Cognitive Science - Second Mid-Term

Front 1

myth: we only use 10% of our brain

Back 1

we use all of our brains known because: people suffer deficits when damaging parts of the brain; evolution would not waste energy building parts of your brain not utilized

it might be possible to be more productive if we used all our brain at once, but quantities are not known

believed because: availability cascade (widely accepted knowledge, believe what others around us believe); wishful thinking

* people CAN lose whole hemispheres and still function relatively normally

Front 2

myth: ESP (extra sensory perception) is real:

Back 2

* believed because: confirmation bias (coincidences happen, things happen by chance and we see it as something that happens regularly); neglect of negative results (only positive results published); wishful thinking

Front 3

myth: listening to Mozart makes babies smarter:

Back 3

* believed because: wishful thinking (we hope its true)

Front 4

myth: IQ tests are biased:

Back 4

* panels of scientists conclude they are not biased, they may be a bad test of future success but it;s the best we’ve got, item analysis is used to identify bad biased questions

Front 5

myth: money makes you happy:

Back 5

* two kinds of happiness: pleasure and day-to-day, life satisfaction

Front 6

myth: child abuse leads to psychological disorders:

Back 6

* believed because: often these things come together

Front 7

myth: AI is a failure:

Back 7

* believed because: moving goal posts, mysterians (searching for something that agrees with preconceived ideas, things dismissed once implemented even though monumental compared to years ago); almost implemented (things aren’t “true” AI or “real” intelligence)

Front 8

myth: the full moon makes people act differently:

Back 8

* believed because: confirmation bias (notice aggression when moon is present, normal and aggressive behavior with no moon is unnoticed)

Front 9

how science works:

Back 9

* quasi-experiments - observe the real world, problems with causation

Front 10

results and statistics:

Back 10

* significant means probably not due to chance, “alpha level” or typical threshold of <0.05 means that 1 out of 20 experiments will find significance by chance

Front 11

science as a culture: science as an epistemology (theory of knowledge)

Back 11

* science’s self correcting nature - publishing makes it public, other scientist attempt to disprove your theory, bad findings get weeded out as they are tested and retested, modifying nature

Front 12

Daniel Gilbert's TED Talk:

Back 12

* our longings and worries are to some degree overblown because we have the capacity to manufacture the very commodity we are constantly chasing when we chose experience

Front 13

dreaming can occur in:

Back 13

REM (rapid eye movement) state: rapid eye movement, muscle atonia (paralyzation of body so you don’t act out your dream), often dreaming

non-REM states: 75% of our sleep is NREM, short and dull dreams

Front 14

interference from the world while dreaming

Back 14

(Ex. needing to pee, hearing an alarm - these incorporate themselves into your dreams)

Front 15

the dreaming brain:

Back 15

very active brainstem, sensing information forward

DLPFC (dorsolateral prefrontal cortex) is involved with executive function and is deactivated

perhaps explains reduced reasoning during dreams - no shock or noticing of weird things

possibly why we have difficulty remembering dreams - evolutionary survival trait to not mistake dreams for reality)

Front 16

dream recall:

Back 16

* animals and infants cannot report dreams

Front 17

recording dreams

Back 17

* subject to memory biases

Front 18

dream characteristics:

Back 18

dreams are NOT like: films, visual image, recent social situations, recent episodic memories

only realize how bizarre dreams are until we are awake (selection bias - bizarre dreams are easier to remember)

dream emotion matches content - bizarre dream but consistent emotions common

experiences “first person” - always about you, things happening to you

tend to be narrative - have a story, cause, effects

scene shifts - shifting from one scene to another with no correlation

Front 19

Threat Simulation Theory

Back 19

* Westerners dream of things we rarely experience, ancestral threats are over represented

Front 20

dreaming as play theory

Back 20

play as practice for needed survival skills as an adult

Front 21

dream incubation

Back 21

how you can affect your dreams: dream incubation - pre-sleep attention to a specific concern that gets brought into the dream with you to help solve

Front 22

AIM Model for Conscious States

Back 22

* dream recall cessation is almost always caused by forebrain lesions

Front 23

people who have no dream imagery (visual anoneria) tend to also have:

Back 23

people who have no dream imagery (visual anoneria) tend to also have a waking deficit in imaging memories (visual reminiscence)

Front 24

lucid dreaming

Back 24

* training aids in getting lucid dreams (dream diaries, improving dream recall, reality checks)

Front 25

sleep paralysis

Back 25

sleep paralysis - carryover of muscle atonia from sleep to waking, feeling awake, possible chest pressure, unable to move, hallucinations (presence of malevolent character, feeling terror)

Front 26

morals

Back 26

morals - evolved to help us take care of others in our group, not people outside our group

Front 27

the expanding circle:

self interest:

friendship:

tribalism:

I care about all people/creatures able to have positive or negative experiences:

Back 27

* when resources are tight, people are less generous

Front 28

how do we know mortality is evolved?

Back 28

* utilitarianism vs. deontology

Front 29

Haidt’s moral foundations theory

Back 29

* the strength of these moral foundations depend on the person, cultural nurturing, environmental effect and genetic influences

Front 30

politics and morals:

Back 30

* political values are 60% genetic

Front 31

moral dumbfounding

Back 31

* evidence that mortality is based on evolutionary instincts and learning

Front 32

should you trust your instincts?

Back 32

should you trust your instincts? - feeling vs. principles

people look to their feelings to judge whether something is moral or not

bad smells/bitter drinks will make people see things as more immoral

Front 33

animal mortality:

Back 33

* Ex. fairness, reciprocity, friendship in chimps (will kiss/embrace after a fight)

Front 34

solipsism

Back 34

* reality vs. virtual reality (how do you know reality is reality and not a dream)

Front 35

Asimov’s three laws of robotics:

Back 35

Ex. I,Robot (2004) - deals with Asimov’s three laws of robotics:

1. a robot must protect its own existence as long as it does not conflict with the first or second laws

Front 36

AI rights:

ethical agents

ethical patients

Back 36

* ethical patients - something that should be treated morally

Front 37

AI rights:

pain and suffering:

Back 37

pain and suffering:

identity theory - the mental state is the brain state, a physical definition of the mental state, pain is the process of the brain reacting

functionalism - the idea that mental states are defined by the functions they perform, we define pain functionally (Ex. distress and avoidance shown by a functioning system in the presence of a harmful element)

if a robot can feel things is it okay to mistreat them? what if it is the only way for them to function properly? do they become moral/ethical patients that we need to be concerned about?

Front 38

what is a model?

Back 38

across fields, a representation of something that excludes unimportant detail and information, but makes other bits of information easier to get

Ex. scale model of a home made of cardboard, simulation of a hurricane, models of cognitive thought (modeling of human beings and how they behave)

Front 39

cognitive models

Back 39

cognitive models - typically a computer program that models some aspect of thought (Ex. model of how people do categorization, how a mouse learns to navigate a maze)

model makes predictions that can then be compared to data

if predictions match the data, it supports the theory underlying the model

Front 40

cognitive architecture

Back 40

* typically includes constraints on how cognition works in ALL people (speed of learning, memory retrieval, ignores cultural and other learned aspects)

* a theory about the structure of the human mind

Front 41

world issues with:

problems on decline:

Back 41

* maybe seem to not be in decline due to the availability heuristic

Front 42

world issues with:

problems getting worse:

Back 42

social capital (bad real life social networks, biggest factor developed world happiness is in the level of quality face to face social connections

environmental damage - Ex. how do we fix climate change?

technological solution:

come up with an alternative, safe energy source

fund science and engineering to do it

get more people to care so our representatives will make it happen

get money, market the problem

this one problem involves many sub-problems which require solving, why it cannot be fixed with one step

social solution:

convince people to use much less oil

still need to market the problem - need money for this

Front 43

issues with solving problems - all problems are intellectual:

Back 43

all problems are intellectual: (why cogsci is the most important, it can be applied everywhere and help us solve these problems)

the reason we can’t solve all the problems in the world is because we ultimately don’t know how to do it

figuring out how to do it requires thinking and problem solving

cogsci helps us

Front 44

cognitive science and problem solving:

Back 44

human modeling - cognitive science studies how people actually solve problems and all of the other cognitive functions it takes to do it

designing AI/programs that can solve problems for us, and better than us

AI is better than us when: stock trading, arithmetic, statistics, scheduling, search engines and aggregation, certain games

we are better than AI with: language, physical movement, creativity, science, social interaction, vision, certain games, most everyday tasks

but AI often makes humans more effective when used as a tool

Front 45

Richard Dawkins's TED Talk:

Back 45

how we understand and our human capability to understand the world

cogsci help us understand our limitations and biases in understanding what the world is like, cogsci also helps us get past those biases

questions whether there are things that will be forever ungraspable by our and even superiorly intelligent minds?

“Our brains have evolved to help us survive within the orders of magnitude of size and speed which our bodies operate at.”

it is useful for our brains to construct notions of solids and impenetrability because such notions help us navigate our bodies through the middle sized world in which we have to navigate

middle world - the medium-scaled environment (small being at the atomic level, large being at the cosmic level) in which we’ve evolved the ability to take act

reality for an animal is whatever the brain needs for it to properly interact and survive within the world (Ex. a flying animal needs a different world model than a land or water animal)

we live in a social world, a social version of middle world, we are evolved to second guess the behavior of others by become intuitive psychologists

relevant to us as many majors are fields trying to understand something about the world

our cognitive system affects how we “see” the world

our fields affects how we see the world

Front 46

kinds of cognitive architecture:

symbolic

Back 46

kinds of cognitive architecture:

symbolic - operates at the level of discrete symbols

production systems:

Ex. assuming words beginning with capital letters are variables and other words are constant

characteristics:

declarative/procedural memory distinction:

justifications: HM and other brain damage patients

our inability to consciously retrieve and reflect procedural memories

goals are subset of declarative memory

production compilation models automatization

Front 47

kinds of cognitive architecture:

sub-symbolic

Back 47

sub-symbolic - operates using smaller, more detailed number representations, which in aggregates constitute symbol (like a pixel of an image)

associative network - uses Hebbian learning to learn patterns, when it gets incomplete input it can complete it based on the weights between the model (associations between activations that happen simultaneously strengthen weights)

early connectionism - using sub symbols to form letters, which are then connected with weights to possible words

connectionism - supervised learning uses backpropagation, great for classification, needs thousands of examples to work, unlike human beings

input, hidden and output layers; connection weight strength depends on the repeated activity between connections

the brain recognizes certain elements in what it is experiencing and connects it to further elements, the strength/weight of these connections determine the final assessment of the experience

Front 48

kinds of cognitive architecture:

neural modeling

Back 48

neural modeling - Ex. nengo and spaun - simulated neurons at a biological level, spaun took in visual information and interpreted it so that it could draw a resultant with a neuron driven arm

Front 49

kinds of cognitive architecture:

hybrid symbolic/sub-symbolic

Back 49

hybrid symbolic/sub-symbolic - not based

Front 50

kinds of cognitive architecture:

brain

Back 50

brain - models of cognition at the level of biology, but speak to cognitive issues