Programming & Artificial Intelligence - Chapter 1: Information Technology and Society

Comprehensive practice flashcards covering IT stages, physical limits, emerging computing directions, the AI hierarchy, machine learning, industrial applications, ethical concerns, and ethical principles from Chapter 1.

What characterized the birth of the computer stage in the $1940s$?

Large, room-sized computers such as ENIAC were used mainly for military and scientific calculations.

What social transformation occurred with the introduction of personal computers (PCs) in the $1970s$?

Individuals gained the ability to use computers for personal tasks at home and work.

What key change occurred during the commercialization of the Internet in the $1990s$?

Information became globally accessible through e-mail and the web spread into ordinary services, communication, and business.

How did smartphones transform information access in the $2000s$?

Smartphones made Internet services portable and accessible almost anywhere away from a desk.

What shift took place with the arrival of cloud computing in the $2010s$?

Storage and computing power transitioned into online services accessed remotely through the Internet.

What is Moore's Law, and what does it actually describe?

Moore's Law is the observed historical trend that the number of transistors on an integrated circuit roughly doubles about every $2$ years; it describes transistor density growth rather than a guarantee of doubled computer speed.

What is quantum tunneling in integrated circuits?

Quantum tunneling is the effect where electrons pass through extremely thin barriers (only a few atoms thick) that are meant to block them, presenting a physical limit to shrinking transistors.

What is leakage current and why is it problematic in computer chips?

Leakage current is electrical current that escapes unintentionally through tiny transistor barriers, wasting electrical power and generating heat.

What is parallel processing and how does it improve computing performance?

Parallel processing is the simultaneous use of multiple processor cores to handle work without relying solely on shrinking individual transistors.

How does a qubit differ from a classical bit?

A classical bit holds a single value of either $0$ or $1$, whereas a qubit uses quantum superposition and is not simply in a classical state of $0$ or $1$.

What is quantum computing?

Quantum computing is computing that utilizes the principles of quantum mechanics to speed up specific complex calculations that are difficult for classical computers.

How do Cloud Computing and Edge Computing differ regarding processing location?

Cloud Computing processes data on remote servers reached via the Internet, while Edge Computing processes data directly on the local device close to where data is generated.

Why is Edge Computing preferred over Cloud Computing for autonomous driving braking decisions?

Edge Computing processes decisions onboard immediately, avoiding network delays that could cause dangerous outcomes in split-second braking decisions.

What is the main distinction between Augmented Reality (AR) and Virtual Reality (VR)?

AR overlays digital content onto real-world views while keeping the real world visible, whereas VR completely replaces the real view with a computer-generated virtual environment.

What are the four nested levels of the AI family hierarchy in order from broadest to most specialized?

Artificial Intelligence (AI) contains Machine Learning (ML), Machine Learning contains Deep Learning (DL), and Deep Learning contains Generative AI.

How does Narrow AI differ from general human intelligence?

Narrow AI is specialized in performing one particular task (such as speech recognition, recommendation, or translation), whereas general human intelligence encompasses a broad range of intellectual capabilities.

What is the difference between a traditional fixed-rule program and a Machine Learning system?

In a fixed-rule program, human developers write explicit rules in advance; in Machine Learning, the system processes training data to learn patterns and build a model used to evaluate new data.

What is a neural network in Deep Learning?

A neural network is a structure used in Deep Learning to learn complex patterns from large-scale data, progressing from simpler features like pixels and edges to complex representations.

What is an AI hallucination in Generative AI?

An AI hallucination is a factually incorrect output produced by Generative AI that sounds plausible, fluent, and confident.

Why is human verification necessary when using output from a Generative AI chatbot?

Because chatbots generate responses using learned language patterns and can produce plausible-sounding hallucinations, requiring important facts to be verified against reliable sources.

How does a Recommendation System operate on platforms like YouTube?

It analyzes a user's past behavior data to predict preferences and display personalized content recommendations.

How does Predictive Maintenance differ from repairing a machine after failure?

Predictive Maintenance uses operational data (such as temperature and vibration) to estimate and prevent a likely breakdown before it happens, whereas repair after failure responds only after a machine breaks.

What is the Black-box Problem in AI applications?

The Black-box Problem is a situation in which an AI system reaches a judgment or prediction, but the underlying decision-making reasoning is unclear to humans.

Why is human confirmation required for high-risk decisions like medical image diagnosis?

Because AI provides probabilistic reasoning and pattern recognition, but final diagnosis, clinical evaluation, and responsibility require expert human review.

What is Algorithmic Bias and what primary factors cause it?

Algorithmic Bias is unfair discrimination in AI judgments, primarily caused by biased training data, unsuitable model design, or improper proxy variables.

What is a proxy variable in the context of AI bias?

A proxy variable is a variable in a model that indirectly stands in for another attribute, which can inadvertently introduce bias regarding a protected characteristic.

What ethical concerns are raised by surveillance through face recognition and mass collection of personal data?

They raise severe Privacy concerns regarding the tracking, collection, and analysis of personal data belonging to ordinary individuals.

What is Explainable AI (XAI) and which ethical principle does it directly support?

Explainable AI (XAI) is technology that enables humans to understand why AI made a particular judgment, directly supporting the ethical principle of Transparency.

What is the distinction between Responsibility and Accountability in AI ethics?

Responsibility addresses the unresolved question of who answers for an AI error among developer, user, and operator, whereas Accountability is the ethical principle that someone must be answerable for AI decisions.

What are the four core AI ethics principles and which ethical concern does each correspond to?

Fairness responds to Algorithmic Bias, Transparency responds to the Black-box Problem, Privacy Protection responds to Privacy issues, and Accountability responds to Responsibility questions.