ENGL 2105 : Workplace-Based Writing and Research

Home | AI | Calendar

Artificial Intelligence

Artificial Intelligence is only indirectly relevant to work-based writing and research. If you need to know what your target audience does, knows, and thinks, you need to be informed by them. AI's are essentially high-powered auto-complete tools that use algorithms to identify the statistical probability of what word comes next in a sequence based on what word came next in similar circumstances previously. In other words, AI has no idea what a given person might say except in so far as a given person might be more likely to say one thing rather than another, which happens more often than you might think. Consider the use of cliches and tropes. But if your goal as a researcher is to know what a specific group of people would say based on knowledge of how they think and feel, you would need to train the AI on their pool of words and their specific context. You could do that, given a large customer feedback database and the money and tools required to AI it. But absent a giant data set you need data from relevant people to inform your decisions.

So while you could successfully use AI to understand important concepts in a field new to you, develop general background information, summarize existing content, vet questions for a questionnaire, you can't get relevant current knowledge about a real audience's goals and needs from anywhere other than representatives of that audience.

You also need always to remember that whatever information you give it to work on it remembers, which is both good and potentially very bad (if you don't want that information out in the world).

One other, very important, point. Do not accept without reflection what an AI offers. Relying on AI without reflection is a bit like thinking you can become a better musician just by listening to music, a better writer by just reading, a better artist just by going to museums. You can't just enjoy, you have to analyze. Think: What's right with this? What's wrong? What assumptions has it made? What have I assumed that it seems to have neglected? Don't just accept what an AI suggests. Work with it. Revise the prompt you gave it and go again. Try a different AI.

Your first assignment is to brainstorm problems you might find solutions to. I suggested you "Think about the various kinds of people involved in each of the problems on your list -- suffers, buyers, impeders, competitors, investors -- and how each type contributes to both the problem and the solution. People who have a problem and people who sell a solution to it are not always the same people. There are also people who profit from the problem and so might get in the way of a solution."

If you brainstormed the fist assignment without an AI assist, good for you. If you came to this screen before you did that self-directed brainstorm, you should stop reading this and brainstorm on your own. Assuming you have brainstormed, what follows is what Claude offered when I prompted it. Read my prompt closely. Think about how I specified a context and how it picked up and ignored that context.

I am taking a class called "Workplace-based writing and research." It's like usability studies for English majors. We need to identify a problem that college students or perhaps just young people might experience and develop an easy to implement solution. Such a solution might be a form to fill out, a lesson plan, or even an app. Please give me 5 problems that fit that description, the audience for solution to that problem, and a suggestion about what might solve the problem.

Okay, here are 5 problems that college students or young people might experience, along with potential solutions:

1) Time Management and Organization

Problem: Difficulty managing time and staying organized with schoolwork, extracurriculars, and personal life
Audience: College students, young professionals
Possible Solution: Develop a comprehensive digital planner/organizer app that helps users schedule tasks, set reminders, track deadlines, and visualize their weekly/monthly schedules. This could incorporate features like calendar sync, to-do lists, habit tracking, and goal setting.

2) Career Exploration and Job Search

Problem: Uncertainty about career paths and difficulty navigating the job search process
Audience: College students, recent graduates
Possible Solution: Create an interactive career exploration tool that helps users discover potential career fields based on their interests, skills, and values. This could include a questionnaire, a database of career profiles, and resources for resume writing, networking, and interviewing.

3) Mental Health and Self-Care

Problem: Struggles with mental health, such as stress, anxiety, or depression, that impact academic and personal life
Audience: College students, young adults
Possible Solution: Design an online self-care toolkit that provides guided exercises, resources, and support for managing mental health. This could include modules on mindfulness, stress reduction, counseling referrals, and building healthy coping mechanisms.

4) Personal Finance Literacy

Problem: Difficulties with personal finance, such as budgeting, saving money, and understanding credit/loans
Audience: College students, young adults
Possible Solution: Develop a financial literacy curriculum or interactive workshop that teaches foundational personal finance concepts, money management strategies, and tools for tracking expenses and building savings. This could be offered as a for-credit course or a campus-wide workshop series.

5) Essential Life Skills

Problem: Lack of practical life skills, such as cooking, cleaning, doing laundry, or basic home maintenance
Audience: College students, young adults moving out on their own
Possible Solution: Create a series of "adulting 101" tutorial videos or a guidebook that covers essential life skills. This could include step-by-step instructions, checklists, and tips for developing self-sufficiency in daily tasks and household management.

Your turn. Revise my prompt or start from scratch given your own context and see what an AI can suggest. Maybe take the best suggestion and try to get an AI to improve in it by giving it more prompts. Ask it questions. Persist.

Links worth following

  • Google Learn About