CERAWeek: Assistant Professor of Engineering Cong Chen | Realtime Pricing & AI Agents

Item: CERAWeek: Assistant Professor of Engineering Cong Chen | Realtime Pricing & AI Agents
Rating: 4
Author: PostSingularity Institute

by Cong Chen

Applied Sciences & Engineering AI — News & Applications UYComputer science UYQArtificial intelligence (AI)

🎙 Cong Chen 👥 2K 📅 April 2, 2026 ⏱ 20 min 👁 213 📄 original study 🧭 2026-08-15

Available in: English (current) Français

Keywords

AI agentselectricity pricingenergy storagedemand responsemarket design

Summary

Cong Chen, Assistant Professor of Engineering at Dartmouth, presents his research on real-time pricing and AI agents for energy systems. He begins with an overview of his work on California’s duck curve, storage operation, and distributed energy resources. The main focus is on using large language models (LLMs) to create AI agents that simulate energy customer behavior. He describes building personas (e.g., PhD, actor, grandma) and embedding them in LLMs to observe their electricity usage decisions, including responses to a simulated blackout. The agents exhibit heterogeneous preferences and provide textual explanations for their actions. He then discusses a real-time pricing framework that addresses intertemporal opportunity costs of storage, proposing two new pricing methods (MDCP and MTRMP) that eliminate out-of-market payments and incentivize truthful bidding. He concludes with a Q&A session where he addresses questions about affordability, agent control of batteries, and token costs.

143 words

Critical Evaluation

Value of the Information & Strength of the Argument

The presentation offers valuable insights into the application of AI agents for energy behavior simulation and proposes novel pricing mechanisms. The argumentation is supported by experimental results, such as the clustering of agent decisions and the reduction of out-of-market payments. However, the talk is high-level and lacks detailed evidence or comparisons with existing methods, which limits the depth of the argumentation.

Scientific Rigor, Source Quality, Title Accuracy

The speaker is a credible researcher with relevant academic background. The talk is based on his own research, but specific sources are not cited in the video. The title accurately reflects the content. The presentation is well-structured and technically sound, though the lack of explicit references reduces the scientific rigor.

126 words

Title / Content Match

The title accurately reflects the content, focusing on real-time pricing and AI agents in energy systems.

Quality & Reliability

7/10

The presentation is based on original research by the speaker, with references to specific methods and results. However, the talk is a conference presentation with limited detail, and the sources are not explicitly cited in the video.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

CERAWeek 2026 at Dartmouth — Description link providing context for the talk

Concurring Sources

CERAWeek 2026 at Dartmouth — Official page for the event, supporting the context of the talk

Contribution & Novelties

The presentation introduces a novel approach to simulating energy customer behavior using LLM-based AI agents, which can provide insights for market design and grid resilience. The proposed pricing methods (MDCP and MTRMP) offer a potential solution to eliminate out-of-market payments and incentivize truthful bidding. The work is original and has practical implications for the energy sector.

Pour aller plus loin :

Large language models as human simulators — Relevant to the AI agent methodology.
Locational marginal pricing — Background on the pricing mechanism discussed.
Distributed energy resources — Context for the DER integration.

93 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quantity and technical level, indicating a well-rounded presentation with strong technical content.

Reliability 7/10