section a- term 1

Critical thinking

The systemic evaluation or formulation of beliefs, or statements, by rational standards

Logic

The study of arguments and the rules that govern good reasoning

Informal vs formal fallacy

Informal- defective argument due to the content of the premises

Formal- defective arguments to the structure of the argument

2 kinds of informal fallacies

Relevance- premises not relevant to conclusion

Evidence- premises don’t support the conclusion in the way intended but are still relevant to the conclusion

List all informal fallacies of relevance

1. Ad hominem

2. Attacking the motive

3. Appeal to hypocrisy

4. Two wrongs make a right

5. Appeal to fear

6. Appeal to pity

7. Appeal to popularity

8. Straw man

9. Red herring

Deductive argument

the conclusion is true given the premise is true

Inductive argument

the conclusion is very likely to be true given that the premises are true

rule of truth

Something ought to believe if there is sufficient evidence that it is true or very likely to be true

When is an argument valid

If the premises are true and the conclusion clearly follows and leaves no doubt

Only occasion it’s invalid

TTF

Good deductive argument?

Valid and sound

When is an argument sound

Valid, and all premises are actually true

Valid argument forms

MP, MT, ES

Consequent implied

Only if, must, required, necessary for

Antecedent implied

Sufficient for, unless

Unless?

Unless becomes if not

Eliminative syllogism

Either P or Q

Not p

So Q

Fallacies of evidence

Circular reasoning, slippery slop, false dichotomy

Circular reasoning

Uses the conclusion as a premise

False dichotomy

Choose either X or Y but Z exists

Slippery slope

Chain of events

Strength

Good reasons to accept the conclusion with certainty

Cogency

Strong and the premises are actually true

How to make an argument stronger

Increase sample size

Independent verifiers

Reliable sources

Modest conclusion

Analogy

A comparison between two different things

Analogical reasoning

Reasoning based on analogies

Strong analogical reasoning

More relevant analogies

Less counter examples

More modest conclusion

When critiquing an analogical argument

Point out disanalogy

Counter analogy unintended consequences

Prudential reasoning

Reasoning based on what is good/bad for you

Cognitive biases

A systematic error in a persons way of thinking- ingrained patterns of thought that affect the way we process information

List all cognitive biases

Confirmation bias, cognitive dissonance, optimism bias, pessimism bias, negativity bias, choice overload, sunk cost fallacy, the framing effect

Confirmation bias

Focus on evidence that fits with what we already believe.

It prevents us from objectively assessing sources or information

Cognitive dissonance

A tension between two or more beliefs that are healed simultaneously.

To resolve this tension we rationalise our behaviour even when it conflicts with evidence available to us.

It is irrational to hold conflicting beliefs. We rationalise to avoid either belief

The sunk cost fallacy

Double down on bad choices. The more you invest in something the more likely you are to keep trying it. The more you have invested the harder it is to walk away.

The framing effect

Our decision making is affected by the way information is presented. The same information can be more or less attractive depending on how it’s presented to us

Optimism bias

A tendency to overestimate the chances of things turning out well for us and underestimate the chances of things turning out badly for us

Pessimism bias

A tendency to overestimate the chances of negative events and the underestimate the chances of positive events

Negativity bias

Negative experiences stick with us much more than positive ones

Choice overload

The more options we have available, the more difficult it is for us to choose between them. The more options we have available the less confidence we feel about our eventual decision

What is generative AI

Identifies statistical patterns .

Generates output based on patterns identified

LLMs

Large language models.

Massive data sets with linguistic input and output

Dangers of AI

Malicious AI, AI misinformation, algorithmic bias, model collapse, cognitive offloading

AI and misinformation

2 keys types:

Harmful speech: racism, sexism etc, eg, Microsoft Tay trained on twitter conversations.

Misinformation and disinformation: hallucinations, intentional disinformation

Algorithmic bias

AI relies on massive datasets so uses patterns in data sets to generate output

Model collapse

AI trained on its own output will gradually become worse at its given task (intellectual inbreeding)

Minor errors taken as new training errors are reinforced and reproduced

Cognitive offloading

Delegating cognitively demanding tasks to AI tools