Rotary Corporation's Paul Peterson, Senior Solutions Architect, and Tyler D'Alto, Senior Business Intelligence Analyst, talked to Glennville Rotary Club members at their February 14, 2024, meeting on "Generative AI (Artificial Intelligence) and Impacts on Modern Business."
The meeting began with an explanation of how AI works, distinguishing between Descriptive Modeling and Generative Modeling. Descriptive Modeling, as Paul explained, classifies existing data points into categories, such as images of cats and guinea pigs. This method is primarily used in supervised machine learning tasks. On the other hand, Generative Modeling seeks to understand dataset structures and generate similar examples, like creating realistic images of guinea pigs or cats, and is mostly utilized in unsupervised and semi-supervised machine learning tasks.
The discussion then turned to the practical applications of AI, which encompass a wide range of tasks, including image and text generation, text-to-image translation, text-to-speech conversion, audio and video generation, language translation, Chat Bots, pattern or anomaly detection, automating repetitive tasks, as well as manufacturing and quality control.
Paul highlighted AI's ability to recognize patterns, citing an example of AI composing a musical piece similar to Bach's works if trained on a catalog of his compositions.
Paul then explained his use case of utilizing generative AI as a junior developer, decreasing the amount of time it takes to go from an idea to a working software solution.
Tyler emphasized AI's role in quality control and its capacity to handle repetitive tasks, freeing up employees' time for more discerning and rewarding work. He also outlined the process of integrating AI into workflows, which includes researching AI, understanding its basics, identifying relevant use cases, selecting the right AI tool for the job, starting with small projects, and collaborating with experienced users.
Despite its numerous benefits, the speakers acknowledged the challenges and concerns associated with AI, such as hallucinations, gaslighting, context limitations, legal issues, perpetuation of biases, potential weaponization, and job displacement. They noted that AI requires significant amounts of data and electricity to operate, and highlighted potential risks, such as the creation of fabricated videos in legal cases and the mimicry of singers' voices in the music industry.
In addressing these concerns, Paul emphasized the importance of viewing AI as a collaborator rather than a replacement for human creativity and discernment. He noted that while AI can make jobs easier, it is crucial to consider its implications, such as its potential influence on elections through unsavory practices.
The meeting concluded with a demonstration of AI's ability to write in the style of a given author, highlighting its capabilities as a creative tool when used responsibly.
