The themes of how people use generative artificial intelligence (AI) and the problems they aim to solve were revealed in a research report by Marc Zao-Sanders on Harvard Business research. During interviews with many owners of AI applications like ChatGPT, many complained about its lack of practical use: "When I think of ChatGPT, I can't think of any use case in my life; everyone is crazy about it." Others believed this technology might err: "It's actually wrong in many things, enough to make me doubt all its answers."
The internet is filled with superficial examples like "text summarization," "generating marketing copy," or "code reviews." However, these streamlined generic phrases read like items on a feature list and are of limited use to the unfamiliar.
Marc Zao-Sanders and his team analyzed and distilled tens of thousands of posts, identifying over 100 use cases of generative AI covering various aspects of home and work life. They categorized these use cases into 100 categories, summarizing them into 6 top-level themes that describe the applications of generative AI from the perspectives of demand and users:
Technical assistance and troubleshooting (23%):
Users may use GenAI to solve specific technical problems or find troubleshooting steps.
Content creation and editing (22%):
Users utilize GenAI to generate article drafts, marketing copy, or perform text editing.
Personal and professional support (17%):
Includes using GenAI for personal life planning, career development advice, or assisting with professional tasks.
Learning and education (15%):
GenAI is used for educational purposes, aiding learning, providing explanations, and educational content.
Creativity and entertainment (13%):
Users use GenAI to inspire creativity and engage in entertainment activities, such as writing stories or creating artworks.
Research, analysis, and decision-making (10%):
Used to support users in research work, data analysis, and aiding decision-making processes.
These themes demonstrate the extensive practicality of generative artificial intelligence, useful for both work and leisure, aiding in creativity and technical efforts. This list was compiled based on examples reported by ordinary people who have had better, faster, or more enjoyable experiences using generative AI. This also reflects the eternal pursuit of individuals: learning, communicating, and thinking.
From the perspective of specific functions and solving problems, among the top 100 use cases of large models and generative AI, examples that support users in research work, data analysis, and aiding decision-making include:
Idea generation: Used for brainstorming, helping users generate and summarize ideas.
Specific search: Assisting users in finding specific information or items.
Text editing: Helping users check and improve their writing, identifying logical errors.
Drafting emails: Assisting users in saving time drafting formal or business emails.
Simple explanations: Explaining complex concepts to non-professionals in simple language.
Excel formulas: Assisting users in writing and simplifying Excel formulas for tasks like reconciling data.
These use cases showcase how generative AI creates value for individuals and organizations in different fields and contexts. According to research from the Harvard Business Review, researchers identified over 100 specific use cases of generative artificial intelligence.
Key Point Q&A:
What were some common complaints about AI applications like ChatGPT mentioned in Marc Zao-Sanders' research report?
Many users complained that ChatGPT lacked practical use in their lives, while others expressed doubts about its accuracy.
How did Marc Zao-Sanders and his team identify the various use cases of generative AI?
They analyzed and distilled tens of thousands of posts to identify over 100 use cases covering aspects of home and work life.
What are some examples of how generative AI can assist users in research, analysis, and decision-making?Generative AI can aid in idea generation, specific searches, text editing, drafting emails, providing simple explanations, and simplifying Excel formulas for tasks like data reconciliation.