Estimating the Increase in Emissions caused by AI-augmented Search

2024-06-17

arXiv.org

AI Summary

Existing benchmarks for search engine energy consumption are outdated, and the climate impact of generative AI–enhanced search remains poorly quantified. Method: We establish an updated empirical energy consumption baseline for traditional search and directly measure inference power draw of large language models (e.g., BLOOM, GPT-3). Integrating energy modeling, parameter-count–power mapping, and cross-model efficiency comparison, we systematically assess the embodied carbon cost of AI-augmented search. Contribution/Results: A single generative AI search query consumes 60–70× more energy than a conventional search, increasing per-query energy use by nearly an order of magnitude. This work delivers the most rigorous, empirically grounded energy-efficiency benchmark for AI-powered search to date—providing critical quantitative foundations for designing green AI search algorithms, optimizing compute scheduling, and enabling carbon-aware system design.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Efficient ML / Green AI
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

AI-generated answers to conventional search queries dramatically increase the energy consumption. By our estimates, energy demand increase by 60-70 times. This is a based on an updated estimate of energy consumption for conventional search and recent work on the energy demand of queries to the BLOOM model, a 176B parameter model, and OpenAI's GPT-3, which is of similar complexity.

Problem

• Artificial Intelligence
• Energy Consumption
• Environmental Emissions

Innovation

• AI Energy Consumption
• Search Query Efficiency
• Environmental Impact