Week 6 · Lesson 3 · 50-65 minutes

Present, Review, and Explore AI Careers

Finish strong: present your project honestly — including its limitations and the human oversight it needs — give and take helpful feedback, and explore what people who build, govern, and design AI actually do.

Materials: This lesson in a web browser · Paper and pencil, or a notes app
This lesson: In progress
Auto-saves on this device

What you'll be able to do

Present a project clearly: the problem, your design, your tests, and its limitations.
Explain the oversight and responsible use your project needs to be trustworthy.
Give and receive specific, kind, useful feedback in a review.
Describe several real AI careers — people who build, govern, and design AI.

Think about it

Two teams present waste-sorting projects. One says, 'It works perfectly!' The other says, 'It's right most of the time, but it struggles with greasy paper, so a person double-checks those.' Which team do you trust more, and why?

An honest presentation that names limitations and oversight earns more trust than a big claim.

Make a prediction

Predict: will the strongest projects be the ones that claim no mistakes, or the ones that clearly explain their limitations and how a person keeps watch?

Vocabulary

See the full course vocabulary

Limitation: A situation where a system gets things wrong, can't decide, or shouldn't be trusted — something honest projects state openly.
Oversight: A person staying responsible for checking and correcting an AI's decisions, especially when a mistake could matter.
Responsible use: Using AI in a way that is fair, honest, safe, and respects people's privacy, being clear about what it can and can't do.
AI career: A job where people build, study, govern, or design AI systems and how they are used.

Present honestly: problem, design, tests, and limitations

A strong presentation tells a clear story: here's the user need and problem definition, here's my design (input, output, features, labels, rules), here's how I tested it, and here's what I found — including its limitations. Naming limitations is not a weakness; it shows you understand your own system.

Avoid overclaiming. 'It works perfectly' is almost never true and quickly loses trust when someone finds a failure. 'It gets most cases right but struggles with X' is honest, and it's exactly what a thoughtful audience wants to hear.

'Input: a photo of one item. Output: recycling, compost, or trash.'
'It passed 7 of 9 test cases; it fails on greasy paper and shiny foil.'
'Because it can be wrong, a person checks the uncertain cases.'

Oversight and responsible use make a project trustworthy

Because AI can be wrong or unfair, a responsible project plans for human oversight: a person stays responsible for checking and correcting the AI, especially where a mistake could matter. Your presentation should say who oversees the system and when.

Responsible use also means being fair, honest about what the system can do, protecting people's privacy, and not using AI where the risks are too high. Tie together everything from the course: fairness, privacy, limitations, and human decisions all show up in how responsibly a project is designed and described.

'A student reviews any item the sorter is unsure about before it's thrown away.'
'We don't collect anyone's name or face — only photos of the items.'
'We'd never use this to decide something serious about a person.'

Reviews work best when feedback is specific and kind — and AI is a career

In a review, you give and receive feedback. Useful feedback is specific ('your greasy-paper test case is a great catch; could you add a rule for it?') rather than vague ('it's good') or harsh. Receiving feedback well means listening, asking questions, and treating it as help, not attack. Iteration continues here: reviews often spark your next improvement.

Building AI is also a growing field of work. Machine-learning engineers and data scientists build and train models. Data labelers and annotators create the labeled examples systems learn from. AI ethicists, auditors, and policy makers govern AI — checking it for fairness, safety, and privacy, and writing the rules for how it may be used. UX and product designers decide how people actually interact with AI and stay in control. Teachers, doctors, artists, and many others increasingly use AI as a tool. You don't have to be a programmer to shape how AI is built and used.

Machine-learning engineer: builds and trains the model.
Data labeler / annotator: creates the labeled examples for learning.
AI ethicist or auditor: checks systems for fairness, safety, and privacy.
Product / UX designer: designs how people use AI and keep oversight.

Worked example

A trustworthy two-minute presentation

State the problem: 'New students can't tell which bin their waste goes in, so we sort a photo into recycling, compost, or trash.'
Show the design: 'Input is a photo of one item; output is one of three labels; features include material and food-grease.'
Report the tests: 'We ran nine test cases and passed seven; it fails on greasy paper and foil.'
Name limitations and oversight: 'Those are real limits, so a person checks any item it's unsure about before it's thrown away.'
Close with responsible use: 'We collect only item photos, never faces or names, and we wouldn't use it for anything high-stakes.'

Takeaway: The most trustworthy presentation states the problem, the design, honest test results, the limitations, and the human oversight that keeps use responsible.

Overclaiming vs. honest presenting

Before (overclaiming): 'Our AI sorts waste perfectly!' — no limitations, no oversight, and it loses trust the moment someone finds a failure. After (honest): 'It passes most cases but fails on greasy paper and foil, so a person checks uncertain items, and we only use item photos.' — states limitations, oversight, and responsible use, and earns more trust.

Honesty about limits and oversight makes a project more trustworthy, not less.

Who builds and governs AI

Role	What they do	A skill they use
Machine-learning engineer	Builds and trains models	Designing and testing systems
Data labeler / annotator	Creates the labeled examples	Careful, consistent labeling
AI ethicist / auditor	Governs AI for fairness and safety	Spotting harms, setting rules
Product / UX designer	Designs how people use AI	Understanding real users

A table of AI careers. Machine-learning engineer: builds and trains models; core skill is designing and testing systems. Data labeler/annotator: creates labeled examples; core skill is careful, consistent labeling. AI ethicist/auditor: governs AI for fairness, safety, privacy; core skill is spotting harms and setting rules. Product/UX designer: designs how people use AI and keep oversight; core skill is understanding users.

Activity

Present and Review Studio

Activity briefing

Present your project honestly and give and receive specific, useful feedback using a review rubric.

Prepare a two-minute presentation covering problem, design, test results, limitations, oversight, and responsible use.
Present to a partner or small group.
As a reviewer, use the review checklist to give at least two specific, kind pieces of feedback per project.
As a presenter, write down the feedback you receive and pick one improvement to make (your next iteration).
Discuss which AI careers each project connects to and which ones interest you.

You will prepare a short, honest presentation of your project — problem, design, tests, limitations, oversight, and responsible use — and then take part in a review, giving other teams specific, kind feedback and receiving theirs, using a shared checklist. Reviews here can spark one more iteration before the Final Project.

You will need

Paper and pencil, or a notes app

Success looks like

A presentation that includes limitations, oversight, and responsible use — not just what works.
At least two specific, kind pieces of feedback given to others.
Feedback received is recorded and one next improvement is chosen.
At least one AI career your project connects to is named.

Built-in dataset

Project review rubric

A built-in review checklist teams use to give feedback. Rows cover: clear problem definition; sensible input/output/labels/features; honest test results; named limitations; a human-oversight plan; and responsible use (fairness, privacy, honesty). Each row has a prompt and space for a specific comment.

For now, work through this activity with the built-in materials, your notes, and discussion. An interactive version is coming in a later update.

Check your understanding

Check that you can present responsibly and describe real AI work.

Question 1. Why should a project presentation include its limitations?
Question 2. Decide if the statement is true or false.
You must be a computer programmer to have any role in building or governing AI.
Question 3. Which response gives the most useful review feedback?
A classmate presents a plant-disease detector that passed most tests but wasn't tested on blurry photos, and there's no plan for who checks its decisions.

0 of 3 answered

Try it yourself

Meet an AI job

Research one AI career and connect it to the project you designed this week.

Choose one role: machine-learning engineer, data labeler, AI ethicist or auditor, policy maker, or AI product/UX designer.
Write three to four sentences on what they do day to day and one skill they use.
Explain which part of your own project that role would handle.
Note one thing about that job you'd want to learn more about.

Success looks like

One AI role described accurately with a skill it uses.
A clear link between the role and part of your project.
One genuine question you have about that career.

Reflection

Saved on this device as you type

Please don't include private information. Your answers are saved only in this browser and are never sent anywhere.

Across the whole course, what is the most important thing you'll remember about using AI responsibly?

Which AI career sounds most interesting to you, and what would you want to learn next to explore it?

Recap

A finished project is presented honestly — problem, design, tests, limitations, oversight, and responsible use — improved through review, and connected to the real people who build and govern AI.

Present the problem, design, and honest test results — including limitations.
Plan human oversight and responsible use so the project can be trusted.
AI is built and governed by many roles, not only programmers.

Grades 7–8 extension

Write your project's 'responsible use' label

Some AI systems now ship with a short 'model card' or use label: what the system is for, what it's not for, its known limitations, and who is responsible for overseeing it.

Write a one-paragraph responsible-use label for your own project. State what it should and should not be used for, its main limitations, who provides oversight, and how it protects people's privacy.

Finish this lesson

Attempt the knowledge check — not done
Save a reflection (recommended) — not done

Attempt the knowledge check to finish this lesson.