Task 1 - Outsourced Intelligence Flashcards
Artificial Intelligence (AI)
Systems that display intelligent behavior by analyzing their environment and taking actions to achieve specific goals.
Technology designed to perform activities that normally require human intelligence.
AI System
an AI-based component, software and/or hardware
–> achieves rationality = the best action to take in order to achieve a certain goal
Rational AI System
They modify the environment but they do not adapt their behavior over time to better achieve their goal.
Learning Rational System
A rational system that, after taking an action, evaluates the new state of the environment to determine how successful its action was, and then adapts its reasoning rules and decision-making methods.
Machine (Supervised) Learning
Programming computers to learn from example data or past experience –> generalization principle
Deep (Unsupervised) Learning
The neural network has several layers between the input and the output that allow to learn the overall input-output relation in successive steps
–> accuracy
–> less human guidance
Reinforcement Learning
We let the AI system free to make its decisions, over time, and at each decision, we provide it with a reward signal that tells it whether it was a good or a bad decision
–> learning from experience
General AI System
Intended to be a system that can perform most activities that humans can de
Narrow AI System
A system that can perform one or few specific tasks
Goal-Directed AI Systems
They receive the specification of a goal to achieve from a human being and use tome techniques to achieve such a goal
Prediction
The ability to take information you have and generate information you didn’t previously have
–> anticipating what will happen in the future
–> major advances in prediction may facilitate the automation of entire tasks
Judgment
The ability to make considered decisions
–> understand the impact different actions will have on outcomes in light of predictions
–> an AI incorporates the feedback on actions and outcomes to develop more accurate predictions and new strategies
–> the role of human judgment will become limited as prediction-driven translation improves
The Challenge of Prediction
- Prediction is not the same as automation
- The most valuable workforce skills involve judgment –> human ethical judgment, emotional intelligence, and creativity is still needed
- Managing will require judgment both in identifying and applying the most useful predictions and in being able to weigh the relative costs of different types of errors
Application of AI (example)
Medicine:
- artificial intelligence can improve diagnosis, lead to more effective treatments, and better patient care
- treatment and care will still rely on human judgment
AI-Enabled Avatars
- life-like virtual human patients for use in clinical training and skill acquisition
- have the potential to be used for all types of person-to-person interactions in mental health care including psychological treatments, assessments, and testing
Advantages of AI-Enabled Avatars
- easy accessibility
- provision of basic assessments, recommendations, and referrals for further treatment
- more privacy
- less stigma
- reduces uncertainty in screening outcomes
- capability to process complex data
Super Clinician
Integrated AI technologies can provide a simulated practitioner with capabilities that are beyond those of human practitioners.
- advanced sensory technologies
- could conduct sessions with complete autonomy or serve as an assistant to practitioners during clinical assessment and treatment
Expert System
A computer program designed to incorporate the knowledge and ability of an expert in a particular domain
Rule-Based Expert Systems
Systems that have facts and rules preprogrammed and therefore require a priori knowledge on the part of the decision-maker
Fuzzy Expert Systems
Expert systems that use fuzzy logic
–> a method of reasoning that deals with approximate values and is useful for working with uncertainties during decision-making
Virtual Reality
A form of human-computer interface that allows the user to become immersed within and interact with a computer-generated simulation environment
–> AI makes artificial companions more life-life, interactive, and capable of doing things that are adaptive to a patient’s need
Implication of AI in Psychological Practice
- computers should not be allowed to make important decisions because computers lack the human qualities of compassion and wisdom
- artificial intelligent agents must be able to process and make value decisions and judgments that involve complex abstract thinking and reasoning
- AI innovations improve and advance psychological practice and research, as well as have the potential to supplant mental health care professionals in core activities that require human intelligence and social interaction
- implanted AI technologies also have the potential to repair or improve general cognitive abilities in humans by making people into cyborgs (partly human and partly machine)
Singularity
The unpredictability of what will happen at that transformative point in human history when machines develop superintelligence
– arises around 2045
The Seven Deadly Sins of AI Prediction
Mistaken predictions lead to fears of things that are not going to happen, whether it’s the wide-scale destruction of jobs, the Singularity, or the advent of AI that has values different from ours and might try to destroy us.
- Overestimating and Underestimating
We tend to overestimate the effect of technology in the short run and underestimate the effect in the long run
- Imagining Magic
Problem with imagined (magical) future technology:
- if it is far enough away from the technology we have and understand today, then we do not know its limitations
- and if it becomes indistinguishable from magic, anything one says about it is no longer falsifiable
- Performance versus Competence
Today’s robots and AI systems are incredibly narrow in what they can do
–> human-style generalizations do not apply
- Suitcase Words
Suitcase Words: words that carry a variety of meanings
–> suitcase words mislead people about how well machines are doing at tasks that people can do
- Exponentials
Moore’s Law: computers get better and better on a clockwork-like schedule (double every year)
–> Moore’s Law and other seemingly exponential laws can fail because they were not truly exponential in the first place
- Hollywood Scenarios
The plot for many Hollywood science fiction movies is that the world is just as it is today, except for one new twist
BUT if we are able to eventually build such smart devices, the world will have changed significantly by then
–> we will change our world along the way, adjusting both the environment for new technologies and the new technologies themselves
- Speed of Development
Almost all innovations in robotics and AI take far, far, longer to be really widely deployed than people in the field and outside the field imagine
Cognitive Science
Cognitive (neuro)science is the scientific study of the biological processes and mechanisms underlying human behavior = the study of the brain.
- a combination of psychology, linguistics, philosophy, and computer science
CRUM
Computational-Representational Understanding of Mind
–> it assumes that the mind has representations of things that have similar ideas to those of a computer
mental representation + computational procedures = thought
The Alignment Problem
How can we make sure that the goals of a strong AI align with our human ideals?
