Unit 5: Ethics and Technology

Question 1

What is an algorithm?

Answer 1

A procedure or formula for solving a problem, based on conducting a sequence of specified actions. It can therefore be described as a process,based on rules, that if followed solves a problem or produces a decision.

Question 2

What is a moral agent?

Answer 2

A ‘being’ who has the ability to discern right from wrong.

Question 3

Can a machine be an agent?

Answer 3

No.
An agent is being who has the ability to
discern right from wrong. Machines cannot make these decisions.

Question 4

How can moral agents make decisions?

Answer 4

They can make a considered judgment based on professional experience, seek further information or gain a second opinion.

Question 5

What moral responsibility do moral agents have?

Answer 5

To not cause deliberate and unjustified harm

Question 6

Traditionally, who is moral agency assigned to?

Answer 6

Only to those who can be held responsible for
their actions.

Question 7

Why should we expect people to act as moral agents?

Answer 7

By expecting people to act as moral agents, we hold them accountable for the harm they cause others.

Question 8

When are we morally responsible?

Answer 8

If:
* We are causally responsible for the effects (or lack of effects) of our actions or inactions.
* We intentionally brought about the effects or could have foreseen them.
* The effects are morally important.

Question 9

If someone accidentally causes harm to someone else, will they still be regarded as morally responsible?

Answer 9

They can be, but we could consider them less blameworthy because it was not intentional.

Question 10

What is a machine learning algorithm?

Answer 10

Machine learning algorithms are a specific kind of algorithm which ‘learn’ their own ways of analysing data and reaching decisions.

Question 11

How is big data used by banks?

Answer 11

Big data is a term used to describe extremely large sets of data that can be analysed by computers to reveal patterns and trends, especially in relation to human behaviours,
decisions and interactions.

Question 12

What are key ethical concepts related to data analysis?

Answer 12

Privacy and consent

Question 13

When dealing with personal data from which people can be identified, what provisions must organisations comply with?

Answer 13

Australian Privacy Principles (APPs).

Question 14

What do the Australian Privacy Principles (APPs) set out?

Answer 14

The rights and data subjects and the obligations of organisations
which collect, process, analyse, use and store personal data

Question 15

What two elements are needed to build an algorithm?

Answer 15

1. Data (what happened in the past).
2. A definition of success (what we’re looking for which becomes embedded in the code).

Question 16

How can autonomous algorithms express bias?

Answer 16

Data can represent historical bias, as can the criteria for success, which is often based on opinion.

Question 17

What are ethical problems algorithms have that human agents do not have?

Answer 17

Decisions made by algorithms are considered objective, but the reasons behind their decisions are not transparent. There is little room to appeal the decision. See below for more:

1. They make bias appear objective (can can cement social biases they reflect).
2. Decisions resulting from algorithms are not transparent - we can ask humans to defend their decisions, but when an algorithm makes a decision the exact processes and calculations that went into that decision are not available to the person at the other end of that decision.