Answers

A14.8 Moral, ethical, social and cultural issues

Software development, law and ethics · A level · OCR H446 1.5.2, AQA 7517 4.8.1 · about 45 min

What this lesson is about

Automated decisions, AI, surveillance, personal data at scale, censorship, piracy, the environment and accessibility, with an autonomous robot that explains every decision.

• Why computer scientists carry responsibility
• The words in the question
• The issues the specifications name
• Case study: the autonomous delivery robot
• Structuring a discussion

Questions 5 marks in all

1. [1 mark]A CV-screening system trained on a company's past hiring decisions rejects most applicants from one group. What is the most likely cause?

   1. ABias in the training data, which reflects unfair past decisions
   2. BA syntax error in the program
   3. CThe computer deliberately chose to discriminate
   4. DToo few features in the user interface
   Answer: A. A machine learning system learns the patterns in its data, including unfair ones.

2. [1 mark]Which of these are named by OCR as moral and ethical issues of digital technology?

   Tick every answer that is true.

   1. AAutomated decision making
   2. BCensorship and the internet
   3. CLayout, colour paradigms and character sets
   4. DClock speed
   5. ENormalisation
   Answer: A, B, C. OCR also names computers in the workforce, AI, environmental effects, monitoring behaviour, analysing personal information, and piracy and offensive communications.

3. [1 mark]AQA says software and its algorithms embed moral and cultural values. What does this mean?

   1. AEach rule in a program is a decision someone made about what should happen to people
   2. BPrograms contain comments about ethics
   3. CSoftware can only be written by philosophers
   4. DComputers have their own values
   Answer: A. A robot that gives way to some people and not others has a value built into its code.

4. [1 mark]Why does the scale of software raise the stakes of a developer's decisions?

   1. AOne program can run on millions of devices, so one choice affects millions of people
   2. BLarger programs are always slower
   3. CScale only matters for hardware
   4. DBig programs cannot contain bias
   Answer: A. Scale lets an individual developer do great good, or great harm, at once.

5. [1 mark]Why does an autonomous robot that logs each decision with its reason help with accountability?

   1. AIts decisions can be checked, questioned and corrected afterwards
   2. BIt makes the robot drive faster
   3. CIt means no human is ever responsible
   4. DIt removes the need for testing
   Answer: A. An explainable decision can be examined when something goes wrong; an unexplained one cannot.

The task: explain every decision

# the two lines every program starts with: the commands, then the robot
from bugbot import *
connect()

last = ""      # the previous decision, so a line is printed only when it changes

# the two lines every program starts with: the commands, then the robot
from bugbot import *
connect()

last = ""
while True:
    gap = distance()
    if gap > 40:
        decision, colour = "drive", "green"
    elif gap > 25:
        decision, colour = "slow", "orange"
    else:
        decision, colour = "stop", "red"
    if decision != last:
        led(colour)
        print(f"{clock()} s: {decision}, because the gap is {gap} cm")
        last = decision
    if decision == "drive":
        forward(60, distance=5)
    elif decision == "slow":
        forward(30, distance=2)
    else:
        break

tone(440, 0.5)
print("handing over to a human")