PHIL 340 Second Half

UM F23

extinction learning

When a conditioned stimulus (CS) no longer predicts the unconditioned stimulus to which it was paired in the past, the CS gradually stops eliciting the conditioned response; while traditional school of thought states that the prefrontal cortex modulates the amygdala to inhibit fear, Pessoa asserts that this behavior is when the conditioned stimulus is safe

cognitive control

If a habitual action is uncalled for, it should be possible to exercise this to recalibrate it, modify it, or call it off completely

cultural technology

the various inventions (language, writing, print, libraries, internet search, wikipedia, LLMs, Foundation models) that have enabled societies to pass down skills, beliefs, and practices from one generation to the next; discussed by Gopnik

direction of fit

"Describing the relationship bewteen beliefs and the world:

- beliefs aim to fit the world (i.e. ideally fit the belief to the world)

- desires aim to fit the world to them (i.e. ideally fit the world to the desire)"

feedback loop

A type of causal structure where the outputs of the function impact future outputs

causal structure

A relationship between inputs (causes) and outputs (effects)

scale

Describing the size of impact

models

opinions embedded in mathematics

WMDs

Algorithms that are high-opacity, large-scale, and large-damage; its not the WMD itself which is unfair or discriminatory rather the use of WMDs which can cause damage.

proxies

values in WMDs that replace variables hard to represent but often times lead to inaccurate results.

glitches

unexpected and often harmful consequences of algorithmic decision-making

fairness

human-prioritized value

calibration

The alignment of a model's design to the intended outcomes

efficiency

model-prioritized value

suitcase words

words that have so many different meanings that people can understand different things by them; Described as 1 of 7 sins in predicting AI by Rodney Brooks because the meaning of such words are improperly extrapolated

exponentialism

The belief that technological progress can exponentially advance society indefinitely; Described as 1 of 7 sins in predicting AI by Rodney Rooks because of physical limit, lack of demand, etc.

AGI

Autonomous agents that operate like beings in the world; Described by Rodney Brooks 1 of 4 misconceptions about AI because society is far from this

the singularity

Idea that eventually an AI-based intelligent entity with goals and purposes will be better at AI research than humans are; Described by Rodney Brooks as 1 of 4 misconceptions about AI because it is a far-away state

misaligned values

Idea that AI-based machines get really good at execution of tasks and are super human at getting things done in a complex world, but don’t share human values, which leads to all sorts of problems; Described by Rodney Brooks as 1 of 4 misconceptions about AI because it is a far-away state

Amara’s law

We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run; 1 of 7 sins by Rodney Brooks

ideology

A set of beliefs; are implicitly reflected in models according to O'Neil

alienation

via Karl Marx, describes how work will become meaningless with the adoption of AI according to Wooldridge

strong AI

machines that are, like us, conscious, self-aware, truly autonomous beings; Wooldridge believes we are far from this technology because we poorly understand mind and consciousness

weak AI

machines that are designed and trained for a specific task

consciousness

a suitcase word; ability to experience things subjectively/from your own personal mental perspective

qualia

A sensation of internal mental phenomena OR the mental sensations that we all experience individually

what it’s like

A question posed by Nagel in determining consciousness; if and only if this answer can be answered for an entity, then it has consciousness

the hard problem of consciousness

the problem of explaining why any physical state is conscious rather than nonconscious

the homunculus problem

The idea of a little man in our heads that sees the images (and words) we see and understands these words and images for us; a recursive solution to the hard problem

the charioteer model

Plato's model of human behavior in which reasoning is pulled by two different horses, one for rational desires, the other for irrational desires. (e.g., Plato’s charioteer model)

the physical stance

Explaining system behavior using the laws of nature (physics, chemistry, etc.)

the design stance

Explaining system behavior from the goals of the system's designer

the intentional stance

Explaining system behavior using the desires of the system

the emotional stance

Explaining system behavior using the emotions of the system

intentional system

Entities whose behavior can be most usefully understood and predicted by the method of attributing to them beliefs, desires, and rational choice

Sally-Anne test

A test given to humans to test for theory of mind

theory of mind

Practical, common-sense capability that fully developed adults have to be able to reason about the mental states (beliefs, desires, etc) of others and themselves

indistinguishability

the ability to discern between to entities; tested by the Turing test; Wooldridge asserts that this value should only be assessed on the behavior of AI and humans (not if there's understanding)

Umwelt

the world as it is experienced by a creature.

Merkwelt

the world that can be sensed by a creature

Wirkwelt

the world that can be interacted with by a creature

blindsight

the ability of people who are cortically blind to respond to visual stimuli that they do not consciously see due to lesions in the primary visual cortex

Cathy O’Neil

argues in WMD that flawed mathematical models which embeds human opinions are increasingly used to evaluate people, as well as promote inequality and other harmful outcomes

Gary Marcus

Discusses the negative impact LLMs have on internet content quality and validity (enshittification)

Thomas Nagel

Discusses qualia; developed test for consciousness be considering whether the question “What is it like to be an X?” makes sense when applied to entities

Daniel Dennett

Developed various stances, or perspectives on explaining system behavior

Golnaz Tabibnia and Dan Radecki

Argue that neuropolasticity and resilience can be improved by developing behavioral pathways (stress management, physical health, social connection) and cognitive pathways (emotion regulation, cogntiive training)

Ruha Benjamin

"Biases and discrimination are ingrained within society as a result of historical and institutional discrimination, and the lack of correction for them results in continued negative effects today

""if we situated racist ‘glitches’ in the larger complex of social meanings and structures, we too could approach them as a signal rather than as a distraction"""

Abeba Birhane

"Argues that machine prediction of social behaviour is not only erroneous but also presents real harm to those at the margins of society

""Certainty and order have always been highly sought after in Western philosophy and sciences. Through the process of elimination of all things that can be doubted, Descartes attempted to get rid of unreliable and fallible human intuitions, senses, and emotions.”"

“. . . extinction is more than a simple form of inhibition. It is a sophisticated form of learning. . . ”

Luiz Pessoa

“The computer learned from the humans how to discriminate, and it carried out this work with breathtaking efficiency.”

Cathy O'Neil

“What we really need is AI that is smart enough to know when it is being used to spew falsehood.”

Gary Marcus

“. . . perhaps if we situated racist "glitches" in the larger complex of social meanings and structures, we too could approach them as a signal rather than as a distraction.”

Ruha Benjamin

“ML systems, tools that fundamentally classify, order, and predict, I argue, are practices that reincarnate Cartesian and Newtonian worldviews that seek stable, predictable, and complete understanding.”

Abeba Birhane