AI Explorers · lesson 21 of 24 · 25 min
People argue about whether AI will take jobs or create them. Here’s a better question: who will build the AI, and will you be one of them?
AI and work
AI is changing many jobs, mostly by taking over certain tasks rather than whole jobs. A doctor might use AI to help spot patterns in X-rays, but still talks with patients and makes decisions. A small business owner might use AI to draft emails, but still knows the customers.
In general:
- Tasks that are repetitive and follow clear patterns are the easiest for AI to help with or automate.
- Tasks that need care, trust, leadership, hands-on skill in messy places, or deep knowledge of a community are much harder for AI.
- New jobs appear too: people who train, test, audit and design AI systems, and people in many fields who know how to use AI well.
Nobody knows exactly how this will play out, and anyone who says they know for sure is guessing. What helps in every case is understanding how AI works, which you are doing right now.
Who builds AI matters
Remember the bias lesson. Teams make choices about data, features, testing and goals. When those teams include people from many backgrounds, they are more likely to notice problems, test on more kinds of people, and build tools that serve more communities.
Black technologists and researchers are already shaping AI, especially work on fairness and accountability. Joy Buolamwini’s Gender Shades research and the Algorithmic Justice League are one well-known example. Timnit Gebru, who co-authored Gender Shades, also co-founded Black in AI, a community that supports Black researchers in the field. There are many more people doing this work in universities, companies, nonprofits and startups.
Many paths into AI
You don’t have to be a math genius or have a PhD to work with AI. Roles include:
| Role | What they do |
|---|---|
| Data scientist | Finds patterns in data and builds models |
| Machine learning engineer | Builds and runs AI systems inside products |
| Data labeler or annotator | Creates the labels models learn from |
| UX designer | Designs how people interact with AI tools |
| AI ethics or policy specialist | Studies fairness and safety, and helps write rules |
| Product manager | Decides what an AI product should do and for whom |
| Any field plus AI | Nurses, musicians, farmers, coaches and teachers who use AI well in their work |
Skills that help: curiosity, asking good questions, math and logic, writing clearly, working on teams, and caring about people.
Try it
- Interview a family member, neighbor or church member (in person, with their OK) about their job. Ask: what tasks do you do every day? Which ones are repetitive? Which need people skills?
- Sort their tasks into two lists: “AI could help” and “Needs a human.”
- Pick one task from “AI could help” and describe how an AI tool might help, and what could go wrong.
- Look at the roles table and choose one role that interests you. Write two sentences about why.
- Write one question you would want to ask someone who works in AI.
Check yourself
Does AI usually replace whole jobs or take over certain tasks?
It more often takes over or helps with certain tasks inside jobs, especially repetitive ones.
Why does it matter who builds AI?
Builders make choices about data, testing and goals. Diverse teams are more likely to notice problems and serve more people.
Name two jobs that involve working with AI.
Examples: data scientist, ML engineer, data labeler, UX designer, AI ethics specialist, product manager.
Key words
- Automation: using machines to do tasks people used to do.
- Task: one piece of work within a job.
- Data scientist: someone who finds patterns in data and builds models.
- AI ethics: the study of how to make AI fair, safe and responsible.
Previous lesson: Privacy, deepfakes and trust
Back to AI Explorers
Next lesson: Capstone part 1: pick a problem and plan your data