Concern 301 · Working draft · editorial review in progress
Training energy
Training energy is the electricity used while an AI model learns patterns from examples.
Understanding the concern
Training involves repeating calculations and adjusting the model until it performs better. Large training runs may keep many chips busy for weeks. But training is more than the final successful run: trials, failed experiments and tuning also use electricity. A report that counts only the final run can miss part of the cost.
Think of preparing a stage production. The opening night does not show all the work done during rehearsals. Similarly, a model's size alone cannot tell us its training footprint. The chips, training time, cooling and electricity supply matter. Energy use and carbon emissions are related, but they are not the same measurement.
An example
Two models can use similar amounts of electricity while producing different emissions because they train on different power grids.
Illustrative unless identified as a documented case.
Words in plain English
- AI
- Software that learns patterns from examples or uses rules to perform tasks such as recognizing images or generating text. It can make mistakes and does not understand every situation as a person would.
- Model
- The learned part of an AI system. It uses patterns found during training to produce an answer or prediction.
- Training
- The process of using examples to shape how an AI model behaves. Learning patterns from material does not establish that the material is true or fair.
Evidence and its limits
The BLOOM study provides a measured case with defined accounting boundaries, not a universal energy figure for large models.
Luccioni, Viguier and Ligozat · Research paper
Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model ↗ (opens in a new tab)Case study including training, idle computing and equipment manufacturing.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Does the published training footprint include experiments, cooling and the source of electricity?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Stewardship
Stewardship means caring for resources and systems that others depend on, including people who will use them later. It looks beyond the benefit of a quick launch.
Plan for maintenance, energy use, data protection, and eventual shutdown. Ask who will look after the tool when the original team moves on and how future users can avoid inheriting hidden costs.
Ask together: What are we responsible for caring for over time?
Justice
Justice means fair treatment, fair procedures, and attention to who receives benefits and burdens. It asks whether people with less power can be heard and seek repair.
Look at outcomes across different groups and individual cases. Provide a usable way to challenge decisions and correct harm. Equal wording in a policy does not guarantee equal treatment in practice.
Ask together: Who benefits, who bears the cost, and who can challenge the decision?
Reading reality
Reading reality means looking closely at the situation people actually face. A visible problem may be connected to habits, incentives, or rules beneath the surface.
If an AI tool gives a poor result, look at the task, data, workplace pressures, and available support. Do not assume that changing the software alone will solve the problem.
Ask together: What is happening here, and what conditions help explain it?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.