/
AI Flashcards
Save to my account
Sign up
AI Flashcards
Foundations of AI
Study
1
Question
What is the key difference between AI systems and traditional human-like programming?
Page 1
Answer
AI systems learn and make decisions from data instead of only following fixed rules; traditional programming requires writing explicit rules that the software follows.
2
Question
What does the progression AI → ML → DL represent?
Page 1
Answer
Artificial Intelligence encompasses Machine Learning, which encompasses Deep Learning.
3
Question
Define Narrow/Weak AI and give an example of its common use.
Page 1
Answer
Narrow/Weak AI performs specific tasks; most real-world systems today use it for task automation and insights.
4
Question
What characterizes General/Strong AI?
Page 1
Answer
Can think and learn like a human across many different tasks (research stage).
5
Question
What is Super AI?
Page 1
Answer
AI beyond human intelligence across many different tasks (research stage).
6
Question
Who invented the first mechanical calculator in 1642?
Page 1
Answer
Blaise Pascal.
7
Question
What is the Turing Test from 1950?
Page 1
Answer
A machine is considered intelligent if a human cannot tell whether they are talking to a machine.
8
Question
When and by whom was 'AI' coined, and what modern example passed the Turing Test?
Page 1
Answer
1956 by John McCarthy; Google Duplex passed in 2018.
9
Question
What were the 'golden years' of AI?
Page 1
Answer
1956-1974.
10
Question
What is the current industrial revolution that AI is part of?
Page 1
Answer
The 4th industrial revolution (starting around 2017).
11
Question
What are the five steps to design an AI system?
Page 1
Answer
1. Identify the problem. 2. Prepare data. 3. Choose algorithms. 4. Train algorithms with data. 5. Run on selected platform.
12
Question
What is data labeling/annotation?
Page 1
Answer
Adding correct 'answers' to raw data; e.g., labeling images as 'cat' or 'dog'.
13
Question
What is data anonymization?
Page 1
Answer
Removing personal identifiers from data.
14
Question
When is synthetic data used?
Page 1
Answer
When real data is missing, too sensitive, or more examples are needed.
15
Question
What does data cleaning involve in preparation?
Page 1
Answer
Fixing messy data: errors, missing values, duplicates.
16
Question
Define feature engineering.
Page 1
Answer
Constructing new features from existing data to train a model.
17
Question
What is data wrangling?
Page 1
Answer
Transforming and mapping data from one raw format into another.
18
Question
What is data mining?
Page 1
Answer
Discovering patterns in large datasets.
19
Question
What is data warehousing?
Page 1
Answer
Collecting data from multiple sources and storing it in a structured way.
20
Question
Distinguish between structured and unstructured data.
Page 2
Answer
Structured: organized; Unstructured: text, audio, video, images (most business value).
21
Question
What is big data?
Page 2
Answer
Unstructured complex data sets too big for traditional tools to store, process, or analyze efficiently.
22
Question
Define money/dollar density of data.
Page 2
Answer
How strongly a piece of data affects a company's revenue or costs.
23
Question
Give examples of high money density data.
Page 2
Answer
Transactions, purchases (direct revenue/cost impact).
24
Question
What are examples of low money density data?
Page 2
Answer
Clicks, likes.
25
Question
What is core business data?
Page 2
Answer
Customer actions and other info closely tied to business value (high money density).
26
Question
Why is achieving the last 10% accuracy in models extremely difficult?
Page 2
Answer
Processed data is key, but fine-tuning yields diminishing returns.
27
Question
What is Machine Learning (ML)?
Page 2
Answer
Application of AI where systems automatically learn and improve from experience without explicit programming, by learning patterns from data.
28
Question
What is Symbolic AI?
Page 2
Answer
Type of AI using explicit rules, logic, and human-defined knowledge to solve problems (symbols/nouns + relations/adjectives/verbs + logic AND/OR/NOT).
29
Question
How does ML differ from symbolic AI approaches?
Page 2
Answer
Symbolic relies on hand-written rules; ML learns patterns from data.
30
Question
What is Supervised Learning?
Page 2
Answer
Machine learning where the model learns from labeled data (inputs paired with correct answers).