Module 5- AI & Blockchain and Cryptocurrency

Put the six steps of the of the Amazon Web Services (AWS) deep learning process in the correct order.

1. Collect data. 2. Choose and optimize the algorithm. 3. Set up and manage the environment for training. 4. Train, retrain, and tune the models. 5. Deploy models in production. 6. Scale and manage the production environment

Put the steps for how an image of a house can be created using GANs in the correct order.

1. The GAN is fed actual images of houses. 2. The generator is fed random inputs. 3. The generator creates a sample based on these inputs. 4. If the discriminator chooses the real image instead of the generated image, the system will continue to develop/adjust generated images.

In a recent study by Deloitte, _____ percent of executives responded they are "not comfortable" using data from advanced AI systems.

67, sixty seven, or sixty-seven

Select the true statement about artificial general intelligence (AGI).

AGI can potentially perform any intellectual task a human can.

Which of the following are true statements about bias in AI?

AI bias is defined as variance that exists in the output provided by AI algorithms. Cognitive bias includes conscious or unconscious errors in cognition that impact an individual's judgments, assumptions, and decision making. Biases in AI often result from two factors: cognitive bias or incomplete data.

Select the true statements about AI detection software.

AI detection can help to ensure submitted work is unique, authentic, and not generated by AI. AI detection software is offered by OpenAI. AI detection software is designed to counter the risk of plagiarism.

Select the true statements about AI language models.

AI language models are computer software programs that use AI to process and generate human languages. AI language models, specifically transformer-based models, use neural networks that can process input text accurately. There are a number of AI language models in use including GPT-3 and GPT-4.

Design AI technologies that are socially responsible.

AI technologies should be advantageous for diverse populations and society as a whole.

Detection software tools are designed to interpret text and potentially detect if _____ generated the text.

AI, ai, Ai, aI, or artificial intelligence

Which of the following are considerations of AI chatbot ethics?

Accurate data representation Transparency Chatbot identity and persona

What were the notable achievements in the field of AI between 2012 and 2020?

Alibaba's AI surpassed human performance in a reading comprehension test. Facebook's AI chatbots developed their own language. Google's neural network learned to identify cats using Internet images.

Which of the following are uses of ChatGPT?

Analysis of computer code Language translation Sentiment Analysis

AI

Artificial Intelligence

What is the primary goal of the AI Act passed by the European Union?

Balance innovation with the safeguarding of individual rights

Bias and discrimination

Bias and discrimination in AI can be intentional or unintentional. The goal is to minimize the occurrence of bias in AI systems.

Which of the following are true statements about Chat GPT?

ChatGPT is an extensive language model that is trained on vast amounts of text data. Chat GPT was developed by OpenAI Unsupervised learning uses machine learning algorithms to identify patterns in text data without the input or interference of humans.

Which of the following should be considered before a student uses information provided by an AI chatbot?

Chatbot Plagiarism Misinformation False Representation

Chatbot plagiarism

Chatbot responses may contain information from preexisting sources.

Chatbot identity and persona

Chatbots need to be trained to ensure they behave appropriately. If chatbots are not properly trained, they could exhibit traits such as racism, sexism, and abusive language.

Which of the following are initial steps in building an AI model?

Choosing the appropriate algorithm Identifying the problem statement and goals Preparing and gathering the necessary data

Surveillance

Countries across the globe have installed millions of cameras to actively monitor public spaces. This rapid deployment of monitoring systems is concerning to many civil rights activists.

Reinforcement Learning

Covers decision-making processes and focuses on using machine learning to analyze a scenario which in turn creates learning.

Transparency

Creating output that can be shared with the public about how systems use data and interact with populations should be available.

Input Layer

Data and information from a variety of sources is entered into the neural network through this layer.

Disadvantages of Cloud-based AI

Data privacy Internet connectivity concerns Lack of platform control

Avoid forming or reinforcing unfair bias in AI systems.

Data sets and algorithms used in AI systems often include (whether intentional or unintentional) unfair or misrepresentational prejudice.

Incorporate transparency and accountability concerning AI practices.

Disclosure of AI practices, procedures, applications, and methods is encouraged in order to create and maintain a positive AI culture.

Human Judgement

Do humans need to be involved in AI training and for AI systems to optimally run? Should this be a concern?

What roles are important in an in-house AI team?

Engineers Data scientists Domain experts

Advantages of Cloud-based AI

Enhanced security Cost savings Increased efficiency Data management

What roles does historical data play in AI modeling?

Enhancing predictions in sales forecasts Training AI to identify fraud patterns

The FTC acknowledged that AI has the potential to advance almost every industry, but all of this potential comes with some negative risks that must be addressed. Concerning AI, which of the following does the FTC urge organizations to provide evidence of?

Ensure that decisions are fair, robust, and empirically sound. Hold themselves accountable for compliance, ethics, fairness, and nondiscrimination. Explain how algorithms make decisions.

What ethical considerations should be addressed when developing and deploying AI systems?

Ensuring AI systems are free from biases Maintaining transparency in AI operations and decisions Protecting against misuse of AI technologies

What considerations are important when determining the architecture of an AI model?

Ensuring the architecture impacts performance positively Configuring layers and connections in the neural network

What are important considerations when selecting AI tools for an organization?

Evaluation of custom and off-the-shelf options Development of integration policies Alignment with organizational goals

Collaboration

Experts agree that a multidisciplinary effort to advance AI is needed. Authorities across many disciplines need to establish a common set of standards and platforms.

Which of the following are guiding principles for the ethical use of AI?

Follow laws, rules, and standards covering AI. Design AI technologies that are socially beneficial. Avoid forming or reinforcing unfair bias in AI systems.

Which of the following are popular AI detection software programs?

GPT Zero Turnitin AI Writing Detection Model AI Text Classifier by OpenAI

How is the United States contributing to AI safety and governance?

Housing the AISIC under the National Institute of Standards and Technology (NIST) Creating the U.S. AI Safety Institute Consortium Implementing an executive order on AI safety by President Biden

Proper Communication

How the software handles abusive language and generates controversial content.

Human Values in AI

Humans have to calculate the potential risks and consequences of their actions based on a number of factors, including morals and values.

Select the true statements about information architecture.

IA is a crucial component of the overall AI strategy of an organization. Many experts agree that there is no AI without the analysis and implementation of IA. IA is about helping people understand their surroundings and allowing them to find what they're looking for in the real world and online.

The _____ Ladder proposes a series of steps/considerations that breaks down AI strategy into individual pieces or rungs on a ladder to better connect data and AI systems.

IBM

Which of the following are guiding principles for the ethical use of AI?

Implement data security and privacy principles. Follow laws, rules, and standards covering AI. Incorporate transparency and accountability concerning AI practices.

Which of the following are true statements about trust in AI platforms?

In a recent study by Deloitte, 67 percent of executives responded they are "not comfortable" using data from advanced AI systems. Creating transparency is an important step in building trust in AI platforms. Collaboration can help to build trust in AI platforms.

Output Layer

In this layer, information from the previous layers is analyzed. After this analysis occurs, a final prediction or conclusion is made.

Select the true statements about unsupervised learning.

In unsupervised learning, machine learning algorithms are used to examine and cluster unlabeled data sets. Unsupervised algorithms are used to uncover hidden associations within data sets or patterns within data sets that may go unnoticed without the assistance of humans. In unsupervised learning, clustering is used.

Potential Positive Outcomes of AI

Increased productivity Decreased human risk and injury Diminished human-caused errors

IBM's AI Ladder Approach to AI

Infuse Analyze Organize Collect

Select the true statements about neural networks.

Interconnected artificial neurons are the main building blocks of simple neural networks. A neural network is a method in artificial intelligence that teaches computers to process data in a way that is inspired by the human brain. Neural networks are a type of machine learning process called deep learning.

What distinguishes internal data from external data in AI applications?

Internal data is generated within an organization, while external data is sourced from outside entities.

Which of the following best explains why data governance is important in AI implementation?

It is crucial for maintaining data quality, consistency, and ethical standards.

Potential Drawbacks of AI

Job losses due to automation Privacy concerns

Unsupervised Learning

Machine learning algorithms are used to examine and cluster unlabeled data sets.

Select the true statements about machine learning.

Machine learning systems improve their accuracy gradually through a continual process of learning. Machine learning uses data and algorithms to emulate the way humans learn. AI can learn from machine learning models created by inputs and analysis of anticipated outputs.

Select the true statements about how machine learning can be used to solve a problem.

Machine learning uses various data inputs to build an algorithm. Algorithms created by machine learning take data and answers to questions and use this to develop a machine learning model. If-then statements are often used in machine learning models.

NLP

Natural Language Processing

NLU

Natural Language Understanding

Which of the following terms are associated with AI chatbots?

Natural Language Understanding (NLU) Natural Language Processing (NLP) Artificial Intelligence (AI)

Select the true statements about deep learning.

One of the benefits of deep learning is that portions of the learning process are automated, eliminating some of the need for human intervention. Deep learning is a type of artificial intelligence that uses data and algorithms to create learning. Deep learning is often referred to as scalable learning.

What company is responsible for the development of ChatGPT?

OpenAI

Why is continuous monitoring important in AI model management?

Provides ongoing improvements based on performance assessments

Interagency Coordination

Sharing of experiences and best practices for AI governance should be maintained between governmental agencies.

Transparency covers the consequences of interacting with AI-assisted chatbots in which ways?

Site ownership Data Collection Information privacy and security

Select the true statements about supervised learning.

Supervised learning uses data class labels within data sets that specify what the data represents. In supervised learning, class labels are used to create the classification model. A classification model is designed when machine learning allocates a label value to a specific class and then seeks to recognize these values to decide what categories they fit into.

Transparency

Technological transparency concerning AI-assisted chatbots refers to awareness regarding all aspects of the use of the software.

Scientific Integrity and Information Quality

The establishment and maintenance of high standards as they relate to AI information quality and decision making can assist in gaining public trust and confidence in AI.

Privacy

The goal of machine learning practitioners is to design models that garner ideas and insights from data sets without revealing specific or sensitive data found within the data set.

Misrepresentation

The large volumes of data that AI chatbots are trained on pose the risk of inaccurate or misleading response generation.

Select the true statements about generative adversarial networks (GANs).

The neural networks in a GAN include a discriminative network. The neural networks in a GAN include a generative network. GANs are designed to compete against each other (this is why they are termed adversarial) to create artificial instances of data that are interpreted as real data.

Disclosure and Transparency

There must be a method to detect when AI is in use as well as a way to address possible impacts of AI on end-users.

Hidden Layer

This layer processes data sent by the input layer or from other associated layers.

What was the primary function of the Turing Test as described by Alan Turing in 1950?

To determine whether a machine can engage in conversation indistinguishably from a human

Which of the following are steps of the Amazon Web Services (AWS) deep learning process?

Train, retrain, and tune the models. Scale and manage the production environment. Choose and optimize the algorithm.

Which of the following are the types of data used in AI?

Unstructured data from social media posts Structured data from customer databases External data from competitor analysis

Accurate Data Representation

Users of AI chatbots should take precautions to ensure that the data and information that is returned are accurate.

Supervised Learning

Uses data class labels within datasets that specify what the data represents. These class labels are used to create the classification model.

What are potential misuses of AI that need to be guarded against?

Using AI for deepfakes AI-enabled cyberattacks AI surveillance applications

Public Participation

When financially feasible and including adherence to legal requirements, organizations should allow the public to be part of the AI rule-making processes and procedures.

Select the true statements about student use of AI chatbots and chatbot plagiarism.

You should assume that responses provided by AI chatbots may contain information from other sources. Most AI chatbot providers report that responses provided by AI chatbots are only produced with unique responses. However, in some instances this has been proven false. Most AI chatbots do not provide source or attribution information.

The White House Office of Science and Technology Policy has identified ten principles to consider when articulating regulatory and non-regulatory methods for the development and use of AI. Principles identified by the White House Office of Science and Technology include which of the following?

analysis of safety and security cost/benefit analysis create public trust in AI

The variance that exists in the output provided by AI algorithms is referred to as AI _____.

bias

The variance that exists in the output provided by AI algorithms is referred to as AI:

bias

Which of the following are Cloud-based AI applications?

chatbots AIaaS IoT

The use of computer models that are designed to simulate human thought processes is referred to as _____ computing and is often used in complex environments where solutions may be ambiguous or difficult to uncover.

cognitive

A type of machine learning model that uses two neural networks is referred to as a(n) _____ adversarial network. These networks are designed to compete against each other (this is why they are termed adversarial) to create artificial instances of data that are interpreted as real data.

generative

In which of the following healthcare areas is AI used?

health tracking robotic surgery pharmaceuticals

According to John McCarthy from Stanford University, artificial intelligence (AI) is "the science and engineering of making intelligent machines and intelligent computer programs. It is related to the similar task of using computers to understand _____ intelligence, but AI does not have to confine itself to biologically observable methods."

human

Machine learning systems improve their accuracy gradually through a continual process of learning and rely on _____ interaction and input to assist in creating learning.

human

Which of the following are neural network layers?

input layer hidden layer

Two subfields if AI include which of the following?

machine learning deep learning

In which of the following areas of government is AI used?

military policing automation of routine tasks

Artificial Intelligence (AI) language _____ are computer software programs that use artificial intelligence to process and generate human languages.

models

Cognitive systems use a decision-making approach similar to that used by humans to analyze large data sets to make decisions. These systems use _____ language processing to read and process text like a human. They break down statements and sentences by analyzing them for grammar, structure, and relationships within the words.

natural

Which of the following are ethical considerations of AI?

privacy human judgement

Which of the following are potential drawbacks of AI?

privacy concerns job losses due to automation

Identify current business applications and uses of AI.

sales and marketing AIOps accounting

Which of the following services are provided by Cloud computing networks?

storage processing software applications

Which of the following are machine learning methods?

supervised learning reinforcement learning unsupervised learning

Learning models that are _____ use neural networks that can process input text and accurately produce output sequences.

transformer-based