Cog Sci Final

implicit memories

(without conscious recall) memory that influences one's behavior or thought but does not itself enter consciousness: cannot be verbally transmitted

Brain Computer Interface

-"neural prosthetics" -computer chip is implanted in motor cortex and communicates directly with external device -allows animal (or person) to directly control a robotic arm with their thoughts -has been used in tandem with exoskeletons (robot suits) to restore some movement to paralyzed patients. The user's EEG waves are read and translated into signals that control the robot's limbs. At the same time, feedback is sent to user's brain from sensors in soles of robot's feet -a cap with electrodes can now be used instead of implants, but an extensive calibration process is required -could potentially be used to replace input devices such as mouse, keyboards, touchpads. Would result in faster information exchange between human and machine

Learning the English Past Tense in Children

-3 stages 1. They employ a small number of very common verbs in the past tense (e.g. "got" "gave" "went" "was") Most of these verbs are irregular, and assumption is that children learn these by repetition of it. At this stage, children are not capable of generalizing from the words they have learned. They also tend not to make too many mistakes. 2. They use a much greater number of verbs in the past tense, some of which are irregular but most of which employ the regular past tense ending of "-ed" During this stage, they can generate a past tense for an invented word (e.g. "ricked") Children at this stage take a step backward and make mistakes on the past tense of irregular verbs that they had previously given correctly (overregularization errors: applying a regular past tense to a word that doesn't apply), ex. saying "gived" instead of gave bc -ed is the regular application for past tense but it doesn't work in this case) 3. They learn more verbs and cease to make overregularization errors

expression of emotions

-90% or more of emotional communication occurs through nonverbal channels, e.g. facial expression, gesture and voice. Some researchers, like Raymond Dolan, have pointed out that emotions, unlike thoughts, are embodied -Ekman has identified seven basic emotions: surprise, fear, disgust, anger, happiness, sadness, and contempt. Each is expressed by a unique set of muscle contractions in the face. These seven emotions are universal. Ekman study on cross-cultural similarities in emotional expression -children who are born blind, or both blind and deaf, manifest emotions in the same basic way as sighted children. ***Expressions do not have to be learned through observations!!! -there are cross-cultural differences in display rules: this is when and where it is considered appropriate to display particular emotional expressions (e.g. when one uses certain gestures). ex: eyebrow flash: Japanese, who are more reserved in their social expressions, use it mainly when greeting young children, whereas Samoans greet nearly everyone that way -Can you guess an Olympic athlete's nationality? generally not -- unless they are smiling! Australian smile is more friendly, American smile is more dominant, and British smile is more polite (think Prince Charles) -differences have also been found between the emotional expressions of Japanese-Americans and Japanese nationals. Two groups were indistinguishable when displaying neutral facial expressions

new jobs created by AI

-AI designer, software engineer, cybersecurity developer, machine relations manager. -unfortunately, ppl whose jobs are taken away may not necessarily be able to get jobs created by AI development -out of 10 most happiest jobs, 7 are related to AI: quality assurance analyst, .NET developer, marketing specialist, senior software developer, data analyst, IT specialist, systems analyst

Situated vs. Embodied Cognition

-Brooks also differentiates between robots that are situated and those that are embodied. -A situated creature is one that is "embedded in the world, and which does not deal with abstract descriptions, but through its sensors with the here and now" -an embodied creature is "one that has a physical body and experiences the world directly through the influence of the world on that body" -Airline reservation system is situated but not embodied -An assembly line robot that spray paints parts in an automobile manufacturing plan is embodied but not situated - it doesn't interact dynamically or adaptively with the environment

Natural Language processing in AI

-ELIZA: early computer program that could engage in some very elementary forms of conversational exchanges: -not capable of anything that really resembled linguistic understanding -simply programmed to respond to certain cues by making one of a small set of responses -basic idea behind ELIZA was to create the illusion of conversation by rephrasing statements as questions and by programming the computer to give certain fixed responses where this is not possible -depending upon who one asks, ELIZA was either based upon or intended to parody typical conversational exchanges between psychotherapists and their parents

identifying microexpressions

-Ekman: student of Silvan Tomkins, Princeton professor whose ability to read people's emotions was legendary. He identified 46 distinct muscular movements (action units) in the face. Facial Action Coding System (FACS) -emotion is conveyed by tone of voice. Physician's malpractice study by Nalini Ambady: research participants were able to very accurately predict which doctors would later get sued based on the pitch, intonation, and rhythm of their speech

Applications of AI

-Healthcare: 1. healthcare robots can take orders from and delivery items to a patient, act as an around the clock sitter, assist frail and elderly patients out of a bed or chair, and provide a video connection to a distant doctor (e.g. in remote clinic or cruise ship) 2. diagnosis of disorders. MIT study found that AI was equal or better than radiologists at reading mammograms for high risk cancer lesions needing surgery. This reduced the number of unnecessary due to inaccurate diagnosis surgeries by more than 30% compared to existing approaches. -UCSD study also found that AI outperformed junior physicians (but not senior physicians) in diagnosis of childhood illnesses like mono, the flu, and chicken pox -Google AI algorithm found to be as effective as ophthalmologists in diagnosing diabetic retinopathy, a leading cause of blindness. -AI analysis of smartphone images can now be used to diagnose diabetic retinopathy with high degree of sensitivity -*potential hazard: a radiologist who misreads a scan may harm one patient, but a flawed AI system in widespread use could injure so many -ex: IBM watson can be used to determine optimal treatment, including type and dosage of drugs, and best diet for the individual e.g. to avoid glucose spikes after eating. This is important because this may vary depending on patient's unique gut microbiome which means ppl's glucose spikes in different ways after eating and this program would tailor diets for individuals. 3. precision surgery without human artifacts like shaking hands. A study comparing computer-controlled robots with human surgeons in performing intestinal surgery on a pig found that the robot sutures were much better -- more precise and uniform with fewer chances for breakage, leakage, and infection

polygraph lie detectors

-a method of lie detection -galvanic skin response (more or less electrically conductive) -work by measuring several of the physiological responses accompanying emotional arousal (perspiration, heart rate, blood pressure, breathing), assumption being that lying is stressful -participant is asked a control question and a relevant question. If relevant question has a spike as opposed to the control, then the participant is lying

reconsolidation

-a process in which previously stored memories, when retrieved, are potentially altered before being stored again - Researchers are experimenting with manipulating reconsolidation to treat people with traumatic memories -people are asked to recall the traumatic or negative experience. At the same time, they are given propanolol, a memory blocking drug or a brief painless electric shock which disrupts reconsolidation of the memory erasing it in part. -treatment is still in experimental stages at this point

mindreading

-ability to understand other people's mental state -allows us to make sense of other people -allows us to coordinate our behavior with theirs -*key to human social interaction -*roots of mindreading in early childhood lie in pretend play

Scary Statistics on medical students/interns

-according to the center for disease control, being awake more than 24 hours impairs performance as much as having a blood alc level of .10% which is legally drunk. Interns work 80 hours a week typically in shifts of 30 hours -medical students trained on traditional schedule, rather than alternative schedule with fewer weekly hours, made 5.6 times more diagnostic errors

Total Physical Response (TPR)

-acquiring a language through movement -ask language partner to give you a list of commands or actions -can use hands to pantomim movements -can also use gestures to represent words -do 50-100 per session

FEELIX GROWING

-affective computing -consortium of universities and robotic companies across Europe that are developing roots that learn, interact, and respond to humans as children might do -programmed to learn to adapt to the actions and moods of their human caregivers -programmed to become attached to human agent -able to express anger, fear, sadness, distress, happiness, excitement, and pride -research is also exploring nonverbal cues and emotions associated with physical postures, gestures, and body movements -aim is for robots to provide caregiving and companionship, e.g. in a hospital setting

FaceSense

-affective computing -program at MIT analyzes facial expressions, head gestures, etc. Along with top-down predictions of what the expected affect should be like in a given situation, can be used to evaluate customer satisfaction. Also used as a training tool for autistic spectrum disorders

Robot "Nao"

-affective computing -will begin replacing tellers at a number of bank branches in Japan -has cameras in his face to analyze customers' facial expressions -has microphones to judge their mood by tone of voice -can greet customers in 19 different languages and ask which service they need -**construction of these types of robots requires a large team of different experts, including psychologists, linguists, computer programmers, etc.

brain fingerprinting

-alternative method of lie detection -EEG technique based on idea that brain emits a particular type of brain wave pattern when one views emotionally significant material that one has seen before -some studies have found high accuracy rates, but overall, controversial and of questionable validity

fMRI brain imaging

-alternative method of lie detection -in general, deception is associated with increased activity on both sides of prefrontal (linked to response inhibition) Prefrontal is inhibiting natural response when you lie -however, particular brain regions that are activated depend on various factors, such as whether lie is spontaneous or memorized and well-rehearsed, and whether lie has high or low salience -in general, not widely used at this point as it is very expensive and of questionable accuracy

guilty knowledge test

-alternative to the polygraph test -relies on the premise that criminals harbor concealed knowledge about the crime that innocent people don't -ex: were you at the scene of the crime? What kind of clothes was the deceased wearing? They will show you stuff from the scene of a crime and if suspect shows spike in physiological measures only to items present at crime scene but claims they were not there, that indicates that they may be lying

SHAKEY

-an example of GOFAI (Good Old-Fashioned Artificial Intelligence) Robotics -early robot developed by Stanford Research Institute -called SHAKEY because of its jerky movements -first robot that was able to move around, perceive, follow instructions, and implement complex instructions in a realistic environment (as opposed to just in a virtual micro-world like SHRDLU) -software that allowed SHAKEY to operate was run on a separate computer system that communicated with SHAKEY via a radio antenna -Programs permit SHAKEY to plan ahead and learn how to perform tasks better -1. physical environment in which SHAKEY operated was a suite of rooms (overall about 40 ft x 60 ft) that were empty except for some boxes that SHAKEY could move around -first robot that used a layered architecture: complex behaviors are hierarchically organized: -2. low-level actions (LLAs) are SHAKEY's basic behaviors e.g. rolling forward or backward, taking photos with its onboard camera, moving its head -3. intermediate-level actions (ILAs) include chain ILA's and could recruit other ILA's ex: GETTO action routine calls upon the NAVTO routine for navigating around in the current room, as well as the gotoroom ROUTINE 4. STRIPS planner (Stanford Research Institute Problem Solver): -similar to Newell and Simon's General Problem Solver (means-end analysis) -translates a particular goal, e.g. fetching a block from an adjacent room, into a sequence of ILA's 5. PLANEX monitors the execution of the plan ex: calculates the degree of error at a certain stage of executing a plan, on the assumption that each ILA would introduce a degree of "noise" When the degree of error researches a certain threshold, PLANEX instructs SHAKEY to take a photo to check its position

container test

-child is shown a familiar kind of container (M&M's bag) that contains an unexpected object (marble) -asked to predict what other person will think is inside

Brain Health

-anticholinergics such as Tylenol PM, Benadryl can trigger or worsen Alzheimer's. It's also associated with reduced brain volume and lower levels of glucose metabolism particularly in the hippocampus -older people with tooth and gum disease score lower on memory and cognition tests potentially bc inflammation in diseased mouths migrates to the brain -physically active people have lower risk of developing disease! Heart health leads to brain health -studies found people who keep their minds active also tend to show less loss of intellectual functioning in general as they age. Nuns focus on education and intellectual activity seem to protect against Alzheimer's. Study also found degree of sentence complexity and amount of positive effect expressed in writing samples when subjects were in their 20's were negatively associated with incidence of disease and positively associated w longevity. -*Use it so you don't lose it!*

treatment of psychological disorders using AI

-apps that can administer cognitive behavioral therapy for disorders like depression or social anxiety e.g. WOEbot -virtual therapists such as Ellie -these programs may analyze tone of speech, breathing pattern, smartphone keystrokes and communication, and/or physical movements in making diagnoses and generating responses -pros: easy accessibility and affordability. Research has indicated that people would rather share their innermost secrets with an avatar than a human being -cons: can't really replace human empathy, adherence to treatment may be poor -using biofeedback/virtual reality to treat anxiety disorders or ADHD

Bayesian Language Learning

-argues against innatism by showing how much can be learned through sensitivity to statistical regularities in heard speech -one of the most basic challenges in understanding speech is word segmentation: segmenting a continuous stream of sounds into individual words: -in english, an actual physical event, such as a pause, marks a word boundary less than 40% of the time. So, how does an 8 month old infant (which is when this skill starts to emerge) figure out which combinations of syllables make words, and which ones don't? Answer: This can be explained by model of transitional probabilities. The transitional probability between any two sounds is the probability that the second will follow the first sound. High transitional probabilities will tend to indicate syllables occurring within a word, while low transitional probabilities will tend to occur across the boundaries of words. Infants are exquisitely sensitive to the frequency of correlations, and they exploit this sensitivity to parse streams of sound into words -Transitional probabilities are also used by adults to map the boundaries of phrases

source amnesia (source confusion)

-attributing to the wrong source an event that we have experienced, heard about, read about, or imagined -ex: case of donald thompson, a psychologist studying memory who was accused of rape after being interviewed on live television

dishabituation paradigm

-basic idea is that infants look longer at events that they find surprising -proves that babies understand object permanence bc babies show surprise when object seems to just disappear, demonstrating rudimentary understanding of object permanence -if infants perceive the event as we do, then they will look longer at the impossible or "magical" event than at the possible event -ex: Baillargeon's drawbridge experiment illustrating violation of expectation (VOE)

Rodney Brook's robot Allen

-basic layer is obstacle-avoidance layer -over time, more and more layers were added, mimicking how evolution works -semi-autonomous sbsystems operate relatively independently of each other, though some subsystems can override others -there is not central "controller" comparable to PLANEX in SHAKEY maintaining a continuously updated model of the world and its own state -direct perception-action links allows robot to deliver immediate motor responses to sensory input

Effect of Early Exposure of language on cognitive development

-by age 3, a child growing up in poverty would have heard 30 million fewer words in his home environment than a child from a professional family -also, the greater the number of words children heard from their parents or caregivers before they were 3, the higher their IQ and the better they did in school -tv talk not only doesn't help, it is detrimental. Study found that infants (12-18 months) actually could not learn vocabulary by merely watching DVD in which spoken words were linked with appropriate objects -*some researchers have argued that the racial and socioeconomic gap in academic performance can be wholly accounted for by disparities in exposure to language alone. New intervention programs target this problem

cognitive interview technique

-can be traumatizing -reinstate physical and emotional context of the crime as fully as possible, e.g. by returning to the scene of the crime for the interview -ask witnesses to visualize scene and describe every detail before detective begins questioning. -*technique has generally been found to result in recall of 30% to 35% more information without any increase in erroneous recall

inderal

-can make you dizzy -beta blocker that was originally approved for treatment of high blood pressure; off label (prescribed for one thing but you can prescribe for a lot of other things) use for treatment of social anxiety -many professionals take this medication before important presentations or tests. -professional musicians given inderal felt more in control, had heart rates that were 40 beats a minute slower, and gave performances that were rated significantly higher by judges -SAT scores of hs students with test anxiety went up an average of 120 points when given this medication -*raises question of should drugs like adderall or propranolol be legal given risks involved in using psychometric medications? Does this given performers or athletes an unfair competitive advantage?

adderall

-can make you lose weight, serve as behavioral therapy -prescription medication for adhd that is often used by college students to boost academic performance -*raises question of should drugs like adderall or propranolol be legal given risks involved in using psychometric medications? Does this given performers or athletes an unfair competitive advantage?

How do we test the claim of innatism of language learning?

-can try to disprove by constructing models that stimulate the trajectory of human language learning without explicitly representing any rules -Connectionist approach

eidetic imagery (photographic memory)

-characterized by relatively long-lasting and detailed images of visual scenes that can sometimes be scanned and "looked at" as if they had real existence -ex: groups of children were shown picture for 30 seconds. picture was taken away and children asked to describe what they saw if they could still see anything -rare, only 5% or so of tested schoolchildren have it -not an especially useful form of mental activity

evidence of massive modularity hypothesis

-comes in part from research indicating that humans tend to be much better at reasoning with deontic conditional than they are with ordinary, nondeontic conditionals -deontic conditionals have to do with permissions, prohibitions, etc. ex: if you are drinking beer, then you must be over 21 years of age. -This phenomenon is illustrated by the Wason card problem: Statement: "if a card has a vowel on one side, then it has an even number on the other side" Which cards below would you need to turn over in order to find out whether this rule is valid? Answer is A and 9 out of A,K,6,9

dynamical systems

-complex -self-organizing -emergent -nonlinear -basically relationship between a robot and its environment is what makes it work ex: when our brain issues a command to move our hand to grasp an object, the perception we receive of our initial movement will feed back and alter our subsequent movements, allowing us to fine-tune our movement -non-predictable -chaotic systems, such as the brain or mind, can exhibit a wide range of possible behaviors -it is not a random process, but neither is it as predictable as a simple linear system; it only has short-term predictability -a defining feature of chaos is sensitivity to initial conditions -two slightly different starting conditions in a chaotic system will eventually diverge to produce two very different outcomes

evidence in support of comprehensible input

-complexity of language learning, e.g. of vocab and grammar, wipes out skill building as a possibility. Average native English speaker knows 40,000+ words -study found that second language readers who read a lot have larger vocabularies than native speakers who didn't read a lot -it's possible to acquire language without any conscious learning -Implications: -what is of primary importance is NOT pushing a person to speak from day 1. Rather, it is listening and picking up comprehensible input

AI to treat Autism

-computerized glasses that provide specialized feedback to coach children with autism in emotional identification and social skills such as making eye contact

affective computing

-computing that relates to, arises from, or deliberately influences emotions -goals: to create computers and robots with ability to recognize emotions in people. ex: computer could tell when you are tired or frustrated based on your facial expression and recommend a rest break, and to imbue machines with ability to express emotions -Cynthia Breazeal's Kismet project was a first step in this direction -FaceSense program -A number of computer programs already exist that are capable of recognizing human emotion from spoken auditory information alone. Use neural networks to process features in speech, such as energy, speaking rate, and fundamental frequency. Can recognize emotions with about 79% accuracy - equivalent to human-level performance. As mentioned earlier, programs like these have been used to develop virtual therapists and chatbot therapists. Part of the difficulty of creating simulations is that people are extremely sensitive to subtle deviations - early roboticists had tough time creating robots that didn't creep people out -FEELIX GROWING -Robot "Nao" -Paro therapeutic robot seal

Micro-worlds

-consists simply of a number of colored blocks, colored pyramids, and a box, all located on a tabletop -SHRDLU can carry out various actions through a virtual robot arm ex: pick up blocks and pyramids, move them around, and put them in the box -in order to do that, SHRDLU needs to be able to carry out: -syntactic analysis: needs to parse the sentence to work out which units in a sentence are performing which linguistic functions, e.g. nouns (picking out objects) and verbs (characterizing events and processes) -semantic analysis: needs to assign meanings to individual words in a way that reveals what sentence is stating -integration of the information acquired with the information the system already possesses in order to obey a command or answer a question

posterior hippocampus

-contains place cells, which are neurons involved in spatial navigation -volume of the posterior hippocampus of London taxi drivers was found to be larger than that of control participants -also, the longer an individual had been a taxi driver, the larger was the volume of the posterior hippocampus

home robots

-cooking, cleaning, fetching -increasingly used in instances that can be described as dangerous, dirty or dull -in retirement homes, robots can provide interaction with the elderly who would otherwise be devoid of human interaction for long periods of time -home robots can also permit family members to visit elderly parents or elderly parents to visit with the family -application of AI

how to defeat polygraphs

-create a baseline spike during control question by pricking urself or something

use of neuroscientific evidence to predict criminal behavior

-decreased prefrontal activity and increased subcortical activity (including amygdala) have been found in brains of convicted murderers -people with antisocial personality disorder showed an 11% reduction in volume of gray matter in prefrontal cortex -research indicates that frontal lobe damage sustained in childhood through abuse may be mediating factor -less activity in frontal lobe in kids who grew up in bad homes, but kids who have strong relationship with one parent, it won't happen! -people with antisocial criminal tendencies, the emotion-controlling amygdala is smaller, but shows greater reactivity -compared with people who feel and display empathy, brains of those with antisocial personality disorder also respond less to facial displays of others' distress -also, those with the low-activity form of the MAOA gene which INHIBITS aggression, tend to behave aggressively when provoked (ex. one study found that those who have the low-activity as opposed to high activity form of the MAOA gene AND were abused as children were SIX TIMES more likely than normal controls to be convicted of violent crimes -neurological evidence could thus potentially be used in determining criminal sentence length or need for personal rehabilitation

recurrent neural networks

-designed to deal with time series and sequence data -there are feedback connections that result in the output at Time 1 serving as the input at Time 2 -this functions as a type of memory -can be used to model object permanence: -activation associated with "sight" of the hidden object at a previous temporal stage is transmitted to the current stage -this information is then used to predict what the next set of inputs will be -the network's learning (which works via standard backpropagation) is driven by the discrepancy between the predicted input and the actual input -as training progresses, the network becomes increasingly proficient at predicting the reappearance of occluded objects over longer and longer periods of occlusion -because the network learns gradually, sensitivity to object permanence is a graded phenomenon -in the early stages of training, it may be strong enough to drive perceptual "expectations" but too weak to drive motor behavior (e.g. manually search for a hidden object)

AI for psychotherapy

-diagnosis/identification of psychological disorders -AI system that analyzed fb posts of consenting patients in an emergency department was able to generate predictions of depression risk that were as accurate as standard depression screening tests. Indicators included references to sadness, loneliness, hostility, rumination, and increased self-reference, e.g. words like "alone" "ugh" "tears" and higher frequency of the use of "I" and "me" Length and timing of posts were also considered -AI has also been used to analyze Instagram photos to successfully screen for depression. Photos posted by depressed ppl tended to be bluer, darker, and grayer. The more comments instagram posts had, the more likely they were posted by depressed ppl, but the opposite was true for likes received. Depressed participants were more likely to post photos with faces, but had a lower average face count per photo than healthy participants. -The screening models created from the data were able to outperform general practitioners in correctly diagnosing depression without the assistance of assessment instruments

Alzheimer's Disease

-disease occurring in latter part of life that is characterized by deterioration of memory, reasoning, and language abilities -common form of neurocognitive disorder: deterioration of intellectual abilities; another common cause is vascular dementia. This occurs in 7% of population above age of 65 and up to 40% of people older than 80 years -Associated with loss of neurons in cortical and sub-cortical regions: ventricles may be enlarged- patients might end up losing as much as 50% of their brain mass. Can provide severe degeneration of large parts of the brain: can eventually destroy most of the hippocampus and cortical gray matter -Brain of patients contain many amyloid plaques, which contain a core of misfolded Beta-amyloid protein surrounded by degenerating axons and dendrites, and neurofibrillary tangles, dying neurons that contain twisted filaments of tau protein -Deficiencies of acetylcholine: failure to show eyeblink conditioning

example of chaotic systems

-does the flap of a butterfly's wings in Brazil cause a hurricane in Texas?" -weather prediction: -involves plotting paths that particles in atmosphere will take given different initial positions -if two particles start very close to each other, their paths will stay very close together for a short period of time, but eventually the distance between them becomes larger and their trajectory, completely unpredictable -this is an example of "sensitivity on initial conditions" -it is this sort of sensitivity on initial conditions that makes weather difficult to predict -the period of time that we are able to accurately predict the weather is usually around 5-10 days

direct vs. indirect access hypotheses

-dual approach to reading -Do readers recognize a word directly from the princed letters (known as direct access)? Or do they convert the printed letters into a phonological code to access the word and its meaning (indirect access)? -PET study explored this: -condition 1 (looking): Participants asked to focus on a fixation point (a small cross-hair) in the middle of a screen -condition 2a (reading silently): participants were presented with words flashed on the screen but told not to make any response -condition 2b (listening): participants listened to the same words being spoken -condition 3 (reading out loud): participants were asked to say out loud the word appearing on the screen -condition 4 (speaking): participants were presented with nouns on the screen and asked to utter an associated verb ex: saying "turn" when presented with word "handlebars" results: areas of activation in conditions 3 and 4 (speaking words) did NOT include areas in condition 2a and 2b (reading silently and listening to words, respectively) Conclusions: patterns of activation identified across the different tasks thus supported a parallel rather than a serial model of single-word processing. In addition, the results support the direct access hypothesis: we do NOT need to sound words out (or subvocalize) to access meaning of words -Moreover, research in general has indicated that, though readers use both direct and indirect access when reading, direct access is more efficient: -skilled adult readers are more likely to use direct access. -Beginning and less skilled readers are likely to sound out words to understand meaning

jobs that are hard to replace by AI

-emotionally demanding jobs such as babysitters, human resources, politicians therapists although this can depend on the individual bc some ppl like talking to ai -creative jobs such as writers, software/graphic designers

dissociation

-evidence for separate memory systems -when brain damage affects two behaviors very differently, this suggests that the two behaviors are produced by different processes -Clive Wearing, a musician, suffered damage to his hippocampus as a result of encephalitis. If you walk out of the room, then come back 10 minutes later he won't remember you, but he can still conduct a choir and play the piano and harpsichord beautifully

Wason card problem

-evidence of massive modularity hypothesis -shows that people respond better to deontic conditionals -Statement: "if a card has a vowel on one side, then it has an even number on the other side" Which cards below would you need to turn over in order to find out whether this rule is valid? Answer is A and 9 out of A,K,6,9 -*reframing the selection task in such a way that what is being checked is a condition having to do with permissions, entitlements, and/or prohibitions improved performance. ex: Statement: if a person is drinking beer, they must be over 21 years of age. the bouncer should check the person holding a beer and the 16 year old. -73% of students who tried drinking-age problem made correct selections, as opposed to 0% in the standard, abstract form of task (A,K,6,9)

Schemas

-generalized information about a situation or event (e.g. things people do at birthday parties) -in general, people show enhanced recall for schema consistent material but there are exceptions

Are results of polygraph admissible in court?

-generally no, unless both parties agree to its use. -however, commonly used for screening purposes for certain types of jobs, such as law enforcement, and in questioning of suspects -*if a person believes polygraph is 100% accurate, test may approach 100% accuracy

Causes of Alzheimer's

-genetics, -higher risk in those who have had strokes or head trauma, -conditions associated with cardiovascular disease such as obesity, diabetes, high blood pressure, high cholesterol, smoking, and lack of exercise -low levels of vitamin D and certain B vitamins -exposure to lead and toxic substances, such as air pollutants -hearing loss is the largest modifiable risk factor for developing neurocognitive disorders -anticholinergics such as Tylenol PM, Benadryl can trigger or worsen Alzheimer's. It's also associated with reduced brain volume and lower levels of glucose metabolism particularly in the hippocampus -older people with tooth and gum disease score lower on memory and cognition tests potentially bc inflammation in diseased mouths migrates to the brain

the stepping reflex

-example of dynamical system -in the first few months of life, infants are able to make stepping movements -they stop making these movements during the "non-stepping window" -the movements reappear when the infant starts walking at around 11 months of age -this is a U shaped development trajectory -explanation of U Shaped development trajectory: infant's initial stepping movements are purefly reflexive. They disappear during the non-stepping window because the cortex has matured enough to inhibit reflex responses - but is not sufficiently mature to bring stepping movements under voluntary control -challenge to this view: research that indicates stepping movements could be artificially induced or inhibited in infants by manipulating features of the environment. Infants in the non-stepping window will make stepping movements. When they are suspended in warm water and when they are placed on a treadmill. On the other hand, stepping movements can be inhibited before the start of the non-stepping window by attaching small weights to baby's ankle. -Conclusion is that stepping movements vary independently on how the cortex has developed. They also depend on a number of parameters, such as leg fat, muscle strength, gravity, and inertia. -*in other words, walking does not involve a specific set of motor commands that "program" the limbs to behave in certain ways (top down) Rather, the activity of walking emerges out of complex interactions between muscles, limbs, and different features of the environment (bottom-up)

cochlear implants

-example of neural prosthetics aka BCI -external part: sits behind the ear and picks up sounds with a microphone, then analyzes the sound and sends signals to the internal part of the implant -internal part: stimulates different regions of the basilar membrane, producing perceptions of sounds of different pitches. High frequency sounds cause the base of the basilar membrane (near the oval window) to flex. Low frequency sounds cause the apex (opposite end) to flex, and the whole basilar membrane also flexes back and forth in time to the sound wave

subsumption architecture

-examples include webb's robot crickets and the Yokoi hand -these robots do not operate by executing algorithms to map their surrounding, etc. but rather with a set of relatively simple stimulus-response mechanisms. -their intelligence aggregates from the bottom up, rather than being organized explicitly from the top down. -based on idea that intelligence - and consequently, performance of efficacious action - does not require formal symbolic representation -robots constructed with this architecture make reflexive responses to environmental stimuli. Representations exist as production rules (if-then statements) or reflexes that map a stimulus onto a behavior -knowledge does not exist in isolated representations; rather knowledge is embodied -subsumption architectures are made up on layers that are built up from behaviors. ex: obstacle avoidance layer: -directly connects perception (sensing an obstacle) to action (either swerving to avoid the obstacle, or halting when the obstacle is too big to go around_ -whatever other layers are built into the subsumption architecture, the obstacle-avoidance layer is always online and function. For instance, there may be a higher layer that directs the robot toward a food source, but the obstacle-avoidance layer will still come into play whenever the robot finds itself on a collision course with an obstacle

attentional explanation

-explanation for deja vu -fully processed perceptual experience that matches a minimally processed impression received moments earlier produces a strong feeling of familiarity -the original impression may not have been fully processed due to a physical distraction or a mental distraction, such as preoccupation with other thoughts

memory explanation

-explanation for deja vu -implicit familiarity without explicit recollection -ex: seeing a lamp in your friend's apartment that is similar or identical to one that used to be in your autn's house

Dual Processing Explanation

-explanation for deja vu -incoming sensory data follow several different pathways -a slight alteration in transmission speed in one pathway could cause the brain to interpret the data as two separate experiences

Long term memory

-explicit/declarative memory (with conscious recall): recall of recognition of information: can be verbally transmitted. two types: episodic memory: recall of personal facts semantic memory: recall of general facts -implicit/non-declarative memory (without conscious recall): memory that influences one's behavior or thought but does not itself enter consciousness: cannot be verbally transmitted -procedural memory: recall of how to do things

Retrieval failure

-failure to access information that is stored in long-term memory - "forgotten" material is not completely erased but merely inaccessible -common causes: lack of appropriate retrieval cues (due to encoding specificity principle) ex. inability to recognize student from your biology class at a dorm party - repression of painful or anxiety-provoking information. There have been documented cases of individuals who had been treated in hospital emergency rooms for childhood sexual abuse, yet these individuals failed to recall the episode when interviewed as adults

things humans can do that AI cant do well

-feel or show empathy -have insights. As anthony goldbloom said, machines can't connect seemingly disparate threads to solve problems they've never seen before -machines can't tackle novel situations -machines need to learn from large volumes of past data. Percy spencer was working on radar during ww2 when he notices magnetron (used to generate radio signals) melting his chocolate bar which lead to the discovery of the microwave oven -make plans for the distant future. humans can plan their lives years in advance but robots tend to focus only on completing the immediate task at hand. -be conscious. whether machines can have consciousness depends on the definition but no machine today meets all the criteria that we may give to consciousness and that humans have

Infant folk physics

-folk physics: innate understanding of basic principles governing the behavior of physical objects -infant folk physics differs from adult folk physics bc: -infants place more weight spatiotemporal continuity than on featural continuity. this means: -infants won't show surprise if a green scarf disappears behind a screen and a bunch of flowers emerges at the other side of the screen -they show more surprise if a large white rabbit appears in a box, then later in a hat, without seeing a path of movement for it -for adults, in contrast, featural constancy is more important: -if they see a large, distinctive white rabbit appear in a box, then later in a hat, they assume it's the same rabbit, even if they don't immediately see a path of movement for it -if they see a green scarf turn into a bunch of flowers as it passes through a magician's hand while maintaining its trajectory, they assume it is a different object

use of AI in customer service

-help lines -providing information to consumers

Short term memory or working memory

-holds items that are actively being thought about -has limited capacity ex: 7 plus/minus 2 items, number of works you can speak in 1.5 seconds -also limited in time, but longer than sensory memory: lasts 5-30 seconds and information decays rapidly unless maintained in consciousness through rehearsal -working memory is the preferred term -includes a phonological loop associated w the left hemisphere that briefly stores sounds, and a visuospatial sketchpad associated w the right hemisphere that stores visual and spatial information. ex: two visuospatial tasks will interfere with each other if performed simultaneously, as will two items in the phonological loop, but people can perform verbal and spatial tasks simultaneously -episodic buffer: another component of working memory that can hold and combine information from the phonological loop, visual sketchpad and long term memory -central executive: integrates information from phonological loop, visuospatial sketchpad, and episodic buffer: similar to attention/sensory memory

Sensory memory

-holds sensory information very briefly (0.5-4 seconds) -large capacity store -sensory input is held very briefly in sensory memory to allow selection and processing of information -considered to be a part of perception two types of sensory information: Iconic (photographic): visual sensory memory and echoic memory: auditory sensory memory

ACT-R

-hybrid cognitive architecture -Adaptive control of thought - rational -cognitive architecture with modular organization that was developed by John R. Anderson in 1976 -it is hybrid in the sense that it incorporates both symbolic and subsymbolic information processing -computer modelling

Age and Language Learning

-in general, it is much easier to learn a language at an early age before "hardening of the categories" sets in -ex: we are all born with the ability to recognize speech sounds or phonemes (e.g. sound of b vs. p) from all the world's languages but gradually lose this ability: -adult Hindi-speakers and young infants from English-speaking homes can easily discriminate two Hindi t sounds not spoken in English. By age one, however, English-speaking listeners rarely perceive this sound difference -Japanese speakers have difficulty distinguishing between English r and l -at birth or a few weeks after, infants can perceive almost all (95%) of the subtle phoneme differences in non-native languages. However, by 8-10 months, accuracy drops to 70% and, by 10-12 months, to 20%

misinformation effect

-incorporating misleading information presented after an event into one's memory of the event -study in which people were shown a film depicting a traffic accident, then either asked how fast the cars were going when they hit each other or how fast they were going when they smashed into each other. People gave faster speed estimates when word smashed was used. When asked a week later whether there was any broken glass in the accident, people who heard smashed were much more likely to answer yes

Terry Winograd's computer model SHRDLU

-initially presented in his 1970 doctoral dissertation at MIT -illustrates how grammatical rules might be represented in a cognitive system and integrated with other types of information about the environment -one of the first attempts to write a program that was not just trying to simulate conversation, but was capable of using language to report on its environment, plan actions, reason about the implications of what is being said to it -programmed to deal with a very limited virtual micro-world - everything takes place on a computer screen

neurolaw

-interdisciplinary field of study that explores effects of new research findings in cognitive science on forensic psychiatry and legal practice, including eyewitness testimony, lie detection methods including brain-based techniques, use of neuroscientific evidence for legal defense or to predict criminal behavior, use of neuroscience based interventions such as deep brain stimulation, to diminish risk of an offender reoffending, and use of nootropics or mind enhancing drugs to boost performance in sports field

use of neuroscience-based interventions and nootropics

-it is possible that neuroscience-based interventions will be developed in near future that will diminish the risk of an offender reoffending, e.g. deep brain stimulation as a way of reducing sex drive. -may be similar in some ways to treating sexual aggression by using synthetic steroids that inhibit the production of androgens by the testes -What should be the ethical guidelines around that? Currently, people have to consent to take the medicine, even if you're a convicted rapist, but many will take it because they're provided an incentive -adderall -inderal

improving testimony

-justice department has guidelines that are as follows: -line-ups should be administered by "blinded" managers who have no knowledge of which participant is suspected of the crime so they can't give tips even non-verbal ones -witness instructions should be standardized and designed to yield a conservative response. for ex: warning witness that the perpetrator might not be present. Witnesses are often convinced that the culprit must be present in a lineup and so simply pick the person who most closely resembles their memory of the perpetrator. Study found that telling witnesses that perpetrator might not be present reduced number of innocent people who were identified incorrectly by 42% -ask witness to describe what happened before beginning questioning and ask open-ended questions -cognitive interview technique (can be traumatizing): -reinstate physical and emotional context of the crime as fully as possible, e.g. by returning to the scene of the crime for the interview -ask witnesses to visualize scene and describe every detail before detective begins questioning. -*technique has generally been found to result in recall of 30% to 35% more information without any increase in erroneous recall

connectionist approach

-provides an alternative to the rule based conception of language comprehension and learning of nativist approaches -demonstrates that it is possible to learn complex linguistic skills without have any explicit linguistic rules encoded in it. ex: simple recurrent networks have been successfully trained to predict the next letter in a sequence of letters or the next word in a sequence of words -the learning trajectory of these networks strongly resembles the learning trajectory of human infants. ex: learning how to form the past tense of English verbs, both regular and irregular

xenobot

-living robot -small biological machine created by scientists at the University of Vermont and Tufts -built from the ground up using biological cells. Made of skin cells and heart cells in the form of stem cells harvested from frog embryos -designed and programmed by a supercomputer using an evolutionary algorithm. A few hundred simulated cells were reassembled into myriad forms and body shapes. The most successful simulated organisms were kept and refined. -Single stem cells were then cut and joined using tiny forceps and an electrode into close approximation of designs specified by computer -Cells began to work together: Skin cells formed more passive architecture while the once-random contractions of heart muscle cells created ordered forward motion as guided by the computer design and aided by spontaneous self-organizing patterns -are able to move in coherent fashion to explore their watery environment and can survive for days or weeks, powered by embryonic energy stores -Functions: 1. groups of xenobots can move around in circles, pushing pellets into a central location - spontaneously and collectively 2. others were built with a hole through the center and were able to use that as a pouch to successfully carry an object 3. when xenobot was cut in half, it stitched itself back up and kept going -Potential applications: intelligent drug delivery: carrying medicine to a specific place in body, traveling in arteries to scrape out plaque, searching out and break down harmful compounds or radioactive wastes, gathering microplastics in the oceans, serving as new material for technologies that is fully biodegradable unlike steel concrete or plastic that can cause ecological and health problems

development of TOMM

-many stepping stones including the shared attention mechanism (SAM): occurs when infants look at objects (and take pleasure in looking at objects) because 1. they see that another person is looking at that object (ex: I see (mother sees the cup)) OR 2. they see that the other person sees that they are looking at the object (ex. Mother sees (I see the cup)) -requires infant to be able to embed representations - to represent that an agent is representing someone else's representation -*makes possible a range of coordinated social behaviors and collaborative activities: -children with autism have difficulties with this type of joint attention and there is a strong correlation between severity of social impairments and inability to engage in joint attention -*attunement between caregiver and child - child's understanding that caregiver knows how he feels - is critical for normal development

Bilingual people

-may think differently in different languages -Bilinguals reveal different personalities when taking the same personality test in their two languages. China-born students at University of Waterloo were asked to describe themselves in English or Chinese. When describing themselves in English, they expressed mostly positive self-statements and moods. In chinese, they reported more agreement with Chinese values and roughly equal positive and negative self-statements and moods -bilinguals often switch languages, depending on which emotion they want to express -when responding in their second language, bilingual people's moral judgments reflect less emotion - they respond with more "head" than "heart"

applying the idea of morphological computation to robotics

-means building as much of the computation as possible directly into the physical structure of the robot. -ex: Yokoi hand: using a traditional computational approach, grasping an object (e.g. a glass) requires computing an object's shape and configuring the hand to conform to that shape -the Yokoi hand is instead constructed from elastic and deformable materials that allow the hand to adapt itself to the shape of the objects being grasped

AI for the military

-military, police and fire services have used robots to neutralize and destroy bombs, fight fires, and carry out other dangerous missions -can have teams and autonomous aircrafts that go out and do firefighting, either fully autonomously or in tandem with human-piloted aircrafts -devices were initially controlled remotely, with a human supplying the "intelligence" -however, these agents are increasingly being sent into environments where the robots are required to make decisions on their own

recovered memories

-much of the recovered memory literature has focused on cases of child abuse many of which are false -recovered memories can be real! -research has demonstrated that some people may indeed forget about painful childhood memories and recall it years later. There have been documented cases of individuals who had been treated in hospital emergency rooms for childhood sexual abuse, yet these individuals failed to recall the episode when interviewed as adults. Also, case of a college prof who woke up one morning and suddenly remember having been molested by camp counselor who confessed when confronted w the crime -HOWEVER, recovered memories can also be false: -many reported incidences of abuse probably never happened (ex. case of woman who remembered under hypnosis that she was raped by her father) -research has shown that it is not all that hard to implant false memories (ex. study on implanting false memories in children. An adult repeatedly asked child to think about several real and fictitious events and 10 weeks later, a new adult would ask the same question. Found that 58% of preschoolers produced false and vivid stories regarding one or more events they had never experienced)

Problems with polygraph

-not reliable -arousal associated with lying is no different than arousal caused by other stressful situations, such as being accused of a lie -also, some people tend to be more prone to arousal so that they tend to show arousal even when they are not guilty -1/3 of innocent declared guilty; 1/4 of guilty declared innocent

modeling object permanence

-object permanence can be explicitly represented in a body of rules and principles -many cognitive scientists think that object permanence depends upon this sort of explicit symbolic representation -however, Yuko Munakata and her collaborators have suggested an alternative neural network approach to object permanence involving the use of recurrent neural networks

false belief (displacement) task

-one of the best-known tests for mindreading ability -tests whether children are able to abstract away from their own knowledge to understand that someone else can have different (and mistaken) beliefs about the world. ex: sally puts ball in red box and then leaves. anne moves the ball to the blue box. when asked where sally will look for the ball, anne says the blue box bc she knows it's in there. this shows that kids do not understand there is a difference between belief and reality bc sally still believes the ball is in the red box, but anne knows its in the blue. -tests children's theory of mind mechanism (TOMM) - their ability to identify and reason about other people's complex mental states, such as beliefs, desires, hopes, and fears -pretend play emerges during the second year of lie, but children do not typically pass the false belief test until they are nearly 4 -indicates that the BELIEVES operation is much harder to acquire than the PRETENDS operation -however, research by Kristine Onishi and Renee Baillargeon demonstrated that children may develop an implicit understanding of false belief well before age 4 -experiment similar to false belief displacement task measured looking time in 15 month old infants. results indicated that children looked significantly longer - indicating surprise - when actor's behavior violated expectations that someone with an understanding of false belief would have -*suggests that children may develop an implicit understanding of false belief by 15 months, but that explicit understanding, involving explicit conceptual abilities manifested in verbal responses and explicit reflection, develops later

examples of how SHRDLU represents diff types of knowledge

-one of words in SHRDLU's vocabulary is CLEARTOP -we can say that something (say, a block) is CLEARTOP when it does not have anything on it -CLEARTOP can also function as a command to remove anything resting on the block -CLEARTOP is represented by procedure: Go-> does x support an object y? if yes, then get rid of Y and if no then assert that X is CLEARTOP -SHRDLU illustrates how linguistic understanding can result from the interaction of many independently specifiable cognitive processes

why are people better at reasoning with deontic conditionals than with nondeontic conditions?

-one theory states that it's because when they solve problems with the deontic conditionals, they are using a specialized module for monitoring social exchanges and detecting cheaters. -This is the cheater detection module -Why should there be a cheater detection module? answer: cooperative behavior presumably has a genetic basis. However, an individual who takes advantage of cooperators without reciprocating will likely do better than one who cooperates. ex: they gain your trust then steal all your bananas. So how could the genes for cooperative behavior ever have become established? This is where the cheater detection module comes in

Reading Emotions

-our brains are in general amazing emotion detectors. After a quick glance at someone's photo, we have a pretty good sense of the person's extraversion and agreeableness. Watching 90 seconds of people walking and talking will allow us to accurately predict how others evaluate them. -Reading the Mind in the Eyes Test (RMET)- determining emotions based on solely the eyes. Women tend to outperform men on this test

Savant Syndrome

-people who are born with severe intellectual disability but show superior ability in one intellectual domain, such as music, art, or mental arithmetic -about 10% of children with autism have savant talents -analysis of case history of 13 musical savants. All had severe deficits in ability to understand and use language and some were blind, but they all showed an extraordinarily intense interest in performing music beginning at a very young age, usually before 4 -largely attributable to a seemingly limitless memory. -ex. musical savants can play note for note long passages of music heard just once. Artistic savants can reproduce exact copies of animals or people or scenes from memory. Human calculators can tell you the day of the week that corresponds with any given day of any given month and year, past or future. -*Suggests that memory capacities are potentially virtually limitless* -possible to create savant-like memorization skills and artistic abilities in people without autistic traits by disrupting the left anterior frontal lobe with TMS

Hippocampus

-plays an important role in the formation, processing and storing of new explicit memories - "save" button for explicit memories -acts as a loading dock where the brain temporarily holds to-be-remembered information. Items then migrate for storage elsewhere in a process called memory consolidation -removing a rat's hippocampus 3 hours after it learns the location of cheese prevents long-term memory formation, but removal 48 hours does not! -after a training experience, the greater one's heart rate efficiency and hippocampus activity during sleep, the better the next day's memory will be -people with full temporal lobe amnesia (damage to hippocampus and surrounding areas) cannot form new explicit memories though they can form new implicit memories -often have normal IQs and can carry on a normal conversation but cannot remember anything that happened more than a few minutes previously

Frontal lobe

-plays an important role in the processing and storing of new explicit memories -recalling information and holding it in working memory

Cerebellum

-plays important role in formation of implicit memories -involved in learning of procedural memories for skills

basal ganglia

-plays important role in the formation of implicit memories -deep brain structure important in motor sequencing -ex: Parkinson's involves degeneration of parts of the basal ganglia

Object Permanence and Reasoning in Infancy

-proof of weakness of Piaget's Theory that children are more cognitively competent than Piaget recognized -ex: object permanence: -Piaget thought understanding of object permanence developed in infants around the age of 8 months. -However, Renee Baillargeon argued that Piaget's finding was rooted in the lack of motor ability in infants since experiments required infant to manually search for the hidden object by pulling a cover off to reveal the object. -More recent studies (described on succeeding slides) have indicated that infants as young as 3.5 months of age and perhaps younger understand that objects continue to exist when hidden, that they can't just disappear) -babies show surprise when object seems to just disappear, demonstrating rudimentary understanding of object permanence

algorithms for predicting recidivism

-recidivism: which defendants will later be arrested for a new crime -new research indicates that complex algorithms are significantly more accurate than humans in predicting which defendants will later be arrested for a new crime -when assessing just a handful of variable in a controlled environment, even untrained humans can match the predictive skill of sophisticated risk-assessment instruments -however, when a larger number of variables are involved - as generally happens in the real world - the algorithm-based tools perform far better than people -these tools consider factors such as pre-sentence investigation reports, attorney and victim impact statements, and an individual's demeanor -tools approach 90% accuracy in predicting which defendants might be arrested again, compared to about 60% for human predictors -could aid in making critically important decisions such as: which people can be rehabilitated in the community rather than in prison? Which could go to low-security prisons, and which to high? Which prisoners can be safely released to the community on parole?

use of AI in finance and economics

-record keeping -fraud detection -optimizing profits in online trading -predicting market supply and demand

Jerry Fodor's The Modularity of Mind

-rejected the idea that the mind is organized in terms of faculties such as memory and attention that can process any type of information -instead, proposed the existence of specialized information-processing modules for things like color perception, shape analysis, visual guidance of bodily motions, grammatical analysis of heard utterances, detecting melodic or rhythmic structure of acoustic arrays, and recognizing the voices of conspecifics -*modules operate at a low level and work quickly to provide rapid solutions to highly determinate problems -*in addition, there are high-level, open-ended nonmodular processes that can bring a wide range of information to bear on very general problems

objections to traditional GOFAI approach to artificial agents like SHAKEY is

-robot is not embedded in a real-life environment and can never really come to terms with real-life problems and challenges -they can only operate in a highly constrained environment -all gofai cannot learn to solve problems - all solutions to problems are built in. ex: cannot look at a photo it has not seen before of a person in a room with a banana out of reach and suggest a plan of action. Even a young child can solve such an issue, but classical AI can't -*to address these issues, situated cognition theorists propose a dynamical systems like approach to robotics where behaviors emerge out of complex interactions between an organism and its environment

Ellie

-robot therapist -can gage body activity, gaze attention, hesitation

Plunkett-Marchman Multilayer Neural Network Model of Tense Learning

-similarly produced overregularization errors as children, but without sudden increase in size of training set -however, because this is a neural network, it is not really biologically viable since human neurons don't engage in backpropagation *main point is that it is possible to devise neural networks that reproduce typical trajectory of language learning without having encoded into them explicit representation of linguistic rules which suggests that nativist view of language learning is not the only viable model

Situated cognition

-situated cognition theorists propose a dynamical systems-like approach to robotics -they believe that we should start small and focus on basic ecologically valid problems -for instance, studying insects can allow us to better understand how organisms interact with their environment. Insects achieve high degrees of "natural intelligence" by exploiting direct connections between their sensory receptors and effector limbs. ex: female crickets are extremely good at recognizing and locating mates on the basis of the song they can make (phonotaxis). Seems like program to do this would be very complex - one would need to identify the sound, work out where it comes from, then form motor commands that will take the cricket to the right place. However, turns out crickets are simply hard-wired to move in the direction of the ear with the highest vibration (provided that the vibration is suitably cricket-like) There is no "direction-calculating mechanism" no "male cricket identification mechanism" and no "motor controller" -Barbara Webb has used this model to build robot crickets that can identify the source of a sound and move automatically toward that source without any of the systems normally assumed by gofai. This is an instance of biorobotics: using knowledge of living insects, as well as ai, to create agents capable of moving about and solving problems in their environment. It is an example of morphological computation: exploiting features of body shape to simplify what might otherwise be highly complex information-processing tasks

effects of sleep deprivation on cognition

-sleep deprivation causes irritability, fatigue, impaired concentration and creativity, greater vulnerability to accidents -surprisingly, people are oftentimes unaware that their concentration, judgment, etc, are impaired -impairs functioning of the prefrontal cortex which has a negative impact on attention, memory, and decision making -reduces neuroplasticity and the proliferation of cells in the hippocampus, which can result in memory impairments. without sleep, the brain was 40% less able to make new memories -impairs process of making broad connections and gaining creative insight. Participants were presented with a task where discovery of a hidden rule greatly improved speed of performance. Sleep dramatically increased the likelihood of grasping the hidden rule -*Sleep causes our brain to create new links, which is why we often wake up with solutions to previously unresolvable problems. -Sleep also plays an important role in emotional regulation. Aftera night of no sleep, brain scans show a shutdown of the medial prefrontal cortex, which normally helps keep our anxiety in check. Research following youth through time found that sleep loss predicts depression rather than vice versa -can produce hallucinations, delusions, and paranoia (Peter Tripp's wake-a-thon) -case metabolic and hormonal changes that mimic aging and lead to diabetes, obesity, hypertension, and memory impairment (18 year old looks like 60 year old in ability to metabolize glucose after 2 weeks of restricted sleep) -depresses immune system and is associated with shorter lifespan

treatments for insomnia

-sleep restriction: go to bed and get up at the same time everyday -associate bed with sleep (only lay on bed when time for sleep) -relaxation response training

Chomsky's Model

-sophisticated cognitive ability, such as use of language, involves stored bodies of information (about phrase structures and transformation rules). These bodies of information can be manipulated algorithmically -however, humans are also born with a specialized language acquisition device, that is, they are prewired to learn language. ex: baby kittens and human babies are both exposed to the same language, but human (almost) always acquire the ability to understand/produce language while cats do not -all human languages can be understood in terms of different parameter settings in a universal grammar -the universal grammar is an innate fixed structure that holds across all languages -the parameter settings are language specific and learned through hypothesis formation and testing -This is a nativist view (i.e. view that language learning is innate) -it is based on poverty of stimulus arguments: young children are simply not exposed to enough information to allow them to learn a language. Much of the speech that children hear is actually ungrammatical, but not flagged as such. Children are typically only exposed to positive information, i.e. they are not told what counts as ungrammatical

repisodic memory

-specific case of schema -recall of a supposed event that is really the blending of details over repeated and related episodes. For ex. if asked to recall last Monday's lunch, you might produce a repisodic memory based on what you usually do for lunch

Speech perception and AI

-speech perception is an extremely complicated process (which is why computer voice recognition systems are often problematic) -need to separate voice of speaker from irrelevant background noises, which might include other simultaneous conversations -pronunciation varies, depending on vocal characteristics of speaker -speakers often slur or mispronounce words -pronunciation of specific phoneme depends in large on previous and following phonemes, e.g. d in idle vs. d in don't -as mentioned, an actual physical event, such as a pause, marks a word boundary less than 40% of the time. Children's. mispronunciations of lyrics in Christmas carols and Pledge of Allegiance -people use visual cues to facilitate speech perception. Study in which participants watched video of woman making on sound (ga) while different sound played (ba). Responses reflected compromise where participants heard da -Lipnet, a machine developed at Oxford, can now also lipread with 95% accuracy -computers can now replicate human voices extremely accurately. Lyrebird company created a program that can replicate voices of people including powerful political figures after analyzing only one minute of audio -it's more challenging to replicate associated facial movements. Creating a video of Obama required 14 hours of Obama high quality footage to train system to translate audio into mouth shapes

applications of RNN

-speech recognition -speech synthesis -machine translation -music composition -time series prediction -robot control

Human Brain Project

-sponsored by European Commission -aims to: 1. simulate the brain; generate digital reconstructions and simulations of the mouse brain and ultimately the human brain. 2. implement models of the brain in neuromorphic computing and neurorobotic systems 3. develop a model of the brain that merges theory (top-down) and data driven (bottom-up) approaches for understanding learning, memory, attention, and goal-oriented behaviors 4. develop tools to explore new diagnostic indicators and drug targets

false memories

-studies show that it's not all that hard to plant a false memory in a person's mind -participants in one study were falsely led to "recall" that they had knocked over a punch bowl at a wedding when they were six year old. They did 3 rounds and by the third round, 25% of people were convinced it actually happened. -in a more recent study, 70% of students reported a detailed false memory of having committed a crime, such as assaulting someone with a weapon -memories are particularly easy to implant if person is asked to visualize the event

neurolinguistics

-study of relationship between the brain and language -hemispheric specialization: -left hemisphere typically performs most language processing (95% for right-handed; 50% for left handed) -however, right hemisphere interprets a message's emotional tone, decodes metaphors, and resolves ambiguities

Sleep deprivation in adolescents

-teens now average nearly 2 hours less sleep a night than 80 years ago -insufficient sleep in adolescence increases risk of high blood pressure, heart disease, diabetes, risk taking behavior, depression, and car accidents -compared to A and B students, those who received c,d, and f's in school obtained about 25 min less sleep per night, went to bed an average of 40 minutes later, had more irregular sleep/wake schedules -sacrificing sleep time to study actually worsens academic performance by making it harder the next day to understand class material or do well on a test -regular full night's sleep can dramatically improve athletic ability

eyewitness testimony

-testimony by eyewitnesses to a crime about what they saw during commission of the crime -eyewitness testimony is often mistaken -it is estimated that 2000 to 10,000 people are wrongfully convicted each year in the US on the basis of eyewitness testimony -case of a man who spent 11 years in prison for rape before they found out on the basis of DNA testing that he couldn't have been the assailant -a study examined 62 cases in which innocent people were later exonerated on the basis of DNA evidence (including 8 in which person had been sentenced to death). In 52 of these cases, the crucial evidence leading to conviction had come from eyewitness -*There is strong evidence that eyewitness testimony is both unreliable and manipulable, leading to many wrongful convictions.

massive modularity hypothesis

-the mind does not do any central processing; instead, all information processing is modular -human mind is a collection of Darwinian modules, that is, specialized modules, each of which evolved to solve a specific set of problems encountered by our primitive ancestors -typically, Darwinian modules engage in more complex types of information-processing than Fodorean ones and are not informationally encapsulated (ex: emotion detection, intuitive mechanics or folk physics-innate understanding of basic principles governing the behavior of physical objects, e.g. that objects don't magically appear and disappear, folk psychology, cheater detection)

What makes ACT-R. a hybrid architecture?

-the symbolic, modular architecture is run on a subsymbolic base -ACT-R is designed to operate serially, so that at any given moment, only one production rule can be active, but how does it select that one? Answer: the pattern-matching module controls which production rule gains access to the buffer by working out which production rule has the highest utility at the moment of selection, as determined by 1. how likely the system is to achieve its current goal if the production rule is activated and 2. the cost of activating the production rule -***These calculations are subsymbolic and use an artificial neural networks approach -subsymbolic equations are also used to model how accessible information is in declarative memory -the basic units of declarative memory are chunks, but each chunk is associated with a particular activation level, which is determined by 1. how useful the chunk has been in the past and 2. how relevant the chunk is to the current situation and context -the general information processing that takes place in the buffers is symbolic -in contrast, the calculations that determine whether or not a particular item of knowledge ends up in a buffer are subsymbolic

Piaget's Theory of Cognitive Development

-theory of how humans acquire, construct, and use knowledge -Piaget observed that children of different ages made different kinds of mistakes when solving problems -this led him to believe that children are not just "little adults" who know less; rather, they think and speak differently -proposed that humans progress through four developmental stages" sensori-motor, pre-operational, concrete operations, and formal operations

argument that subsumption architecture robots do not constitute intelligent agents because

-they do not really involve decision-making processes -one response to this challenge is that hybrid architectures that have 1. subsumption architecture for low-level reactive control ("scaled-up insects") WITH 2. traditional central planner for high-level decision making ("scaled-down supercomputers") grafted onto them -Another way of meeting this challenge is behavior-based robots. -unlike subsumption architectures, behavior-based ones represent their environments and use those representations in planning actions. However, unlike symbolic architectures, there is no central planning system. ex: TOTO robot that can identify shortest route between previously visited landmarks. It uses a topological, rather than a metric map, that simply contains information as to whether two landmarks are connected, but not as to how far apart they are, and selects the path that goes via the smallest number of landmarks. Bees operate in similar fashion to identify shortcuts between feeding sites.

use of AI in advertising

-tracking customer behavior to target them with personalized promotions

pretend play

-typically emerges around 14 months -considered a major milestone in cognitive and social development -during pretend play, some of infant's primary representations of the world and other people become "decoupled" from their usual functions while preserving their ordinary meaning -both pretend play and mindreading involve metarepresentation: use of a representation to represent another representation, rather than referring directly to the world -children with autistic spectrum disorder show impoverished pretend play, as well as impairments in mindreading

language processing AI

-use of natural language processing of speech to synthesize notes in professional settings

insanity defense

-using neuroscientific evidence in legal defense -a defendant is not responsible for criminal conduct "if at the time of such conduct as a result of mental disease or defect he lacks substantial capacity either to appreciate the criminality of his conduct or to conform his conduct to the requirements of the law" basically, you can't control yourself or didn't know right from wrong -test thus takes into account both cognitive and volitional capacity -percentage of accused felons that plead insanity is only 1%

Sapir-Whorf Hypothesis

-view that language determines thought -underlying assumption of use of affirmations in cognitive therapy -language and color: natives in New Guinea who have words for two different shades of yellow more speedily perceive and better recall variations between the two yellows. Additionally, those who speak Russian, which has distinct names for various shades of blue, remember the blues better

Cynthia Breazeal's Kismet project

-was a first step in this direction of achieving affective computing 1. designed to model the interaction between an infant and its caregiver 2. cute robotic head capable of sensing others and making wide range of facial expressions 3. processed visual and auditory information to detect affective information 4. information then guided its own emotional state, e.g. expressing frustration, anger, surprise, or sadness. If it does not receive enough stimulation, it expresses sadness. If it is getting too much stimulation, it makes a fearful face. If it received a moderate amount of stimulation, it expresses joy

Errors in eyewitness testimony are especially likely if

-witness' attention was stressed and/or distracted (e.g. by presence of a gun) -plausible misinformation was given during questioning (information presented after can change original memory) -witness is pressured to give a specific response by officers -witness is given positive feedback (could be even a simple "ok") -**confidence of a witness is a very poor predictor of whether a memory is accurate; nevertheless juries are strongly influenced by confidence

social robots

-worldwide sale of robots has reached 20 billion dollars and involves more than 200,000 robotic units -in the immediate future, we can look forward to robotic AIs that can reproduce the skills of therapists, caregivers, guides, and security guards

AI for transportation

1. Kiwibot 2. self driving cars -2018 was the target date proposed in 2015 by Elon Musk and by Google for introduction of self-driving cars, but a series of widely publicized crashes, some fatal, have delayed release Are self driving cars really a viable option? Not rly sure rn. The great successes of deep learning have all been in relatively narrowly defined domains including chess and Go, and even things like image recognition, which primarily involves identifying patterns in a data set, then projecting those patterns onto new exemplars. Some believe that self-driving cars need more than sensitivity to patterns and the ability to learn from experiences -*they need to be able to deal with the unexpected - completely unpredictable behavior from other drivers, pedestrians, cyclist, and even wild animals. -Also, human drivers and constantly exploiting their knowledge of how physical objects move and behave (folk physics), as well as their knowledge of other drivers and road-users (mindreading: understanding peoples intentions by watching them) -*perhaps a key challenge is for designers of self driving cars to find how to equip their vehicles with this kind of general knowledge -pros: less human error = more lives saved, accessible to those who can't drive, can engage in other activities during commute -cons: criminal hacking or system glitches, loss of jobs, high initial cost, fewer people using public transportation -*moral issues: cars will need to make moral decisions in unavoidable accidents such as which person to sacrifice the pedestrian or the driver? 3. autonomous ships

Modular processes are characterized by

1. domain specificity -modules are designed to carry out very specific and circumscribed information processing tasks -can only operate on limited range of inputs 2. informational encapsulation -not affected by what is going on elsewhere in the mind -cannot be "infiltrated" by background knowledge and expectations or knowledge from other modules 3. speed -modular processing operates quickly and efficiently e.g. shape analysis 4. mandatory application -repond automatically -cannot be "switched off" e.g. perception of visual illusion

modular processes are also usually characterized by

1. fixed neural architecture -it is sometimes possible to identify determinate regions of the brain that are associated with particular types of modular processing ex. fusiform face area for face recognition 2. specific breakdown patterns -modules can fail in highly determinate ways, which provide clues on the form and structure of processing ex. prosopagnosia: inability to recognize faces

Stages of Piaget's Theory of Cognitive Development

1. sensorimotor: from birth-2 years -children act on objects (e.g. grasping, sucking, stepping), coordinate sensory experiences (e.g. vision and hearing) from these interactions, and form schemas (internal mental representations) about the objects -they learn to think about aspects of the environment outside of the reach of their senses. ex: object permanence: understanding that object continues to exist even though they cannot see it. Infants do not understand this which is why they respond to peak-a-boo. Once they develop object permanence, they quickly lose interest in the game. Infants also won't reach under cloth for a toy that is hidden because they don't know that it's there. -*According to Piaget, object permanence is one of the most important accomplishments of the sensorimotor age 2. preoperational stage: ages 2-7 -child develops ability to symbolize objects and events that are absent -engages in pretend play -however, child still has trouble seeing things from different points of view: thinking is egocentric. ex: three mountain task: child is unable to describe how the mountains would look to the doll sitting across from her. -shift to new stage often happens quite abruptly, within a matter of a couple weeks -understanding at this stage is based on appearances rather than principles. ex: conservation of liquid quantity. When same amount of milk is poured into different shape glasses, child will think glass that is taller and thinner has more milk than the fatter but shorter cup even though the amount is actually the same 3. Concrete-operational stage: ages 7-11 -child develops higher order schemas called operations - understands the reversible consequences of actions such as the conservation of liquid quantity, conservation of mass quantity, and conservation of number 4. Formal operations stage: over age 11 -child develops ability to engage in hypothetical and deductive reasoning and to think about abstract concepts. ex: child may not that 4 + 1 is odd, that 6+ 1 and 8 + 1 are odd. however, before this stage, they wouldn't understand that if you add one to any even number, the result will be odd -child is asked to discover what makes a pendulum go fast or slow (length of string, weight of object, or initial force that sets pendulum in motion). Before this stage, the child doesn't understand that one needs to vary one factor and hold the others constant

How do we learn language?

1. skill building hypothesis: language is acquired as a result of learning language skills, such as vocabulary and grammar: -use skill building: learn grammar, study vocab lists, do drills, take tests -no pain, no gain -*general public (and government) believe this is the way to learn language 2. comprehension hypothesis: language skills such as vocabulary and grammar, result from language acquisition - we acquire language in one and only one way: when we understand messages: -use "comprehensible input:" listen to stories, read books, have conversations, watch movies -immediate gratification: have a good time - the more you enjoy it, the better your comprehension will be -**comprehensible input has won in pretty much every single study comparing the two methods

Natural language Approach to second language learning

1. storytelling -find language partner who is fluent in language you are trying to learn, e.g. friend, family member, or language exchange partner -find magazines (20%), then children's stories (80%) with tons of pictures -language partner is not going to translate story; it will describe pictures and ask you simple questions, and you will ask simple questions: partner doesnt just describe what is in pictures: they describe the picture the way they might to a young child they love. for ex: "This is a spare tire. A spare tire is very important. You could have a blowout and then you would need to use that spare tire. -rules to tell language partner: -no english. If we don't understand each other, we'll use gestures and act or draw. If we still can't understand, we'll say "it's not important" in target language -no grammar: don't teach me any grammar -no corrections: don't correct me at any time -Factors that affect language acquisition: motivation (positive correlation), self esteem (positive correlation), and anxiety (negative correlation; for language acquisition to really succeed, anxiety should be zero) -Affective filter: somewhere in brain is a language acquisition device, according to Chomsky. Our job is to get input into that device. High anxiety blocks the input. If a student thinks language class is a place where his weaknesses will be revealed, he may understand the input, but it won't penetrate 2. Read!!! whatever is of interest -read things that you like. ex: watch baseball game, then go read about it in newspaper. ***This is also the key to teaching kids who don't want to learn how to reaad!!!!!

Three Stages of Memory Processing and forgetting

Encoding: getting information into brain Storage: retaining that information retrieval: getting the information back out *Forgetting can derive from problems in encoding, storage, or retrieval*

Where in the brain is long-term memory stored?

How does storage work? Karl Lashley's rat experiment in which rats learn the maze found that memories do not reside in single, specific spots. He did this by removing certain parts of a rat's brain and seeing if the speed at which they ran the maze changed for a rat that had a specific part of their brain removed

Why don't they use DNA testing instead of eyewitness?

It's too slow or wasn't available

ACT-R layers

Perceptual-motor layer -perceptual module in turn consists of a visual module, audition module, etc. -motor module consists of speech module, manual module, etc -Communication between modules on different layers takes place via buffers (workspaces) Cognitive Layer -declarative memory is organized in "chunks" -procedural memory is encoded as production rules: actions for the system to perform, e.g. retrieve a chunk from memory, send a command to the motor module to perform an action -production rules can be nested within each other, so that output of a given production rule will trigger firing of another production rule -*All of the above modules are encoded in the form of physical symbols!

Is storage decay simply due to the passage of time or to interference from new memories formed during this interval?

Research suggests that storage decay is primarily due to interference -participants learned lists of nonsense syllables, then either slept or engaged in normal daily activities. Recall was significantly better when participants slept during the retention interval. -The degree to which memories interfere with each other depends on their similarity i.e. it's harder to remember a list of letters if all the letters rhyme (V, G, P, D)

Deja Vu

Who experiences it? About 60%. It's negatively correlated with age. Positively correlated with socioeconomic level, education, stress, and fatigue. More common in people who travel

episodic buffer

another component of working memory that can hold and combine information from the phonological loop, visual sketchpad and long term memory

According to Fodor, not all cognition is

carried out by modular mechanisms -modules provide inputs to nonmodular central processing -the latter can evaluate and correct the outputs of cognitive modules ex: evaluating beliefs and decision-making

lesions of the amygdala may

cause loss of ability to learn fear conditioning and may also play a role in aggressive behavior. -ex: man shot 28 people and murdered 16 at a university and was found to have a midbrain tumor that pressed against his amygdala. There are now neurosurgeries that can remove tumors like that, but -**What are the legal ramifications of that? Should accused be exonerated or given lighter sentence after surgery since he presumably would no longer be violent?**

central executive

component of working memory that integrates information from phonological loop, visuospatial sketchpad, and episodic buffer: similar to attention/sensory memory

risks and dangers of AI

deliberate programming of AI to be hostile e.g. by a terrorist group -sci-fi scenario: if human behavior contradicts one of AI's preprogrammed goals, AI could turn malicious. In 2017, two fb computers started communicating with each other in a language they developed on their own -simulation of government leader's image and voice issuing unauthorized orders, e.g. military action -generation of fake emails, phone calls, video chats -loss of privacy -loss of jobs. Some AI experts predict that AI will replace or eliminate 40% of jobs within 15 years. Greatest impact will be on jobs involving tasks that are repetitive and can be automated -development of computer technology has meant an increase in the amount of time we spend looking at computer screens, and that may be problematic. Recent study published found that after controlling for age, gender, and income, 3-5 year old children with higher use of screen-based media had lower measures of structural integrity and myelination and scored lower on cognitive tests. In older ppl, increased tv watching was found to be correlated w cognitive impairment and poor verbal memory. In fruit flies, blue light exposure causes brain neurodegeneration, as well as shortening of lifespan

cases of organic amnesia have provided evidence for the

distinction between different memory systems (explicit vs. implicit)

Which regions of the brain are thicker in successful agers than in regular agers? Those regulating cognition of emotion?

emotions e.g. the midcingulate cortex and anterior insula

prosopagnosia

failure to recognize particular people by the sight of their faces -after stroke, sheep rancher could not recognize people but could recognize sheep

Encoding failure

information never entered long-term memory (lack of attention) ex: being unable to remember what the face of a penny looks like

storage decay

information stored in long-term memory gradually faced -in general, storage decay is not as severe as most people tend to think. -Study found that people remembered nearly 40% of foreign language vocabulary, idioms, and grammar after 50 years and 75% when recognition tests are used -people who had taken psychology class remembered about 70% of broad general facts and research methods 10 years later

Long-term potentiation

mechanism through which learning occurs in brain -a long term increase in the excitability of a neuron to a particular synaptic input caused by repeated high-frequency activity of that output *Stimulating a particular neural circuit will increase the sensitivity of neurons in that circuit, increasing the probability that they will fire again -process involved binding of glutamate to NMDA receptor -Psychologically, LTP means that both positive and negative thoughts tend to be self-reinforcing

phonological vs. surface dyslexiaHow do we learn

phonological: manifests as severe impairment in reading phonetic script (similar to alphabetic system), but preserved ability in reading pictographic script surface dyslexia: manifests as impairment in reading pictographic script (characters) -direct vs. indirect approaches are reflected in these two different types of dyslexia (learning disability that interferes with reading despite average or above average intelligence)

organic amnesia

physical cause; brain injury through accident or stroke -ex: Korsakoff's amnesia (Wernicke syndrome): amnesia caused by brain damage resulting from thiamine deficiency usually as a result of chronic alcoholism

Psychogenic ("hysterical") amnesia

psychological cause -Dissociative identity disorder

Strengths and Weaknesses of Piaget's Theory

strengths: provides good overview of children's thinking at different points, and has fascinating observations Weaknesses: stage model depicts children's thinking as being more consistent than it actually is. Later research found that children are more cognitively competent than Piaget recognized, understates contribution of the social world, and does not explain underlying mechanisms

three-stage model of memory

the classification of memories based on duration as sensory, working/short-term, and long-term

object permanence

the knowledge that an object exists even when it is not in sight

What type of activities will increase chances of remaining sharp into old age?

those that require hard work and cause you to feel somewhat tired, stymied, frustrated such as grappling with a math problem or pushing yourself to physical limits

nativist view of language

view that language learning is innate

Memory as reconstruction

what we think we remember often never really occurred. we actually filter information and fill in missing pieces -whenever we retrieve a memory, the brain rewrites it a bit. It is slightly altered chemically by a new protein synthesis that links it to our present concerns and understanding (reconsolidation) -because the process involved in memory reconstruction are unconscious, we can be convinced that our memories are accurate even when they are partially or even wholly wrong. -ex: 73 ninth grade boys were interviewed, then re-interviewed 35 years later. when asked to recall how they had reported their attitudes, activities, and experiences, most men performed at a rate no better than chance.

explicit memories

with conscious recall -recall of recognition of information: can be verbally transmitted

poverty of stimulus arguments

young children are simply not exposed to enough information to allow them to learn a language and that much of the speech that children hear is actually ungrammatical, but not flagged as such. Children are typically only exposed to positive information, i.e. they are not told what counts as ungrammatical (e.g. the bell ringed)