Chapter 1 Intro to AI

How does AI become biased?

The input that is being fed into the machine is either not factual or controversial

How do Siris and Alexas get smarter?

They use active learning to take in input then processes it then puts in into effect

What machine learning algorithm groups data and finds patterns without the help of training data?

Unsupervised Learning

Are self driving cars an example of strong or weak ai?

Weak AI

What is most of the worlds AI considered?

Weak AI

voice assistant tools like the Google Home or Siri are examples of what type of AI?

Weak AI

Is of a picture guessing AI an example of Strong or Weak AI? Why?

Weak AI because it's only task is to guess the picture you are making

What would be considered in input of an Alexa?

What you say to it

How does Auto Draw work?

When you draw Google runs your colors and shaped through algorithms to find what you are drawing

Can AI automatically adapt to changes in your appearance in face id?

Yes

Is Grammerly an AI?

Yes

What is an algorithm?

a bunch of variables that build off of each other

What seems to be the easiest way to combat algorithmic bias?

after creating algorithms, test for bias, retrain bias algorithms, and prevent bias from occurring

What is a node?

an algorithm that receives input from a different algorithm and passes it onto another

What benefits does this give white men?

better facial recognition results and unfair credit limits

Name three examples of Weak Ai other than Voice Assistant.

chat bots email spam fillers self driving cars apple autocorrect google maps

Do picture guessing AI's use convolutional networks or filters?

convolutional networks

What are some examples of Strong AI?

cyber security language translation machines image recognition machines GPT-3

What does AI's development come with?

ethical considerations

Why is Facial Recognition criticized?

for using data sets that include way more light skinned men than any other groups

How does Apple's TruDepth camera use AI to unlock your smartphone? (face id)

it projects 30,000 invisible dots onto a user's face to create a so-called 'depth map' and compares that to saved data to allow access

What are some of the impacts of facial recognition bias?

misidentification of Black protesters or false positive matches in image databases

What do Siris and Alexas do with word phrases you say to them?

they break them into individual sounds and then runs them through a database, using sophisticated algorithms to find the right answer

What is the purpose of Tensor Flow?

to group data

What is the purpose of neutral networks?

to help humans make intelligent decisions with limited human assistance

Should facial recognition software be used?

yes but with major caution

What percentage of global consumers who have AI-enabled interactions with organizations daily in 2020 and 2018?

2018- 21%. 2020-54%

What about weekly uses with AI-enabled interactions in 2018 and 2020?

2018-31%. 202-27%

How many people use Grammerly to improve their writing?

30 Million

What percent of users turn on Dark mode after 6 PM?

65%

How do consumers feel generally about AI?

76% of consumers are concerned with misinformation about AI

How many times a day do Gen-Z consumers unlock their phones?

79 times

What is Artificial Intelligence?

A computer system that is able to perform tasks that ordinarily require human intelligence

What is Strong AI?

A machine that is able to apply intelligence to any problem

What is Weak AI?

A machine that is able to apply intelligence to one single problem

Which data set is most likely to cause bias results?

A pesticide company tests its product on a variety of different species in a variety of different climates to determine its efficiency

Define Output.

A series of understanding in motion from things that were communicated or experiences

How is AI made intelligent?

AI uses algorithms to learn and process info

How does AI effect future job employment?

AI's can easily be programmed to do a humans job so in result they will be hired instead of humans

Define Input.

Data collected that is then learned from

True or False- Because AI is able to learn on its own, it doesn't require the use of algorithms to develop?

False

True or False: Artificial Intelligence poses an immediate threat to job security around the world

False

True or False: Most Artificial Intelligence is inherently unethical.

False

How might the ability to figure out the emotion in a text be useful?

For sentimental analysis so the AI can come up with the best and most accurate response

How do Netflix recommendations know what to recommend?

It asses various factors such as your viewing history, how you rate titles, what others with similar taste search, and similar movies in the same genres you like. All of these components are fed into the algorithm

How might Natural Language processors determine the emotion of a text?

It implements a keyword-based approach where the emotional state of a sentence is constituted by the emotional affinity of the sentence's emotional words

What is considered the process of an Alexa?

It taking in the info and using the Machine learning algorithm

How does an Alexa learn from what you say?

It uses a Machine learning algorithm to process the request and look up any requests that are similar

How does an Alexa comprehend what you are saying?

It uses a language processor

What are the sub categories of AI?

Machine Learning Natural Language Processor

What is an example of machine learning?

Natural Language Processors

What is one sub set of algorithms in Machine Learning?

Neutral Networks

There are two programs developed to use the same AI, Program A is exposed to 10 users while Program B is exposed to 1000. What is the relationship between these two Programs as a result of the exposer to users?

Program B is more likely to create more accurate output than Program A

When we think of AI what do we associate it with?

Robotics and science fiction

Give an example of a neutral network.

Tensor Flow

What would be the output of an Alexa?

The Alexa learning from what you said

What layers of a neutral network randomly assigns weight into input values in order to evaluate the importance of certain input?

The Hidden layer