RAG · AI Sustainability · UW–Madison ML Marathon 2025
AI is transforming the world.
It's also quietly reshaping the planet.
The environmental cost of AI is real — but the evidence is buried across hundreds of peer-reviewed papers that most practitioners never read. WattBot is a RAG chatbot built to fix that: grounded, citation-backed answers on AI's energy and water footprint, with an honest fallback when the evidence simply isn't there.
Retrieval-Augmented Generation
50+ scholarly articles
Citation faithfulness
Numerical accuracy
Standardized fallback
AI sustainability
The problem
The data exists. Nobody can find it.
AI's environmental impact — energy consumption, water use, carbon emissions — is documented across peer-reviewed literature in machine learning, energy systems, and environmental science. But the research is fragmented, technical, and inaccessible to the practitioners and decision-makers who need it most. The gap isn't a lack of evidence. It's a retrieval problem.
How it works
Answer from evidence — or say so.
WattBot retrieves from a corpus of 50+ peer-reviewed papers and answers only from what it finds: every claim carries a document reference ID, every number keeps its units and source study, and when the evidence isn't there it says so instead of guessing. It was evaluated on four dimensions — fluency, retrieval precision, numerical accuracy, and citation faithfulness — because a fluent answer with a fabricated number is worse than no answer at all.
Behavior 1
Concise natural-language answers
Retrieved passages are synthesized into a direct, readable response — no padding.
Q
How much energy does training a large language model typically consume?
W
Training large language models is highly energy-intensive, with estimates ranging from hundreds to thousands of MWh depending on model scale, hardware, and data center efficiency. [doc_12, doc_27]
Behavior 2
Numeric estimates with units
Numbers keep their units, context, and source study — numbers without provenance are not reported.
Q
What is the water footprint of a typical data center?
W
Studies report water usage effectiveness (WUE) values of 1.0–2.0 L/kWh for modern data centers, with older facilities significantly higher. Cooling method and climate drive wide variation. [doc_08, doc_31]
Behavior 3
Document reference IDs + supporting material
Every claim links to a reference ID, with relevant quotes, tables, or figures surfaced from the source.
Q
Which AI tasks are most carbon-intensive?
W
NLP model training dominates emissions estimates in the literature, particularly transformer pretraining at scale. [doc_04] Fig. 3 in that study shows a breakdown by task type and hardware generation.
Behavior 4
Standardized fallback on missing evidence
When the corpus can't support an answer, WattBot says so instead of guessing — an explicit, consistent fallback.
Q
What is the total carbon footprint of AI inference globally in 2024?
W
The current corpus does not contain sufficient evidence to provide a reliable estimate for this question. No supporting documents were retrieved with adequate coverage of this specific claim.
Reflection
Behind the build
Fluency is easy. Grounding is hard. The standardized fallback was the design choice I'm most proud of — it would have been easy to let the model generate a plausible estimate when the corpus came up short, and instead the system emits an explicit, consistent signal that the evidence isn't there. Responsible AI isn't a constraint you add at the end; it's an architectural decision you make at the start.