UC Merced Explores Topological Materials to Reduce AI Energy Consumption

It is easy to overlook the physical cost of a digital conversation. When we question a chatbot to summarize a meeting or draft an email, the response feels instantaneous and weightless. However, the reality is far more grounded in hardware and electricity. Right here in Merced, researchers at UC Merced are tackling a problem that the rest of the world is only beginning to quantify: the massive energy appetite of artificial intelligence. The sheer scale of electricity required to maintain these systems is no longer just a technical hurdle; it is an environmental and operational challenge that demands a fundamental rethink of how we build our computers.

The numbers are eye-opening. According to recent findings from the UC Merced team, a single standard ChatGPT query consumes approximately 0.34 watt-hours. Although that might sound negligible on an individual basis, the aggregate demand of millions of users performing billions of queries creates a staggering energy load. What we have is why the current push into topological materials is so critical. By exploring these specialized materials, the university aims to create a future where AI can reduce its own energy needs, effectively putting the intelligence of the machine to work on solving the very problem its existence created.

The Institutional Shift Toward Responsible AI

This energy research does not exist in a vacuum. It is part of a broader, campus-wide strategy to integrate AI in a manner that is both responsible and strategically aligned with the university’s goals of advancing knowledge and cultivating dignity. To manage this transition, the university announced the formation of the UC Merced AI Advisory Council on February 11, 2026. This council serves as the strategic guide for how the institution integrates these powerful tools without sacrificing its core values.

The effort to standardize AI access is also moving up the chain of command. The UC Office of the President has completed contract negotiations, and UC Merced is currently coordinating with the vendor to establish a streamlined process for the campus to purchase ChatGPT EDU licenses. This move suggests a shift from fragmented, individual use toward a sanctioned, institutional framework that likely includes the “appropriate safeguards” mentioned by the Office of Information Technology to protect institutional data privacy.

We are already seeing the practical applications of this integration. For instance, the university has successfully automated research data extraction using Amazon Bedrock LLMs. Led by Dr. Christian Fons-Rosen, researchers used this technology to process over 10,000 declassified ARPA documents. These documents were often lengthy and inconsistently formatted, making manual extraction a nightmare. By leveraging AI, the team was able to turn unstructured historical records into structured data, proving that AI’s value lies not just in conversation, but in the heavy lifting of academic research. You can learn more about how these local technology trends are reshaping the Central Valley’s academic landscape.

Navigating the Ethics of the AI Classroom

While the researchers focus on the hardware and the administration focuses on the licenses, the faculty and students are grappling with the ethics of the “prompt.” On April 4, 2025, UC Merced faculty held a discussion titled “Does the Future Need Us? The Ethics of AI in Education,” highlighting a tension that exists in every classroom: where does assistance end and plagiarism initiate?

View this post on Instagram

The university has established a nuanced boundary. For those navigating admissions and personal insight questions (PIQs), the policy is clear: AI is a tool for structure and readability, not a ghostwriter. The university allows limited AI use for editing, grammar, clarity, and sentence-level edits. However, the core content must be the student’s own creation. The distinction is vital—using AI to rephrase a sentence for clarity is generally safe, but allowing AI to generate entire paragraphs or ideas presented as one’s own is strictly prohibited.

To enforce this, UC Merced utilizes AI detection tools and screening software to identify AI-generated content in applications. This creates a high-stakes environment where the “human” element of writing is the primary currency. Instructors are encouraged to develop their own “AI Usage Philosophy,” explaining to students exactly why certain tools are permitted or banned in their specific courses. This approach moves away from a one-size-fits-all ban and toward a pedagogical model where AI literacy is treated as a necessary skill, provided it is used with caution. Those interested in the broader implications of these policies can explore our educational policy guides for more context.

The Balance of Innovation and Integrity

The current landscape at UC Merced reflects a duality. On one hand, there is the embrace of cutting-edge tools like GPT-4, Bard, DALL-E 2, and Midjourney to push the boundaries of content creation and data analysis. On the other, there is a rigorous commitment to academic integrity. The university’s stance is that while an AI tool can facilitate with the “readability” of a narrative, the narrative itself must reflect the individual’s own experiences in their own words.

This balance is essential because as AI becomes more efficient—potentially thanks to the topological materials research—it will develop into even more pervasive. The goal is to ensure that as the energy cost of a query drops, the intellectual cost of relying on AI does not rise.

Local Resource Guide: Navigating AI and Energy in Merced

Given my background as a geo-journalist focusing on the intersection of technology and community, the trends emerging from UC Merced will eventually ripple out into the local business and residential sectors. Whether you are a small business owner looking to integrate AI or a property manager concerned about the energy footprints of new tech installations, you need specific expertise. If these trends impact you in the Merced area, here are the three types of local professionals you should seek out:

Sustainable Energy Infrastructure Consultants: As AI and high-compute hardware enter more local businesses, energy demands will spike. Look for consultants who specialize in “green computing” and energy-efficient HVAC systems capable of handling server loads. The key criterion here is a proven track record of reducing kilowatt-hour consumption in commercial spaces without sacrificing processing power.
Academic and Instructional Design Specialists: For local educators or corporate trainers struggling to implement AI policies similar to those at UC Merced, these specialists are essential. Look for professionals who can help you draft an “AI Usage Philosophy” that balances tool adoption with integrity. Ensure they have experience with AI detection software and the ethical frameworks of generative AI.
Enterprise IT Integration Experts: If you are looking to implement tools like Amazon Bedrock for data extraction or secure EDU-style licenses for your organization, you need an integrator who understands data privacy and institutional safeguards. Prioritize experts who can demonstrate a deep understanding of NLP (Natural Language Processing) models and the security protocols required to protect sensitive institutional data.

Ready to find trusted professionals? Browse our complete directory of top-rated professional services experts in the merced area today.