed AI helping and some showed what it costs. Both are true of AI at once.
The cost is not a reason to stop. It is a reason to ask: is AI the best tool here, and where does its power come from?
AI can help the environment, but it uses energy and resources, so use it where it really helps.
Weighing the benefit against the cost is a choice that people make, not the machine.
Notes
AI can help the environment, but it uses energy and resources, so use it where it really helps.
Ask two questions about any AI project: what good does it do, and what does it use?
Check yourself
A state wants to find places where forest has been cleared across a huge area. Which use of AI fits best?
Ishaan's class wants to guess tomorrow's canteen sales. What is the most careful choice?
An AI tool forecasts air pollution days ahead for a whole city, using electricity at a data centre. Is that worth it?
- A chatbot asked to guess where trees are missing from the forest. A chatbot has not seen the forest. With no pictures or other data, it can only guess.
- A program that compares satellite pictures taken in different months — correct. The pictures are the data, and a patch that turned from green to bare is the clue. The tool fits the problem.
- A model that forecasts next week's rainfall over the whole forest. Forecasts are useful, but this one does not look at trees. The tool must match the problem.
- Try a simple method first, and use a bigger AI only if needed — correct. A simple method may do the job with far less energy. Bigger tools are best kept for bigger problems.
- Train a huge new model from scratch, since bigger models are better. Bigger models cost more electricity, and a small job rarely needs one. Bigger is not the same as better.
- Stay away from AI, because it does more harm than good. AI can also help nature, as the examples showed. The aim is to use it where its benefit is greater than its cost.
- No, because electricity used by a computer tool is wasted. Electricity is used for many useful things. What matters is what we get in return.
- Yes, because a tool that helps people has no real cost. Every tool has some cost. Even useful ones are worth weighing.
- Probably, if many people gain and the power is clean — correct. Weigh the benefit, here early warnings for many people, against the cost. Then choose cleaner power where possible.