A “Secret” Recipe for Responsible GenAI Use to Strengthen Trust in Public Institutions

The views expressed are those of the author and do not necessarily reflect the views of ASPA as an organization.

By Thuong (Annie) Bui
September 6, 2024

Are you ready to revolutionize public service with the power of Generative AI (GenAI)? As public leaders, we stand at the intersection of tradition and innovation, faced with a vital question: How do we use this technology responsibly to build trust, not erode it? In this article, I want to share a fresh, practical approach for leveraging large language models (LLMs) to help you work smarter and serve our communities more effectively.

Understanding how GenAI works

To utilize this tool effectively, it’s important to understand its current capabilities and limitations. As of 2024, GenAI isn’t an expert in any specific field. Instead, it generates responses by predicting the next word in a sequence, drawing from a vast amount of data it has been trained on. This enables GenAI to mimic expert-level conversations, but it doesn’t truly understand the topics it’s discussing. As a result, while GenAI can produce convincing content, it often blends accurate information with less reliable data, making its output occasionally unpredictable and inaccurate.

The “Secret” Recipe

Given these complexities, how can local governments leverage GenAI more effectively? To answer this question, let’s look at this simple recipe that consists of three main ingredients: Data quality, prompt engineering and the human touch.

• Data Quality

Think of high-quality data as the premium fuel for your GenAI engine. In public administration, this means utilizing data that is accurate, consistent, complete and up to date. Achieving this standard doesn’t have to be overwhelming: we can start by focusing on regularly updating datasets and sourcing them from reliable providers. As drivers of this process, we must also watch out for biases in the data, as these can lead to unfair or misleading results.

Pro tips: What if you’re unsure what data or information that you should “feed” your GenAI engine? Just ask it! The model can suggest the types of data it requires for addressing your specific problems. Once you have a list, it’s your responsibility to select the right data and ensure their quality.

• Prompt Engineering

This is all about having great dialog with your GenAI. This involves providing specific clear instruction and precise context, defining specific role (tone or styles) and specifying the desired format (whether it’s a bullet list, a table, a report, etc.). Thoughtfully designed prompts can push GenAI beyond its usual patterns, steering it toward solutions that are truly aligned with your unique needs.

Pro tips: Before using the tool, consider breaking down your problem into specific tasks for each prompt. Every tool has their limits on how much data/text they can handle at once. For example, ChatGPT digest only about 4,000 characters in a single interaction (as of August 2024), so structuring your inputs strategically is key.

• Human Touch

Keeping humans “in the loop” is very crucial for balancing the innovative capabilities of GenAI with the irreplaceable value of human judgment. This human oversight helps safeguard the integrity and maintains public trust.

Pro Tips: Accountability is essential. Always double-check GenAI’s outputs against your own knowledge and ethical standards to avoid risks like hallucinations, biases and privacy breaches.

Let’s Put This Recipe In Action! In a simple scenario, let’s say your city wants to improve the way it communicates with residents. Let’s implement the above “recipe” in a few practical steps:

• Step 1: As strategic leaders, start by identifying specific goals. For instance, you might focus on reducing response times for citizen inquiries or increasing public awareness of city initiatives. A well-defined problem sets the stage for effective use of GenAI.
• Step 2: Gathering high-quality data from trusted municipal sources. For example, data from updated service logs, recent community surveys or demographic data from the latest Census Bureau reports.
• Step 3: Design specific instructions/ prompts according to your identified goal. Ensure that your instructions including the high-quality data that you gathered from step 2, explain clearly your context, desired tone and format in the prompts.
• Step 4: Carefully review the outputs generated by GenAI. Proofread and refine them to ensure they align with your city’s values and ethically meet the needs of your local community. You can also consider asking GenAI to critique its own outputs, offering a chance to further refine the results and make them even more relevant and effective.

Ongoing Evaluation: Ensuring Continuous Improvement

Since AI technology continues to evolve, strategies for its use in public administration must also adapt accordingly. To truly utilize GenAI’s potential, public administrators and leaders must commit to ongoing evaluation and refinement of its applications. This regular reassessment is key to mitigating risks, ensuring ethical considerations are upheld and adapting to new developments in the field. Nonetheless, human intelligence and oversight remain essential elements of this process, roles that AI cannot replace in the foreseeable future.

---

Author: Annie Bui is a Doctoral Candidate in Public Administration major, where her research focus on the integration of GenAI in Local Governments. She is the current President of Student Public Administration Association (SPAA) at University of La Verne. At the same time, she works for Small Business Development Center (SBDC), a non-profit organization funded by Small Business Administration, provides zero cost one-on-one consulting services to entrepreneurs for their new and existing businesses. As both a researcher and a practitioner, she committed not just to discussing theories but to actively seeking comprehensive resources that enhance our understanding of how to leverage this technology effectively in the public sector – a sector traditionally slower to adopt technology than the private sector. Her aim is to explore practical solutions that help public leaders to catch up with this fast-changing technology by being ready and proactive in adopting it. Her monthly article series will cover various topics related to GenAI in the current local government settings. Each article is designed to give government agencies the essential knowledge and tools to prepare for an effective and responsible GenAI adoption.

(1 votes, average: 5.00 out of 5)
Loading...