AI Governance: Key Terms

Choose a study mode

Play Quiz
Study Flashcards
Spaced Repetition
Chat to Lesson

Podcast

Play an AI-generated podcast conversation about this lesson
Download our mobile app to listen on the go
Get App

Questions and Answers

What is the primary objective of 'Accountability' in the context of AI systems?

  • To ensure AI systems operate in a manner that is ethical, fair, and transparent. (correct)
  • To eliminate the need for human oversight in AI operations.
  • To maximize the speed and efficiency of AI decision-making processes.
  • To reduce the computational resources required for AI system deployment.

Why is 'Accuracy' considered a critical metric in AI, particularly in applications like medical diagnoses?

  • Because it ensures the AI system can process data quickly, regardless of the correctness of the output.
  • Because it directly reflects the reliability and effectiveness of the AI system in producing correct outputs. (correct)
  • Because it reduces the amount of training data required to develop a functional AI model.
  • Because it simplifies the AI model, making it easier to deploy across different platforms.

In 'Active Learning', how does an algorithm improve its learning process?

  • By requesting additional data points to enhance its understanding. (correct)
  • By using pre-existing knowledge to fill in gaps in the data.
  • By passively learning from all the data it is initially given.
  • By ignoring irrelevant data to focus on core information.

What is the primary goal of 'Adaptive Learning' in educational contexts?

<p>To provide a personalized and optimized learning experience tailored to individual student needs and abilities. (B)</p> Signup and view all the answers

How do 'Adversarial Attacks' pose a risk to AI models?

<p>By manipulating the model with malicious or deceptive data, causing it to malfunction. (A)</p> Signup and view all the answers

What is the main purpose of 'AI Assurance'?

<p>To measure, evaluate, and promote safe, reliable, and trustworthy AI through various controls and assessments. (B)</p> Signup and view all the answers

What does an 'AI Audit' primarily assess?

<p>Whether an AI system operates as intended and complies with relevant laws and regulations. (A)</p> Signup and view all the answers

What is the role of 'AI Governance' in the AI landscape?

<p>To implement, manage, and regulate the development, deployment, and use of AI technology at various levels. (B)</p> Signup and view all the answers

Which of the following best describes an 'Algorithm' in the context of AI?

<p>A procedure or set of instructions designed to perform a specific task or solve a particular problem using a computer. (B)</p> Signup and view all the answers

What characterizes 'Artificial General Intelligence (AGI)'?

<p>AI with human-level intelligence and strong generalization capabilities across various tasks and contexts. (D)</p> Signup and view all the answers

How does 'Artificial Intelligence (AI)' automate tasks?

<p>By using various computational techniques like machine learning to perform or automate tasks. (D)</p> Signup and view all the answers

What is the key feature of 'Automated Decision-Making'?

<p>Decisions are made by technological means without human involvement, either entirely or partially. (C)</p> Signup and view all the answers

What is the primary source of bias in AI according to the text?

<p>A systematic error or deviation in predictions that originates from the model's assumptions or the data itself. (C)</p> Signup and view all the answers

What is the main goal of the 'Bootstrap Aggregating' method in machine learning?

<p>To aggregate multiple versions of a model trained on random data subsets to make the model more stable and accurate. (A)</p> Signup and view all the answers

How do 'Chatbots' simulate human-like conversations?

<p>By using natural language processing and deep learning to understand and respond to text or speech. (B)</p> Signup and view all the answers

What is the purpose of a 'Classification Model' in machine learning?

<p>To sort input data into different categories or classes. (A)</p> Signup and view all the answers

How does 'Clustering' work in unsupervised machine learning?

<p>By grouping data points into clusters based on their similarity without prior knowledge of categories. (A)</p> Signup and view all the answers

What role does 'Compute' play in AI systems?

<p>It refers to the hardware components necessary for processing data and running applications. (A)</p> Signup and view all the answers

What is the primary function of 'Computer Vision' in AI?

<p>To enable computers to process and analyze visual inputs like images and videos. (B)</p> Signup and view all the answers

What does 'Conformity Assessment' evaluate in the context of AI systems?

<p>Whether an AI system meets specific requirements like risk management, data governance, and cybersecurity. (B)</p> Signup and view all the answers

What does 'Contestability' ensure regarding AI systems and their decision-making processes?

<p>That these processes can be questioned or challenged by humans. (A)</p> Signup and view all the answers

What is the main purpose of a 'Corpus' in AI and machine learning?

<p>To provide a large collection of texts or data for computers to find patterns, make predictions, or generate specific outcomes. (D)</p> Signup and view all the answers

What differentiates a 'Data Leak' from a 'Data Breach'?

<p>A data leak is unintentional exposure of sensitive data, while a data breach is a deliberate, malicious attack. (B)</p> Signup and view all the answers

What is the goal of 'Data Poisoning' in the context of AI?

<p>To manipulate the training process by injecting false data, corrupting the learning algorithm. (D)</p> Signup and view all the answers

Why is 'Data Provenance' essential for AI governance?

<p>It tracks the history and origin of data, ensuring its integrity and quality. (D)</p> Signup and view all the answers

How does 'Data Quality' impact AI outputs?

<p>It directly impacts the quality of outputs and the AI system's performance with high-quality data ensuring better outcomes. (D)</p> Signup and view all the answers

What does a 'Decision Tree' represent in machine learning?

<p>Decisions and their potential consequences in a branching structure. (C)</p> Signup and view all the answers

How does 'Deep Learning' utilize artificial neural networks?

<p>By using multiple layers to model complex nonlinear relationships within data. (D)</p> Signup and view all the answers

What potential harm can be caused by 'Deepfakes'?

<p>Spread of misinformation and disinformation. (D)</p> Signup and view all the answers

Which of the following describes a core function of 'Diffusion Models'?

<p>Iteratively refining a noise signal into a realistic image. (B)</p> Signup and view all the answers

Which of the following tasks would use a 'Discriminative Model'?

<p>Text classification. (D)</p> Signup and view all the answers

What is the key characteristic of 'Disinformation'?

<p>Its manipulative intention to cause harm. (D)</p> Signup and view all the answers

Which concept does 'Entropy' measure in machine learning?

<p>Degree of unpredictability. (A)</p> Signup and view all the answers

An 'Expert System' helps with decision-making in which field?

<p>Medical diagnosis. (D)</p> Signup and view all the answers

What quality is highlighted by 'Explainability'?

<p>Transparency. (A)</p> Signup and view all the answers

In what way does 'Exploratory Data Analysis' enhance model training?

<p>Identifying relationships among variables. (A)</p> Signup and view all the answers

What is the primary goal of 'Fairness' within AI systems?

<p>To prioritize relatively equal treatment regardless of sensitive attributes. (D)</p> Signup and view all the answers

Which data aspect is NOT transferred in 'Federated Learning'?

<p>Training data itself. (D)</p> Signup and view all the answers

An AI system is designed to evaluate loan applications, but consistently denies applicants from a specific ethnic background. Which term from AI governance best describes this scenario?

<p>Fairness (B)</p> Signup and view all the answers

A self-driving car misinterprets a stop sign due to heavy rain and glare. This is an example of a failure in what aspect of AI systems?

<p>Robustness (C)</p> Signup and view all the answers

An AI model used for predicting stock prices begins to make increasingly erratic predictions after a sudden shift in market conditions. This is a failure in what aspect of AI performance?

<p>Reliability (C)</p> Signup and view all the answers

A hospital implements an AI system for diagnosing diseases, but does not inform the medical staff about the AI's limitations or how it arrives at its decisions. This is a violation of which principle?

<p>Transparency (C)</p> Signup and view all the answers

An AI-powered recruitment tool is found to favor candidates who attended a specific group of universities, even though candidates from other institutions are equally qualified. What concept does this scenario exemplify?

<p>Bias (B)</p> Signup and view all the answers

A financial institution uses an AI to detect fraudulent transactions. However, the AI struggles to accurately identify fraud in transactions made by non-English speakers due to a lack of training data in other languages. This is primarily an issue related to?

<p>Data Quality (A)</p> Signup and view all the answers

A company develops an AI system for medical diagnoses but fails to keep detailed records of the data used, the model's development process, or any changes made to the algorithm. This poses a significant problem for?

<p>Accountability (D)</p> Signup and view all the answers

An AI model is trained to predict customer churn, but during deployment, it performs poorly on new customer data that differs significantly from the training data. This is an example of?

<p>Generalization Failure (B)</p> Signup and view all the answers

A research team uses an AI model to generate realistic images of human faces, but some of these images are then used to create fake social media profiles for malicious purposes. This represents a failure in?

<p>Safety (B)</p> Signup and view all the answers

A company uses an AI to automate customer service interactions. A customer asks for a feature that is not yet implemented, and the AI fabricates a response, promising the feature will be available soon, even though there are no plans to develop it. What term describes this scenario?

<p>Hallucination (A)</p> Signup and view all the answers

If a malicious actor gains access to a model's parameters, corrupts the training dataset with fabricated data, and compromises the expected AI performance, this attack exemplifies?

<p>Data Poisoning (C)</p> Signup and view all the answers

What process ensures AI systems and their decision-making are open to scrutiny and challenge, promoting trust and responsibility?

<p>Contestability (A)</p> Signup and view all the answers

What type of AI model combines image and textual analysis to produce a descriptive score, showcasing versatility in handling varied data types?

<p>Multimodal Model (B)</p> Signup and view all the answers

Which AI method prioritizes direct human involvement and oversight throughout operation to ensure alignment with human intentions?

<p>Human-in-the-Loop (B)</p> Signup and view all the answers

Which term describes the detailed analysis and recording of a dataset's origins, ensuring thorough understanding of how data is managed and transformed?

<p>Data Provenance (C)</p> Signup and view all the answers

During AI model training, what term identifies the process of adjusting a pre-existing model to make it highly specialized for a specific function?

<p>Fine-tuning (D)</p> Signup and view all the answers

What emerging AI area focuses on creating diverse content through deep learning, distinguishing itself by generating new data based on prompts?

<p>Generative AI (C)</p> Signup and view all the answers

If an AI model's performance is strong with training data but declines significantly when exposed to new data, what issue does this indicate?

<p>Overfitting (B)</p> Signup and view all the answers

Which process assesses an AI system's risks and societal effects in specific scenarios, ensuring ethical and legal standards are considered?

<p>Impact Assessment (D)</p> Signup and view all the answers

What proactive measure embeds visually imperceptible patterns in AI-generated content to aid in identifying and labeling AI creations?

<p>Watermarking (D)</p> Signup and view all the answers

Which term describes the initial phase of data examination to uncover patterns, anomalies, and variable relationships that informs model training?

<p>Exploratory Data Analysis (C)</p> Signup and view all the answers

What kind of AI system replicates expert decision-making within a specific field by using a knowledge base from human specialists?

<p>Expert System (C)</p> Signup and view all the answers

Which subfield of AI helps computers interpret and apply human language for tasks like translation and sentiment analysis?

<p>Natural Language Processing (B)</p> Signup and view all the answers

What AI training approach involves a model learning through trial and error using rewards and penalties within a simulated environment?

<p>Reinforcement Learning (A)</p> Signup and view all the answers

Which machine learning technique enhances model stability by averaging predictions from multiple models trained on different subsets of the data?

<p>Bootstrap Aggregating (C)</p> Signup and view all the answers

What is the primary goal of 'red teaming' an AI system?

<p>To identify vulnerabilities and potential harms through adversarial testing. (B)</p> Signup and view all the answers

Which of the following best describes the purpose of 'validation data' in the context of machine learning?

<p>To fine-tune the parameters of a model and prevent overfitting during training. (D)</p> Signup and view all the answers

What is the primary function of 'transfer learning' in machine learning?

<p>To use knowledge learned from one task as a basis for learning a different but related task. (C)</p> Signup and view all the answers

In the context of AI, what does 'robustness' primarily refer to?

<p>The ability of a system to be resilient to attacks and maintain functionality under various conditions. (C)</p> Signup and view all the answers

Which of the following best describes the 'Turing test'?

<p>A test of a machine's ability to exhibit intelligent behavior equivalent to a human. (B)</p> Signup and view all the answers

What is the main goal of 'prompt engineering'?

<p>To structure prompts to influence model behavior and generate more desirable outputs. (C)</p> Signup and view all the answers

Which of the following is a key characteristic of 'small language models'?

<p>They are optimized for efficiency and deployment in resource-limited environments. (A)</p> Signup and view all the answers

In machine learning, what does 'variance' refer to?

<p>The measure of how far a set of numbers are spread out from their average value in a dataset. (C)</p> Signup and view all the answers

Which of the following is a core feature of 'open-source software' in AI development?

<p>It provides free and open access to source code for modification and redistribution. (B)</p> Signup and view all the answers

What is the primary role of 'input data' in machine learning?

<p>To form the basis for machine learning models to learn, make predictions, and carry out tasks. (C)</p> Signup and view all the answers

When we talk about 'interpretability' in AI, what are we referring to?

<p>The ability to explain or present a model's reasoning, in human-understandable terms. (A)</p> Signup and view all the answers

Which of the following options explains about AI's 'compute'?

<p>The processing resources that are available to a computer system. (C)</p> Signup and view all the answers

Flashcards

Accountability

Obligations of AI developers/deployers to ensure ethical, fair, transparent, and compliant system operation.

Accuracy

How correctly an AI system performs its intended task.

Active learning

AI algorithm selects data to learn from, requesting more to improve learning.

Adaptive learning

Method adjusting content to individual student needs for optimized learning.

Signup and view all the flashcards

Adversarial attack

Safety/security risk manipulating the AI model with malicious or deceptive input data.

Signup and view all the flashcards

AI assurance

Frameworks/policies measuring/promoting safe, reliable, trustworthy AI.

Signup and view all the flashcards

AI audit

Review/assessment ensuring an AI system works as intended and complies with regulations.

Signup and view all the flashcards

AI governance

System of laws/policies regulating AI development/deployment, ensuring responsible use.

Signup and view all the flashcards

Algorithm

Instructions designed to perform a task or solve a problem.

Signup and view all the flashcards

Artificial general intelligence

AI with human-level intelligence and strong generalization ability.

Signup and view all the flashcards

Artificial intelligence

Engineered system using techniques to automate tasks.

Signup and view all the flashcards

Automated decision-making

Decision making by technological means without human input.

Signup and view all the flashcards

Bias

Systematic error from the model's assumptions or data.

Signup and view all the flashcards

Bootstrap aggregating

Aggregates multiple model versions trained on dataset subsets for increased stability.

Signup and view all the flashcards

Chatbot

AI designed to simulate human-like conversation via text or speech.

Signup and view all the flashcards

Classification model

Model to sort input data into categories or classes.

Signup and view all the flashcards

Clustering

Unsupervised ML method grouping data points into clusters based on similarity.

Signup and view all the flashcards

Compute

Processing resources available to a computer.

Signup and view all the flashcards

Computer vision

AI field processing and analyzing images/videos.

Signup and view all the flashcards

Conformity assessment

Analysis determining if AI system meets requirements.

Signup and view all the flashcards

Contestability

Principle ensuring AI systems can be questioned by humans.

Signup and view all the flashcards

Corpus

Large collection of texts/data used to find patterns.

Signup and view all the flashcards

Data leak

Accidental exposure of sensitive data.

Signup and view all the flashcards

Data poisoning

Adversarial attack injecting false data to corrupt the learning algorithm.

Signup and view all the flashcards

Data provenance

Process tracking records' history in dataset.

Signup and view all the flashcards

Data quality

How well a dataset meets requirements.

Signup and view all the flashcards

Decision tree

Supervised learning model decisions/consequences in branching structure.

Signup and view all the flashcards

Deep learning

AI subfield using artificial neural networks.

Signup and view all the flashcards

Deepfakes

Audio/visual content altered using artificial intelligence.

Signup and view all the flashcards

Diffusion model

Generative model refining noise signal into a realistic image.

Signup and view all the flashcards

Discriminative model

Model maps input features to class labels.

Signup and view all the flashcards

Disinformation

Audio/visual intentionally manipulated to cause harm.

Signup and view all the flashcards

Entropy

Measure of unpredictability/randomness in machine learning data.

Signup and view all the flashcards

Expert system

AI that replicates human expertise within a field.

Signup and view all the flashcards

Explainability

Describing how an AI system generates outputs.

Signup and view all the flashcards

Exploratory data analysis

Process to gain dataset insights before modeling.

Signup and view all the flashcards

Federated learning

Models trained on multiple edge devices.

Signup and view all the flashcards

Fine-tuning

Refining a pre-trained model on a smaller labeled dataset.

Signup and view all the flashcards

Foundation model

Large-scale model with broad AI capabilities.

Signup and view all the flashcards

Generalization

The ability of a model to apply learned patterns to unseen data.

Signup and view all the flashcards

Generative AI

AI that creates content like text, images, or music.

Signup and view all the flashcards

Greedy algorithms

Algorithm making optimal choice for immediate objective.

Signup and view all the flashcards

Ground truth

Objective truth of a dataset used for evaluation.

Signup and view all the flashcards

Hallucinations

A plausible but factually incorrect output.

Signup and view all the flashcards

Human-centric AI

AI that prioritizes human well-being and values.

Signup and view all the flashcards

Human-in-the-loop

Human oversight in AI decision-making.

Signup and view all the flashcards

Impact assessment

Evaluation of ethical, legal, and societal AI implications.

Signup and view all the flashcards

Inference

Using a trained model to make predictions.

Signup and view all the flashcards

Input data

Data powering a learning algorithm.

Signup and view all the flashcards

Interpretability

Explaining a model's reasoning in understandable terms.

Signup and view all the flashcards

Large language model

AI that analyzes and learns patterns from massive datasets.

Signup and view all the flashcards

Machine learning

Algorithms that learn from data without explicit programming.

Signup and view all the flashcards

Machine learning model

Representation of data patterns to do tasks.

Signup and view all the flashcards

Model card

Document that shows info about an AI model.

Signup and view all the flashcards

Multimodal models

Models that process more than one input.

Signup and view all the flashcards

Natural language processing

AI helps computers understand human language.

Signup and view all the flashcards

Neural networks

Models mimicking human brain processing.

Signup and view all the flashcards

Open-source software

The public can access the source code.

Signup and view all the flashcards

Overfitting

Model too specific to data.

Signup and view all the flashcards

Oversight

Effectively monitor the AI system.

Signup and view all the flashcards

Parameters

Values learned through training.

Signup and view all the flashcards

Post processing

Adjust model output after running.

Signup and view all the flashcards

Preprocessing

Steps for preparing data.

Signup and view all the flashcards

Prompt

Instruction for an AI model.

Signup and view all the flashcards

Prompt engineering

Structuring prompts to get desired outputs.

Signup and view all the flashcards

Random forest

Algorithm merging multiple decision trees.

Signup and view all the flashcards

Red teaming

Adversarial testing to find safety issues.

Signup and view all the flashcards

Reinforcement learning

Model optimize through rewards and penalties.

Signup and view all the flashcards

Reinforcement learning with human feedback

Reinforcement plus human preferences.

Signup and view all the flashcards

Study Notes

AI Governance Key Terms

  • Accountability: Ensures AI systems operate ethically, fairly, transparently, and compliantly, tracing actions back to the responsible entity; includes obligations and responsibilities of AI developers and deployers to ensure ethical operation, fairness, transparency, and compliance with regulations.
  • Accuracy: Measures how correctly an AI system performs its task; critical for AI model reliability, especially in high-precision applications like medical diagnoses; measures system performance and effectiveness in producing correct outputs from input data.
  • Active Learning: A machine learning subfield where algorithms selectively learn from data by requesting specific data points to improve learning.
  • Adaptive Learning: A method that customizes educational content to meet individual student needs and learning styles; provides personalized and optimized learning catering to diverse learning styles.
  • Adversarial Attack: Exploits AI model vulnerabilities via malicious input data, causing malfunctions and unsafe outputs; manipulating inputs can cause malfunctions, like fooling a self-driving car into perceiving a stop light as green, impacting safety.
  • AI Assurance: Frameworks and controls ensuring safe, reliable, and trustworthy AI; includes conformity, impact, and risk assessments, AI audits, certifications, testing, evaluation, and compliance w/ standards.
  • AI Audit: Review of AI systems ensuring intended operation and legal compliance, identifying risks, and offering mitigation; review and assessment of an AI system ensures it operates as intended, complies with laws, regulations, standards, helps identify and map risks and offer mitigation strategies
  • AI Governance: Laws, policies, and processes that manage AI tech development ethically and responsibly at international, national, and organizational levels; helps stakeholders implement, manage, oversee, regulate AI technology, manages associated risks to ensure alignment with objectives, ethical development, legal and regulatory compliance
  • Algorithm: A procedure or set of rules designed to perform a specific task or solve a computer problem.
  • Artificial General Intelligence (AGI): AI with human-level intelligence and broad generalization capability; contrasted with "narrow" AI
  • Artificial Intelligence (AI): Engineered systems that use computation to perform tasks, simulate intelligence, and automate; a field of computer science that simulates intelligent behavior, may include automated decision making.
  • Automated Decision-Making: Making choices via AI without human intervention; a process of making decisions by technological means either wholly or partly.
  • Bias: Systematic errors in AI predictions due to model assumptions or flawed data; computational or machine bias is a systematic error from a model's assumptions or the data itself, can impact outcomes and risk individual rights and liberties; cognitive bias refers to inaccurate judgement or distorted thinking, societal bias leads to systemic prejudice, favoritism, discrimination (selection bias ie biases in selecting data for model training)
  • Bootstrap Aggregating: A machine learning method improving model stability and accuracy by averaging multiple model versions trained on random data subsets; also known as bagging
  • Chatbot: AI simulating human conversations through natural language processing and deep learning; designed to simulate human-like conversations and interactions, uses natural language processing, deep learning to understand and respond to text or speech.
  • Classification Model: Categorizes input data in machine learning; sometimes referred to as classifiers
  • Clustering: An unsupervised machine learning to group similar data points; sometimes referred to as clustering algorithms
  • Compute: Processing resources in a computer system, essential for data handling, applications, and visual rendering; includes hardware components such as the central processing unit or graphics processing unit that is essential for memory, storage, processing data, running applications, rendering graphics etc.
  • Computer Vision: AI field for processing images/videos, enabling applications like facial/object recognition; uses computers to process and analyze images, videos, and other visual inputs
  • Conformity Assessment: Analysis to determine if an AI system meets requirements for risk management, data governance, and cybersecurity; often performed by an entity independent of a model developer in order to establish a risk management system, data governance, record-keeping, transparency, and cybersecurity practices
  • Contestability: Ensuring AI systems' decisions are questioned or challenged, promoting accountability; helps promote accountability in AI governance, outputs and actions depend on transparency, ability to contest or challenge the outcomes (principle of ensuring AI systems, their decision-making processes can be questioned/challenged by humans) may be called redress
  • Corpus: A data collection used by computers to find patterns/make predictions; may include structured or unstructured data and cover a specific topic or a variety of topics
  • Data Leak: Unintentional exposure of sensitive data due to security lapses or human error; can be from poor security defenses, human error, storage misconfigurations or a lack of robust policies around internal and external data sharing practices; a data breach is intentional and done in bad faith
  • Data Poisoning: Adversarial attacks that inject false data to corrupt the training process; involves malicious users injecting false data into a model, leading to compromised performance, unintentional, misleading, or harmful outputs
  • Data Provenance: Tracking data history to ensure integrity, transparency, and governance; tracks and logs the history and origin of records in a dataset, encompassing the entire life cycle from its creation/collection to its transformation to its current state, essential for data transparency and governance and promote a better understanding for the data.
  • Data Quality: How well a dataset meets requirements for its intended use and is accurate, complete, valid, consistent and timely; high quality is accurate, complete, valid, consistent, timely, fit for purpose, directly impacts the quality of AI outputs and the system performance.
  • Decision Tree: A machine learning model showing branching decisions and their potential outcomes; supervised learning that represents decisions and their potential consequences in a branching structure
  • Deep Learning: AI subfield using neural networks, excels in processing raw data like images/speech; a subfield of AI and machine learning that uses artificial neural networks
  • Deepfakes: AI-altered audio or video for misinformation.
  • Diffusion Model: Generates images by refining a noise signal.
  • Discriminative Model: Directly maps input features to class labels in machine learning; used in machine learning that directly maps input features to class labels and analyzes for patterns, is often used for text classification tasks, like identifying the language of a piece of text or detecting spam; examples are traditional neural networks and random forests
  • Disinformation: Manipulated content intended for harm; can be spread through deepfakes who have malicious intentions
  • Entropy: Measures unpredictability in machine learning datasets; a higher entropy signifies greater uncertainty in predicting outcomes
  • Expert System: Emulates human decision-making in fields like medical diagnoses using a rule-based AI; a form of rules based AI that draws inferences from a knowledge base provided by human experts to replicate decision-making abilities within a specific field such as medical diagnoses
  • Explainability: Providing sufficient information about the how an AI system works, to maintain transparency and trust; important in maintaining transparency and trust in AI
  • Exploratory Data Analysis: Techniques to gain insights into a dataset, uncover relationships, and identify anomalies before training a machine learning model.
  • Fairness: An AI quality that prioritizes equal treatment in decisions, avoiding adverse impacts on sensitive attributes like race or gender; system decisions should not adversely impact, whether directly, disparately on sensitive attributes
  • Federated Learning: Training models on decentralized data while preserving privacy; allows models to be trained on the local data of multiple edge devices - only the updates of the local model are sent to a central location to improve the global model
  • Fine-tuning: Specialized training of pre-trained learning models to make them perform a specialized task; a process taking a pretrained deep learning model and training it further for a specialized task through supervised learning, involves taking a foundation model that has already learned general patterns to train it for a specific task using a much smaller and labelled dataset
  • Foundation Model: Large-scale model trained on diverse datasets to be a base for specific applications; a large-scale model that has been trained on extensive and diverse datasets to enable broad capabilities, can function as the base for use-specific applications, also called general purpose AI model and frontier AI.
  • Generalization: The ability of a model to apply learned patterns to new, unseen data.
  • Generative AI: AI for creating content from large datasets, responding to prompts; uses deep learning trained on large datasets to create content such as data, images, code, music, simulations, videos in response to user prompts, unlike discriminative models it makes predictions existing data vs new data
  • Greedy Algorithms: Algorithms that make immediate, optimal choices without considering long-term solutions.
  • Ground Truth: True state against which AI is evaluated; serves as the real-world reference against which the outputs are measured for accuracy and reliability.
  • Hallucinations: Factually incorrect information created by AI; also called confabulations
  • Human-Centric AI: Prioritizes human well-being and autonomy; an approach to AI design, development, deployment and use in order to amplify and augment human abilities vs undermining them
  • Human-in-the-Loop (HITL): AI design including human oversight; a design paradigm that incorporates human oversight, intervention, interaction or control over the operation and decision-making processes of an AI system
  • Impact Assessment: Process to identify the ethical, legal and societal impact of AI; an evaluation process designed to document and mitigate the potential implications of an AI system for a specific use case.
  • Inference: Model-based predictions from input data; a type of machine learning process used to make predictions or decisions based on input data.
  • Input Data: Data used by a learning algorithm to produce an output; the basis for machine learning models to learn, make predictions and carry out tasks.
  • Interpretability: Ability to explain a model's reasoning in human terms; involves designing models that inherently facilitate understanding through their structure, emphasizes designing models that facilitate understanding through their structure, features or algorithms, requires significant domain expertise
  • Large Language Model (LLM): Utilizes deep learning to analyze patterns in text-based tasks; a form of AI that utilizes deep learning algorithms to create models (see also machine learning model, foundation model and fine-tuning) pretrained on massive text datasets, performs text-based tasks; generally, two types: generative and discriminative models
  • Machine Learning (ML): Algorithms that learn and improve from data to make decisions; implements various algorithms that learn and improve by experience, includes data collection and preparation, feature engineering, training, testing and validation
  • Machine Learning Model: Representation of data patterns used for future predictions; a learned representation of underlying patterns and relationships in data, created via an AI algorithm to a training dataset, can be used to make predictions or perform tasks on new, unseen data
  • Misinformation: False content spread without harmful intent; can be spread unintentionally through deepfakes by those who lack harm
  • Model Card: A document with information about the purpose and performance of an AI model; discloses information about an AI model like explanations about intended use, performance metrics benchmarked in various conditions
  • Multimodal Models: Processing multiple data types at once; can process more than one type of output, has various uses like image captioning and speech recognition
  • Natural Language Processing (NLP): Helps computers understand human language; helps computers to understand, interpret, and apply human language by transforming into content and enabling machines to translate language, read text or spoken language etc
  • Neural Networks: Models mimicking brain structure for complex tasks; the layered approach enables machine-based neural networks to model complex nonlinear relationships and patterns within data
  • Open-Source Software: Promotes transparency and collaboration; a decentralized development model that provides free and open access to source code, promotes shared collaboration and learning
  • Overfitting: Model that is too specific to training data and fails on new data; unable to generalize to unseen data
  • Oversight: Supervising AI for regulatory compliance and risk management; important for effective AI governance
  • Parameters: Internal variables adjusted during training; the model adjusts to, during the training process so it can make predictions on new data.
  • Post-Processing: Adjusting model outputs.
  • Preprocessing: Preparing data for machine learning; can include cleaning, handling missing values, normalizing, extracting features, etc
  • Prompt: An input or instruction provided to an Al model or system to generate an output.
  • Prompt Engineering: Structuring prompts to optimize desired outputs; a deliberate process of structuring prompts or series of prompts to influence model behavior to generate more desirable outputs.
  • Random Forest: A supervised machine learning algorithm using multiple decision trees to increase accuracy and stability in prediction; helpful to use with datasets that are missing values or are very complex
  • Red Teaming: Testing AI safety and security through simulated adversarial attacks; reveals security risks, model flaws, biases etc and the results of such testing are passed onto the model developers for evaluation and remediation.
  • Reinforcement Learning: Training a model to optimize actions via reward.
  • Reinforcement Learning with Human Feedback (RLHF): Learning through human preferences; human feedback is provided on the model's output to help align the AI's behavior.
  • Reliability: Consistent and accurate AI performance; ensures it behaves as expected and performs its intended function even with new data
  • Robotics: Designing robots that interact with the physical world
  • Robustness: AI's resilience and functionality under attack; ensures system's functionality, performance and accuracy in variety of environments/circumstances even when faced with changed inputs or security attacks
  • Safety: Minimizing AI harms and risks; also encompasses the prevention of existential or unexpected risks
  • Semi-Supervised Learning: Combines supervised and unsupervised learning; generative AI commonly relies on semi-supervised learning
  • Small Language Models: LLMs optimized for efficiency; optimize them for efficiency and better suiting them for deployment in environments that need faster training
  • Supervised Learning: Trained on labeled data; can be useful for classification or regression
  • Synthetic Data: Artificially created data resembling real data; often used for testing or training machine learning models, particularly in cases with limited, unavailable, sensitive data.
  • System Card: Documents on how AI models collaborate; discloses information about how various AI models work together within a network of AI systems, promotes explainability of the overall system
  • Testing Data: Assessing model performance.
  • Training Data: Datasets to train a model.
  • Transfer Learning Model: Using knowledge learned from one task to perform another.
  • Transformer Model: Neural network focusing on sequence data relationships; by attending to the surrounding words a word can be comprehended
  • Transparency: Comprehensibility and openness in AI; implies openness, comprehensibility and accountability in the way AI algorithms function and make decisions, specific meaning can vary
  • Trustworthy AI: Ethical, secure, and accountable AI; interchangeable with terms responsible AI and ethical AI, refer to principle-based AI development and AI governance,
  • Turing Test: Measuring AI's ability to exhibit human behavior; originally thought of the test to be an AI's ability to converse through a written text, such that a human reader would not be able to tell a computer-generated response from that of a human
  • Underfitting: Model failing to capture training data complexity.
  • Unsupervised Learning: Discovering patterns in unlabeled data; with minimal human supervision, the provided AI is provided with preexisting unlabeled datasets and then analyzes those for patterns
  • Validation Data: Assessing model performance during training; and preventing overfitting
  • Variables: Measurable attribute with different values; can be numerical/quantitative or categorical/qualitative, also called features.
  • Variance: How data spreads from the average.
  • Watermarking: Embedding patterns in AI-generated content for transparency.

Studying That Suits You

Use AI to generate personalized quizzes and flashcards to suit your learning preferences.

Quiz Team

Related Documents

More Like This

Use Quizgecko on...
Browser
Browser