AI Training Data and Bias

What is crucial for machine learning to be effective?

• High-quality data and lots of it

• Advanced algorithms

• Fast computers

• Large teams of researchers

Where does training data often come from?

• Only from scientists

• From people like you and me

• From textbooks

• From other computers

How might a video streaming service use training data?

• To improve video quality

• To recommend what you might want to watch next

• To reduce subscription costs

• To create new movies

What is an example of being directly asked to help provide training data?

• Watching a movie

• Identifying street signs in photos

• Reading a book

• Playing a video game

What can medical researchers use as training data?

• Medical images

• Patient names

• Hospital locations

• Doctor's notes

What is the purpose of using medical images as training data?

• To teach computers how to recognise and diagnose diseases

• To improve hospital infrastructure

• To train new doctors

• To create new medical equipment

Why is it important to use high-quality data in machine learning?

• To make the computer run faster

• To ensure the machine learning model is accurate

• To reduce electricity usage

• To make the data collection process easier

What might happen if low-quality data is used in machine learning?

• The machine learning model will be more efficient

• The machine learning model might give incorrect results

• The data will be easier to collect

• The computer will run slower

How do computers often collect training data from people?

• By asking them to fill out surveys

• Without any effort on their part

• By making them solve puzzles

• By tracking their social media activity