The environmental impact of ChatGPT
Maria Mendes
Created on October 25, 2024
Over 30 million people build interactive content in Genially.
Check out what others have designed:
GRETA THUNBERG
Horizontal infographics
FIRE FIGHTER
Horizontal infographics
DEMOCRATIC CANDIDATES NOV DEBATE
Horizontal infographics
STEVE JOBS
Horizontal infographics
ONE MINUTE ON THE INTERNET
Horizontal infographics
SITTING BULL
Horizontal infographics
10 SIGNS A CHILD IS BEING BULLIED
Horizontal infographics
Transcript
The environmental impact of Chatgpt
Maria Mendes
What is ChatGPT?
- ChatGPT is an Artificial Intelligence chatbot.
- It was developed in 2022 by OpenAI
- It is infamous for its human-like conversational response.
- It got public attention due to its language, rapidness and detail on answers.
Why is ChatGPT bad for the environment?
- The rapid advancement of technology has brought numerous benefits giving us comfort and efficiency, with our day to day lifes, but what you don't know is that it also poses significant environmental challenges.
- Technology is a main problem to environment when it comes to carbon emissions, electronic waste for example.
Why ChatGPT?
- It is suggested that training ChatGPT generates fives times more carbon dioxide emissions comparable to those the average American cars over their entire lifetimes.
- The energy consumption for training ChatGPT equates to that of an American household for more than 700 years.
- ChatGPT consumes 25 times more energy than Google.
Carbon emissions
- AI technologies depend on energy sources powered by fossil fuels.
- Fossil fuels are by far the one of the largest contributors to global climate change.
- Fossil fuels gases that are released, absorb solar energy and keep heat close to Earth's surface, rather than letting it escape into space.
Energy consumption
- ChatGPT is a learning system.
- Training a single large model is equivalent to the energy consumption of several households over a year.
- Each interaction with a model involves numerous calculations that consume energy.
Eletronic waste
- The demand for advanced hardware to run AI applications leads to increased production, which in turn generates waste.
- Toxic materials found in electronics can leach into soil and water.
Water waste
- Data centers use water-based cooling systems, which can lead to significant water consumption.
- Data centers use approximately 700,000 litres of freshwater during training. It is equivalent to the amount of water need to produce 372 Teslas.
Potential solutions
- Green Data Centers
- Lifecycle Management
- Government iniciatives
- Public Awareness