Module 1 of 8

AI Evaluation Foundations

What AI evaluation is, how models learn from human feedback, how evaluation differs from training and annotation, and the career paths available.

4 lessons32 minFree

Lessons

1What Is AI Evaluation?7 min
2What AI Evaluation Assessments Actually Test5 min
3AI Evaluation Roles & Specialisations5 min
4Your First AI Evaluation Experience15 min

Try an evaluation task

Explore the kind of thinking used in AI evaluation. No account needed for this sample.

Response rankingSample task · 01

Prompt

“Explain what a VPN does in two sentences for a beginner.”

Which response follows the instructions better?

Response AA VPN encrypts your internet traffic and sends it through another server. This hides your activity from your network and your IP address from the websites you visit.Response BA VPN, or Virtual Private Network, is a technology that uses tunneling protocols such as IPsec, WireGuard or OpenVPN to establish encrypted connections between endpoints, which provides confidentiality and integrity for data in transit across untrusted networks.

Show the explanation

Response A. Both are accurate, but B ignores two instructions: "two sentences" and "for a beginner". Evaluators check the instructions first, then accuracy.

The key skill: check the instructions, then judge the answer.

Continue with Module 1 4 lessons · Free with an account

How this module works

4 lessons.
Each lesson is a short text lecture (5–30 minutes) followed by mini case studies.
Free with an account.

Start this module free

Create a free account to open Module 1.

Start free