Studied by 894 people
Dark Patterns
tricks used in websites and apps that make you do things that you didn't mean to.
Confirmshaming
emotionally manipulating a user into doing something that they would not otherwise have done.
Design Thinkng Process
Empathize
Understanding a problem situation use of technology from another persons perspective
Important to interview many people and with different attributes
Define
Analyzing interviews and other data leads to understanding of problems from the experience of others
Should result in a useful problem statement for the next stage
Ideate → prototype → test (and repeat)
Developing a prototype takes many iterations
Prototypes are not produces that are ideas made into artifacts
UI
screens, buttons, toggles, icons, and other visuals you interact with while using the website or app.
UX
entire interaction you have with a product including how you feel about the interaction.
Intelligence
the ability to acquire and apply knowledge and skills
AI (Artificial Intelligence)
the theory and development of computer systems able to perform tasks that normally require human intelligence
Machine Learning
an area of AI in which a computer generates rules and predictions based on raw data that has been fed into it. Also, a subfield of artificial intelligence that gives computers the ability to learn without explicitly being programmed.
Supervised ML
labeling examples in a specific domain to teach AI to recognize patterns.
Training Stage
take 80% of the labeled data and instruct the ai to look for patterns in good shoes training data and bad shows training data.
Testing Stage
remove the labels to test the AI
Futurism
the study of social and technological advancement for the purpose of exploring how people will live and work in the future
Why is predicting the future so difficult?
Population growth is not linear
GDP growth is not linear
Tech advancements are not linear
Moore’s Law
Exponential increases including computing power per dollar/microchip
The number of translators in an integrated circuit (IC) doubles about every two years
Narrow, Weak AI
Trained to perform a specific task
Can only perform one thing or a specific set of things
Narrow scope of expertise
Generalized AI, Strong AI
Ability to learn and adapt to new situations
Can perform a wide range of tasks at or near a human like level
Can understand and adapt to a board complex unanticipated experience
Instrumental Convergence
An intelligent agent with unbounded but harmless goals can act in surprising and harmful ways
Unintended consequences
Example: AI designed to make humans smile might learn how to tell jokes
DNS
Domain Name System
Domain Name
any text or string that you can enter into your webpage (example: google.com)
IP Address
Four set numbers (172.25.43.123). If you type in the IP address of a particular website you will be routed to that specific webpage.
The IP address is the internet protocol that has a set of rules that helps millions of devices communicate with each other.
DNS Resolver
Acts as a phonebook in the entire thing and bridges the gap between human communication and the DNS and networking world
Intelligent Behaviors
Machines can recognize a visual scene, understand a text written in a natural language, or perform an action in the physical world.
Functions of Machine Learning System
Descriptive: the system uses the data to explain what happened
Predictive: the system uses the data to predict what will happen
Prescriptive: the system will use the data to make suggestions about what actions to take
Supervised Machines
trained with labeled data sets which allow the models to learn and grow more accurately over time (the most common type used today). Example: trained with pictures of dogs and other things all labeled by humans. The machine would learn ways to identify pictures of dogs on its own.
Unsupervised Machine
a program looks for patterns in unlabeled data. Finds patterns or trends that people aren't explicitly looking for. Example: could look through online sales data and identify different types of clients making purchases
Reinforcement Machines
trains machines through trial and error to take the best action by establishing a reward system. Trains models to play games or train autonomous vehicles to drive by telling the machine when it made the right decisions.
Natural Language Processing
machine learns to understand natural language as spoken and written by humans instead of the data and numbers normally used in programming (siri).
Neural Networks
modeled on the human brain- millions of processing nodes are interconnected and organized into layers. Cells are connected, each cell processing inputs and producing an output that is sent to other neurons. Labeled data moves through the cells with each cell performing a different function. Example: in a network trained to identify a picture the different nodes would assess the information and arrive at an output that indicates the picture's features.
Deep Learning
neural networks with many layers. Modeled on the way the human brain works and powers many machine learning uses- the more layers you have the more potential for complex things to go well
Layered Network
process extensive amounts of data and determine the weight of each link in the network.
Recommendation Algorithms
suggest what information appears on your feed and product recommendations are fueled by machine learning
Image Analysis and Object Detection
machine learning can analyze images for different information (identifying people and telling them apart)
Fraud Detection
machines can analyze patterns
How someone normally spends or where they normally shop
Identifies potential fraudulent credit card transactions, log ins, or spam emails.
Automatic Helplines or Chatbots
customers or clients don't speak to humans but instead interact with a machine.
Use machine learning and natural language processing with the bots learning from records of past conversations to come up with an appropriate response
Medical Imaging and Diagnostics
machine learning can be trained to examine medical images or other information and look for certain markers of illness like a tool that can predict cancer risk
Explainability
the ability to be clear about what the machine learning models are doing and how they make decisions
Bias and Unintended Outcomes
Machines are trained by humans and human biases can be incorporated into algorithms
The program will learn to replicate the biased data and perpetuate forms of discrimination
Chat bots on twitter can pick up on offensive and racist language
Teachable Machine
A web based tool that makes creating machine learning models fast easy and accessible to everyone
How to Use Teachable Machine
Gather: group your example sinto classes that you want the computer to learn
Train: train your model to instantly test it out to see whether it can c correctly classify new examples
Export: sites apps and more- you can download your model and host it online
What to use to teach the machine
Use files or capture examples life
Use it entirely on device without any webcam or microphone data leaving your computer
Images: teach a model to classify images using files or your webcam
Sounds: teach a mode to classify audio bu recording short sound samples
Poses: teach a model to classify body positions using files or striking poses in your webcam
2040/2050 (TED Talk)
the year there is a 50% probability that we will have achieved human level machine intelligence.
Intelligence as an Optimization Process (TED Talk)
steers the future into a particular set of configurations
AI Goal: Make Humans Smile (TED Talk)
Weak AI: performs useful or amusing actions that cause the user to smile
Superintelligent AI: realizes there is more effective way to achieve this goal
Take control of the world and stick electrodes into facial muscles of humans to cause constant beaming grins
SOPA
Stop Online Piracy Act
PIPA
Protect IP Act
TED Talk Video
Only few points from the TED Talk video are added. Watch the video for better understanding.
Note 
Flashcard (22)