“At the beginning of 2023, I established Haviland Software to build software and data science solutions for mission driven organizations — especially in the climate and sustainability sector — integrating AI responsibly and effectively. Examples of projects include working on the Energy Score Cards product for Bright Power to measure utility and carbon use in multifamily residential buildings, and the Empower AI toolkit for communitybased organizations with Unified Ground and CIV:Lab. AI improves job efficiency by automating repetitive tasks, reducing human error, and providing predictive analytics to make informed decisions. This includes summarizing large documents, writing emails, generating code, and with what is known as agents, doing actual work on the Internet and making decisions in real time. However, it is very important at this point to maintain people in the loop. AI is dynamic like people and can get things wrong, and so there are many ethical and legal considerations that companies need to consider, such as use in hiring practices or insurance approvals where laws are emerging regarding audit regulations based on bias lawsuits that have been brought on. Haviland Software helps organizations both build software and workflows with responsibility practices in mind. AI is revolutionizing nearly every industry, and understanding it provides a competitive advantage. Learning how to manage tools properly that use AI prepares students to lead in a techdriven world. In fact, I would suggest to my fellow computer science majors to spend time learning what is known as prompt-driven development practices. There are emerging techniques where code is generated through conversations between the developer and AI to exponentially speed up the development process.” Karl Haviland ’06 Principal at Haviland Software, UM Data Science Computer science major “I work as a consultant across a variety of industries including aging services, healthcare, financial services, consumer packaged goods, and automotive. I utilize statistics, machine learning, and AI to extract insights from data. This includes evaluating the effectiveness of marketing plans, forecasting product demand, and developing pricing strategies for consumer goods. I also work with clients to understand their business needs and communicate complex insights into non-technical audiences. I use AI in a variety of ways, from coding and writing assistance to routine data analysis, creative brainstorming, and more. When given enough data and proper instructions, AI can be very good at augmenting decision-making at scale. When you can automate parts of the decisionmaking framework, it allows skilled employees to simply execute tasks in a more efficient manner. We’ve implemented AI solutions to help clients identify their highest quality leads, provide recommendations on timelines to follow up with customers, and prioritize customer service inquiries. In each of those examples, AI is making hundreds or even thousands of impactful decisions each month. I use an AI coding companion, and I can’t imagine not having one now. It allows me to focus more on what needs to be done rather than how to do it. I also find value in using AI as a soundboard when brainstorming ideas. It’s like having a subject matter expert on speed dial for every subject imaginable. The potential for inaccurate output is a risk, however, especially when using AI for critical business scenarios. Because of that, it’s best to have some prior knowledge of what you’re asking AI to do so you can validate the output.” Zackary Collevechio ’16 Data scientist, WildFig Actuarial math and mathematics dual major O’Donnell and Wilson collaborated with April Drumm-Hewitt, Ph.D., assistant professor of psychology, to determine whether assignments editing and critiquing AI-generated writing would make students more effective at writing and editing their own work. They surveyed the students to try to understand how those assignments might affect their thinking about AI’s capabilities and their own scientific knowledge. While the three haven’t analyzed their results yet, they are hoping it gives them at least some ideas for how AI can be used as a tool for students in positive ways. “I think it’s important that we educate our students about how to ethically use AI so that they enter the workforce with up-to-date knowledge about resources that employers may expect them to use,” added Wilson. “However, as we learn more about the environmental impacts of huge dataprocessing centers, I believe we also need to be discerning about when it is appropriate to do so.” O’Donnell agrees. “As a biology professor, I may need to reexamine whether the positive effects of AI outweigh the detrimental impact in terms of water, energy, and pollution. This impact makes me hesitate to recommend that AI be used in the classroom unless the usage is critical to the education of the students.” To get a better understanding of what is happening in industries outside of higher education, I sought input from some Lycoming alumni who are utilizing AI to enhance and streamline their jobs to those who are leading the way in AI innovation. One thing they agree on: Lycoming students will benefit from having experience with various forms of AI to position themselves for successful careers since it will touch every field in some way. No matter how advanced AI becomes, the world will always require highly educated individuals with strong critical thinking skills and a conscious environmental footprint to harness, refine, and guide these emerging technologies. Lycoming graduates, equipped with a deep understanding of their fields and the ability to adapt to an evolving digital landscape, will continue to drive innovation and make meaningful contributions in the workforce. 20 LYCOMING COLLEGE 2025 SPRING MAGAZINE
RkJQdWJsaXNoZXIy NTA3NDk=