MIS 200 Exam 3 - Armeen

What is AI?

Systems capable of performing tasks that typically require human intelligence

What are the different types of AI?

Artificial Narrow Intelligence, Artificial General Intelligence, and Artificial Superintelligence

What is Artificial Narrow Intelligence?

AI that performs one specific task

What are some examples of Artificial Narrow Intelligence?

Facial recognition, translating languages, and recommending movies

What is Artificial General Intelligence?

AI that could think, learn, and reason across any domain like a human

What is Artificial Superintelligence?

AI that surpasses human intelligence in every field

What is the Turing Test?

Checks if an AI can talk like a human so well that you can’t tell it’s a machine

How does AI work?

It takes training data to predict (not understand) and generates a response by taking your input and turns it into results that turned those predictions into texts

What does traditional AI do?

Focuses on classification or prediction

What is an example of traditional AI?

Spam detection in emails

What is testing data?

A specific subset of a dataset used to evaluate the final performance and quality of an AI or machine learning model after it has been trained

What is training data?

Books, websites, images, and other digital content used to recognize patterns and relationships

What are the limitations of AI?

Hallucinations, No Sense of Truth, Struggles with Long or Complex Tasks, and Biased or Outdated Information

What are sources of biases in AI?

Data-level: representation

Human-level: Labeling, cultural

Algorithmic-level: Tokenization (linguistic), stereotyping

What is representation bias?

A data-level bias where the dataset does not include enough variety to represent real-world diversity

What is labeling bias?

A human-level bias where human annotators tag or level data based on personal or cultural views

What is tokenization bias?

An algorithmic-level bias where the model’s language processing favors certain words, languages, or patterns

What is an Large Language Model (LLM)?

trained on massive datasets to understand and generate human-like texts

What is prompt engineering?

The process of guiding a generative AI system to produce useful and reliable results

What is zero shot prompting?

The most simplest way to use AI, relying entirely on the model’s existing knowledge, you ask the AI to perform a task without providing any example or prior context

What is few-shot prompting?

Similar to training an intern, you show what “good” looks like before assigning a new task, use this when tone, structure, or consistency matters (executive summaries or client reports), you give the AI 1-3 examples of the desired style, structure, or tone before asking for your actual output.

What is Chain of Thought prompting?

When you want the AI to reason like an analyst, not just summarize, essentially you are teaching the AI to simulate the reasoning pipeline when you ask the AI to “think step-by-step” before giving an answer, encouraging reasoning and logicHat W

What are the key components of an effective prompt?

Role, task, context, and constraints/format

What is role prompting?

A technique that frames the AI’s expertise, like assigning a consultant role in your company, you assign a specific role or perspective to the AI before giving it a task

What is self-critique prompting?

A technique that is similar other than a quality control loop, you can make the AI check its own work. You ask the AI to evaluate its own output and improve it

What is context window/token awareness?

Understand how much text or context an AI model can “remember” and process at once. Critical for managing long documents, datasets, or contracts efficiently by segmenting or summarizing information to prevent the loss of key details

What are the 3 major purposes of AI implementation for companies today?

Improve customer experiences, boost employee productivity, and accelerate process optimization

What can AI-based personalization do?

Creates individualized recommendations, content, and offers using customer data. Helps increases engagement, satisfaction, and loyalty and powers dynamic pricing and tailored email or app content

What does data augmentation do?

Create simulated data for machine-learning training when real data is limited or sensitive

What does conversational analytics do?

Analyzes unstructured data (such as customer reviews, surveys, call transcripts) to detect patterns and sentiment

What are the barriers to AI adoption?

Data quality and access and the Learning Gap

What is the Learning Gap?

Many systems that do not learn, adapt, or integrate into workflows

How can organizations adopt AI strategically?

Start with business goals, look for processes that are repetitive, time-consuming, rule-based, data-heavy, customer-facing, prioritize by impact and ease, build governance and skills, and monitor and improve

What is a defining characteristic of ANI?

Performs a single task

What is AGI?

Think and learn across domains like a human

What is an ASI?

Surpasses human intelligence in all fields

Which action helps mitigate representation bias in AI datasets?

Collect data from a diverse set of populations

Which statement about tokens in AI prompts is true?

Tokens are pieces of text such as words or subword units

True or false: In reinforcement learning, the model learn by interacting with an environment and receiving rewards that guide behavior

True

Which statement best defines Generative AI?

It generates new content (text, images, or other data) based on patterns it has learned from large datasets

What does the Turing Test measure?

The ability of an AI to imitate human conversation well enough to fool a human

Human annotators tag or label data based on person or cultural views, this is an example of which bias?

Labeling bias

A student is marked “Pass” if their score in B2 is 60 or higher. Otherwise, “Fail”. Which formula is correct?

=IF(B2>=60,”Pass”,”Fail”)

If you ask the AI to “think step-by-step” before giving an answer, you are using the ________ prompting technique

Chain-of-Thought

True or false: Prompt chaining involves breaking down complex problems/tasks into smaller, sequential steps

True

Which Excel function can display a value based on a condition?

IF

They key components of an effective prompt include:

Task, context, role, constraints

What is the main advantage of using the IFS function over nested IF statements?

It allows multiple conditions to be tested without nesting multiple IF functions

Conversational analytics helps businesses by:

Detecting customer pain points from unstructured data

When selecting AI projects, organizations should prioritize:

High-impact, easy-to-implement use cases

True or false: The formula =IF(OR(B2>=18,C2>30000),”Eligible”,”Not Eligible”) correctly returns “Eligible” only if age is at least 18 and income is greater than 30,000

False

True or false: Organizational resistance to AI often comes from employee fear and lack of understanding

True

What is one reason data quality is a barrier to AI adoption?

Poor data leads to unreliable or biased outputs

In Excel, what is the issue with the formula below?

=IF(A2:A10>50,”High”,”Low”)

A range is used instead of a single cell in the logical test

True or false: In Excel, formulas must be efficient because unnecessary steps can make formulas more complex without improving the result

True

In Excel, what is the issue with the formula below if B2 contains a number?

=IF(B2=”10”,”Pass”,”Fail”)

The number 10 should not be in quotes

In Excel, which formula correctly returns “Approved” if A2 is at least 70 and B2 is at least 80?

=IF(AND(A2>=70,B2>=80),”Approved”,”Denied”)

In Excel, why is the condition >10 incorrect when the requirement is “at least 10”?

>10 excludes 10

In Excel, what is the issue with the formula below?

=IF(AND(A2>=50,B2=”Yes”,”Pass”,”Fail”))

Return values are incorrectly placed inside AND

In Excel, what is the issue with the formula below?

=AVERAGEIF(A2:A20,”East”,B2:B20,”Yes”)

AVERAGEIF only supports one condition

In Excel, what is the effect of writing “!0” instead of 10 in a formula?

Converts the value to text

In Excel, what is the issue with the formula below?

=IF(A2>=10,AND(B2<5,C2=”Yes”0,”Accept”,”Reject”)

The logical test is spilt incorrectly; all conditions must be combined into a single logical test

In Excel, you want to write a nested IF formula to assign grades based on the following conditions:

• If A2 is at least 90 = return “A”

• Else if A2 is at least 80 = returns “B”

• Otherwise = return “C”

Which of the following explains the issue with this formula?

=IF(A2>=90,”A”,IF(A2>=80,”B”,IF(A2>=70,”C”)))

The third IF is inefficient because the value_if_false of the second IF a Ready covers all remaining cases

True or false: In a nested IF structure, once earlier conditions are ruled out, the value_if_false represents all remaining possible cases

True

True or false: The formula =SUM(A1+A2) is inefficient because it uses the SUM function unnecessarily when direct addition (=A1+A2) is sufficient

True

True or false: The formula =COUNTIF(A1:A10,”>50,B1:B10,”Yes”) is incorrect because COUNTIF only supports one condition

True

In Excel, what is the issue with the formula below?

=IF(AND(J2>=.1,I2>90,000),”High_Bonus”,”No”)

The number 90,000 is incorrectly written with a comma, causing Excel to treat it as separate arguments

In Excel, what is the key differences between AND and OR?

AND returns TRUE only if all conditions are TRUE; OR returns TRUE if at least one condition is true