1. Foundations Flashcards

1
Q

What is AI?

A

Hallmarks of human intelligence: ability to think creatively, consider various possibilities, and keep
a goal in mind while making short term decisions.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are common definitions of AI?

A
  1. Machines performing tasks that normally require human intelligence (A branch of computer science concerned with creating technology to do things that
    normally require human intelligence)
  2. Alan Turing, a cryptographer and mathematician, developed a test to determine whether a machine is intelligent (1950) (A machine was considered intelligent if it produces responses to human interviewer that fool the interviewer into thinking the responses are human)
  3. Definitions include common elements of AI: technology, autonomy, human involvement and
    output
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are common elements of AI?

A
  1. Technology: use of technology and specified objectives for the technology to achieve
  2. Autonomy: level of autonomy by the technology to achieve defined objectives
  3. Human involvement: need for human input to train the technology and identify objectives for it to follow
  4. Output: technology produces output, e.g., performing tasks, solving problems, producing content
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is machine learning?

A

The process of training machines to display AI behavior.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are the main types of machine learning?

A
  1. Supervised learning: Labeled data that is grouped or classified into categories via the AI system. Used for text recognition, detecting spam in email, etc.
  2. Unsupervised learning: Unlabeled data; typically used for pattern detection.
  3. Reinforcement learning: An AI system is rewarded for performing a task well and penalized
    for not performing it well. Over time, learning to maximize the rewards and develop a system
    that works.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is an example of supervised learning?

A

For email filtering, the algorithm is trained using a labeled dataset
containing both spam and legitimate emails. It extracts the relevant information to create patterns to predict whether future emails are spam or legitimate.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is an example of unsupervised learning?

A

Outliers in the data such as banking data

Reviewing transactions for any fraudulent behavior

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is an example of reinforcement learning?

A

Self-driving cars: the system is rewarded when it keeps a car on the road and gets it to the destination where it is supposed to go. It is personalized if the car goes off the road or hits another object. The system learns over time to maximize the rewards, resulting in a better performing self-driving car.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are the risks in the use of AI?

A
  1. AI systems are implemented in vast and complex environments
  2. The data used for AI will change over time
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are the 5 OECD’s main dimensions for AI?

A
  1. People and planet
  2. Economic context
  3. Data and input
  4. AI model
  5. Tasks and output
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What does the OECD’s ‘AI model’ dimension refer to?

A

Technical type and how the model is built and used.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Who are the relevant stakeholders to consider when working with AI?

A
  1. Individuals who look at the broader societal influences of AI, such as anthropologists, sociologist or others who work in social sciences.
  2. Individuals who develop and implement AI systems.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What does the OECD’s ‘people and planet’ dimension refer to?

A

Identifies individuals and groups that might be affected by the AI system. (ex, human rights, the environment, and society –> privacy comes into play here)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What does the OECD’s ‘economic context’ dimension refer to?

A

The AI system is looked at according to the economic and sectoral
environment in which it operates.

Characteristics include:

  • The sector where the AI system operates (ex, financial, health care or education)
  • The business function or model for the AI system
  • Necessity of the AI system to operations
  • How it is deployed and the impact of the deployment
  • Scale of the system
  • Technological maturity of the AI system (a newer system may not have been tested on as much data over time; more mature systems may be more effective)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What does the OECD’s ‘data and input’ dimension refer to?

A

The type of data used and expert input.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What does the OECD’s ‘tasks and output’ dimension refer to?

A

Tasks that AI systems perform, its outputs and resulting actions from those outputs.

Characteristics include system tasks, systems that combine tasks and actions, evaluation methods used to look at how tasks and systems perform.

13
Q

What is expert input?

A

Human knowledge that gets codified into rules.

Includes characteristics such as how data was collected and hat collection method was used (by machine or by human), structure of the data and data format.

13
Q

What is the difference between AI governance and framework?

A

It is the ‘why’ and ‘how’ of AI governance.

AI governance principles are a set of values, whereas an AI governance framework is a means to operationalize those values.

14
Q

What are AI governance frameworks?

A

AI governance frameworks provide guidance for operationalizing values that come from principles.

While there are similarities among AI frameworks, they are often context-sensitive and fit for specific purposes.

Frameworks are not one-size-fits-all. In addition, an organization can align to one framework in multiple ways.

15
Q

What are some examples of AI governance frameworks?

A
  • International Organization for Standardization (ISO): Several may apply, including ISO 42001 (Management Systems) and ISO 31000:2018 Risk Management Guidelines
  • National Institute of Standards and Technology (NIST) AI Risk Management Framework - U.S.
  • Institute of Electrical and Electronics Engineers (IEEE) 7000-21 Standard Model Process for Addressing Ethical Concerns during System Design
  • Human Rights, Democracy, and the Rule of Law Assurance Framework for AI Systems (HUDERIA) - Council of Europe
  • Other standards specific to jurisdiction or industry
16
Q

Why was the 1956 Dartmouth conference instrumental for AI?

A
  1. Before Dartmouth it was isolated disciplines (psychology, computer science, linguistics, engineering).
  2. John McCarthy (assistant professor in mathematics) and 3 senior researchers proposed that “every aspect of learning or any other feature of intelligence can, in principle, can be so precisely described that a machine can be made to simulate it.”
  3. Allen Newell and Herbert Simon introduced the Logic Theorist, considered by many as the first AI program.
  4. The term AI was adopted creating AI as a field of research.
17
Q

What are the fluctuations of AI development?

A
  1. Mid 1950s - mid 1970s AI research labs are established at top universities
    - John McCarthy - First AI programming language
    - ELIZA (early natural language processing) developed at MIT
  2. Mid-1970s - mid 1980s period of skepticism and funding cuts
  3. Mid 1980s - late 1980s renewed interest in AI
    - Expert and computer systems emulated human decision-making ability
    - Japanese government invests in the development of AI-powered computers (Fifth Computer Systems project)
  4. Late 1980s - late 1990s decreased interest and high cost of expert systems is recognized
  5. Late 1990s - 2011
    - IBM’s Deep Blue defeats world chess champion (1997)
    - Internet champions the big data era
    - Computational power and machine learning improve AI capabilities (shopping recommendation algorithms and smart phone voice assistants)
  6. 2011 - present
    - Deep learning advancements (neural network data training)
    - Google’s AlphaGo defeats Go world champion (2016)
    - Open AI’s Chat GPT (language model capabilities)
18
Q

What is the blossoming of data science?

A
  1. 1960s - 1980s Foundations: the term “data science” is introduces.
  2. 1980s - 1990s Age of databases: Relational Database Management Systems and Structured Query Language transform business relationships with data.
  3. 1990s - 2000s Advent of the internet: “big data” and data mining emerge.
  4. 2000s - 2010s Rise of data science: increased importance of data-driven decision-making
  5. 2021s - present Current trends: machine learning methods are used to extract insights from data and machine generated data has led to real-time analytics and a need for advanced data processing.
19
Q

What is data mining?

A

Process of discovering patterns in large data sets.

20
Q

What are the modern drivers of AI and data science?

A
  1. Cloud computing (drives AI and data processing capabilities).
  2. Mobile technology and social media (AI models learn from social media platform data).
  3. Internet of things (IoT) (data that feeds into AI models contributing to data science).
  4. Privacy-enhancing technologies (PETs) (a viable approach to data security and privacy concerns).
  5. Blockchain (provides a trusted interface for secure financial transactions).
  6. Computer vision (enables machines to understand the world through images and videos)AR/VR (redefines how people interact with digital content).Metaverse (represents a vision of shared virtual space).
21
Q

How does AR and VR redefine how individuals interact with digital content?

A
  1. AR overlays virtual objects onto the real world.
  2. VR immerses users in entirely simulated environments.

Applies to diverse fields (gaming, therapy, medicine).

22
Q

What are the categories of AI?

A
  1. Recognition (image, speech, or facial recognition; retail product matching; product defect recognition in manufacturing; plagiarism detection).
  2. Detection (fraud detection - credit cards or government services; events and sport videos; cyber events and systems management).
  3. Forecasting (sales and revenue forecasting; ridesharing demand peak times; weather forecasting).
  4. Personalization (Unique online customer profiles)
  5. Interaction support (virtual assistants or chatbots).
  6. Goal-driven optimization (used to optimize a particular problem and find solutions ex. supply chain optimization Or driving route optimization).
  7. Recommendation (product or viewing recommendations for customer based or predictive analytics; decision support systems ex. medical diagnoses; disability case adjudication).