MIS560

reactive machines AI

basic operations; simplest, no learning

limited memory AI

predictive; store previous data to make bettter predictions

theory of mind AI

interact with thoughts of humans; do not exist yet, are in development

self-ware AI

evolved to be akin to the human brain that it has self-awareness (has it’s own emotions & understand emotions)

supervised

trained with given sample data with a known output

unsupervised

trained with given sample data without a known output

semi-supervised

combination of supervised and unsupervised learning

reinforcement

trained by trial and error; program explores their environment and receives or negative rewards based on actions

Types of supervised ML algorithms

linear regression, logistic regression, & Naive Bayes classifier

Deep Learning

based on learning examples (like humans) using artificial neural network; mimics neurons in the human brain

types of unsupervised ML algorithms

K means clustering, apriori

Deepfakes

use of AI and ML to create lifelike but fake images of real people, AI is fed images of the person to replicate, even voices

Ethics of Deepfakes

misinformation, cyberbullying, dead actors & influencers, laws

In functional programming, was is data center around?

functions

What kind of programming are we most familar with?

Imperative & OOP programmin

What are some benefits or functional programming?

concurrency, fault tolerance, easier to debug, shorter code, more math-based

Lewin’s 4 Golden Rules of Changes

1. change should only be implemented for good reason, 2. change is most effective when done gradually 3. change should be planned and not sporadic or sudden, 4. strive to include individuals who may be affected by the change as much as possible in planning for the change

3 Stage model

unfreeze(ensure ready for change), change (execute the intended change), refreeze (change becomes permanent)

Things that can help change

help people understand what’s in it for them, some people may be harmed by the change: this is important to actively manage

Critiques of the model

it is quaint & simple for the modern era, linear in nature when projects and change are often not, little to no emphasis on people’s emotional journey

ethics are involved at multiple levels

individual, community, business, government, society

what is ethics?

refer to standards and practices that tell us how humans ought to act in multiple situations and different roles

What is ethics NOT?

feelings, religion, law, accepted norms, science

What are the 6 ethic lenses?

rights, common good, utilitarian, virtue, care ethics, justice

The rights lens

ethical action is the one that best protects and respects the moral rights of those affected; dignity based on human nature, choose freely to live their own lives

The justice lens

each person should be given their due, and what people are due are often interpreted as fair or equal treatment; people are treated as equals but not necessarily the same

different types of justice

social, corrective, retributive, restorative, distributive

The utilitarian lens

ethical action is the one that produces the greatest balance of good over harm for the most stakeholders; the greater good

the common good lens

interlocking relationship of society are the basis of ethical reasoning; respect and compassion for all others are requirements of such reasoning; life in community is a good itself

The virtue lens

ethical actions ought to be consistent with certain ideal that provide for full development of humanity; ancient approach

Virtues are dispositions and habits that enable us to act according to the highest potential of our character. These include:

honesty/integrity, courage/fidelity, compassion/generosity, tolerance/fairness, love/self-control

The care ethic lens

is rooted in relationships and in the need to listen and respond to individuals in their specific circumstances; not merely following rules or calculating utility; connects interpersonal duties to social duties

Main problems of these lenses:

definition of human & civil rights, definition of what constitutes the common good, definition of what is a good and what is a harm

What pending question does the different lenses lead to different answers?

What is ethical?

5 steps within the framework for ethical decision-making

1. identify the ethical issues

2. get the facts

3. evaluate alternative actions

4. choose an option for action & test it

5. implement your decision & reflect on the outcome

What is data analytics?

analyzing raw data to formulate conclusions

descriptive analytics

what HAS happened

predictive analytics

what COULD happen

prescriptive analytics

what SHOULD happen

diagnostics

WHY did something happen

Stages of a Process flow

1. data collection/extraction,

2. data organization/storage,

3. data cleaning & transformation,

4. KPI (Key Performance indicator)/insight derivation

5. data visualization

common use cases of D.A.

1. e-commerce

2. manufacturing

3. IT security

4. healthcare

Benefits of Data Analytics

1. performance measurement

   1. social media usage & engagement

   2. summarize effectiveness of marketing campaigns or business operations

   3. interpreting survey results

2. improvement of business processes

   1. more efficient inventory forecasting

   2. pattern detection to prevent fraudulent activity

   3. risk reduction

What are some ethical concerns around data analytics?

1. privacy & security

2. transparency & accountability

3. fairness & non-discrimination

ethical determining questions

1. is it legal?

2. is it ethical?

3. how will our customers react?

Pros of Proliferation of Data Privacy

1. stronger consumer protections

2. more responsible use of data by corporations

3. improved ability to respond to breaches

4. stronger requirements to notify of breaches and ability to impose penalties

Cons of Proliferation of Data Privacy

1. significant organizational effort to implement

2. different protections in different states

3. frequently law makers do not understand all issues involved

4. largely operates on trust - cannot check that everyone did it correctly

Ethics of AI examples

1. biases in systems

2. liability in failure

3. profiling & unnecessary intrusion

4. plagiarism

5. unnecessary collection

Problems with monolith

1. long build/test/release cycles

2. scalability is compromised

3. lack of agility

4. software architecture is hard to maintain and evolve

5. system operation is difficult

6. lack of innovation

7. slow development pace

Microservices

software architecture style in which complex applications are composed of small, independent processes communication with each other using language-agnostic APIs. (highly decoupled & focuses on smaller tasks)

Benefits of a microservice architecture

1. improves fault isolation

2. eliminates long-term commitment to a single technology stack

3. improved agility

4. provides the ability to “fail fast” and recover fast

What is the trade-off of microservices?

1. can be complex

   1. some services can be up while others are down

   2. remotes calls experience latency

2. premium placed on automated deployment

3. increased security vulnerabilities

4. more complex data management