CS 3001 Exam 2

intellectual property

a general term for categories of rights in intangible creations of the mind

-Establishes protectable legal interests in the goods and services sold by IP owners and the technology and know-how used to produce them.

4 types of intellectual property

trade secrets, patents, trademarks, copyrights

patent

A patent is a right granted by governments to an inventor for a limited time to exclude others from:

Making, using, offering for sale, or selling the invention in that country.

Importing the invention into that country.

utility patents are the most common type but there is also design

-does not last forever (20 for utitlity, 15 for design)

invention, application, examination, allowance

Trademark

identifies which person or company is the source or origin of goods and services and distinguishes the trademarked goods and services from those made or sold by others

-can take the form of names, brands, symbols, colors(trade dress), sounds and scents

-can get from registration, use, or assignment

-rights are territorial, limited to specific good and services, and you must police your trademarks

copyright

Copyright protects works of authorship that are:

Fixed in any tangible form or medium of expression.

Original in the sense that they are

:• Independently created by the author (or authors).

• At least minimally creative

-protects literary works, music works, dramatic works, art works, ect

-does not protect ideas, systems, methods, or processes

-registration not required unless your are bringing a suit for infringement on U.S work

trade secret

A trade secret is:

Business, financial, and technical information.

Kept confidential by its owner through reasonable efforts.

Economically valuable because the information is not generally known

ex: coca-cola recipe, kfc recipe, adobe photoshop

-not registered; protected by making reasonable efforts to keep info confidential

-protection lasts forever until someone finds out

Computer errors: Operational

examples: southwest airlines data center failed and took 4 days to fix, resulting in over 2,000 canceled flights

Starbucks computer failure shut down 8,000 stores for a day

-knight capital lost 440 million in 45 minutes due to a program trading error

-Chernobyl power plant explosion in 1986 due to the failure of a test program; 30km radius will be contaminated for thousands of years

computer errors: data entry and retrieval

-examples: florida disqualifies voters for the 2000 election, becase a database said they were felons even though the committed misdemeanors

-national crime information center had errors that led to false arrests

computer errors: billing errors and malfunctions

Qwest billing software bug caused it to charge $600 per minute for cell phone usage

Renters in NYC were over-charged-Threatened with eviction when refused to pay

450 Prison inmates with a high risk of violence were mistakenly released after being identified by a program to reduce overcrowding.

computer errors: pricing errors

amazon offered new ipaq computers for 7 pounds instead of 275 pounds.

-they refused to fill orders

computer errors: real time system errors

Mars Climate Orbiter- One team used metric units, the other English units

Ariane 5 rocket -Software reuse from Ariane 4 led to counter overflow

T&T 1990 long distance network crashn Server crashes, reboots, sends out "OK" message

Levels of Automation

0: none

1: driver assistance- Small things like help with braking

2: partial automation - More driver assistance; person still in charge

3: conditional automation- Mostly automated, but expect driver to respond in an emergency

4: high automation

5: full automation

Tesla Autopilot and design considerations

May 2016 fatal accident

Level 3 conditional automation

System not fully automated, but driver using it as if it were.

Is it OK to prioritize the safety of occupants over people outside the car?

Who is legally responsible when an accident occurs?

Shrink Wrap Warranties

Consumer (off the shelf) software

Often don't promise that software will work

Limit liability only to purchase price

warranties on goods

a product manufacturer or service provider's documented guarantee of quality as promised to a customer.

Boot example: seller told man boots were waterproof. man wore boots; they were not; he got a fungal infection; Since the boots were not as promised, the seller will be liable for the value of the boots, plus any loss associated with the resulting physical harm.

internal realism

how do we know there is a desk here?

We are all trapped in our subjective perceptions, But our subjective perceptions oddly agree

Assumption: the world exists- Our subjective perceptions agree because the desk exists

Epistemology

study of knowledge

knowledge is justified, true belief

-evidentialism

-reliabilism

-virtue epistemology

epistemology: evidentialism

all beliefs must be supported by evidence

Some beliefs are basic and justified

Nonbasic beliefs are formed by inference from basic ones Can be hard to follow the chain of evidence all the way down to basic beliefs

Reliabilism

A belief is reliable if it is the result of a reliable cognitive process

A reliable process is one that generally leads to correct outcomes

Example: two bird watchers

"Compare a novice bird-watcher and an expert walking together in the woods, seeking out the rare pink-spotted flycatcher. A bird flies by and each person spontaneously forms the belief that there is a pink-spotted flycatcher there. The expert knows this to be true but the novice is jumping to a conclusion out of excitement. The expert has a well-founded belief but the novice does not."

Virtue Epistemology

Knowledge is an achievement

The result of "competent cognitive agency"

Epistemic virtues include:

Curiosity, Intellectual autonomy, Intellectual humility, Attentiveness, open mindedness

social Construction of knowledge

The process of something becoming socially accepted is the process of it becoming a fact

-knowledge is collaborative

peer review

How carefully something is reviewed is a measure of how reliable it is Standard practice for journal articles and peer-reviewed conference articles

three expert reviewers

Wikipedia vs peer review

The more a page gets attention, the more carefully reviewed it is

Journal article: Three expert reviewers, Once published, fixed and can't be updated.

Wikipedia: Non-expert reviewers, Number of reviewers varies Can be updated continuously.

Wikipedia faults

-subject to many biases (selection, participation, group think, gender, racial)

Revenge editing: Inserting false, biased, or defamatory content into articles, especially articles on living people

citogenesis: Journalist looks something up on Wikipedia; Writes article without citing it; Someone trying to improve Wikipedia looks for citations; Finds journalist's article!

John Snow and Cholera

Cholera epidemics in London, mid 1850s

John snow was able to find out the cause of cholera with a good visualization of data (he decided to make a graph of all individuals on a map)

-he noticed these people were mainly centered around the broad street pump

-an outlier proved it was from the broad street pump because lived far away but had servants go to get water from that pump

Challenger Case Study

The rubber o-ring seals used in the rockets are normally flexible and elastic, which allows it to work.

-when it is cold however, they become brittle and shrink

-it was going to be really cold, and the engineers knew this was going to be a problem

-they sent NASA a warning report to urge them to reschedule the launch, but the Report was very hard to read and showed not many signs of credibility

-NASA, was in a lot of pressure to launch on time, and they didn't really understand the report, so they went on with the launch, and the challenger exploded right after takeoff

Algorithmic bias

Implicit bias: we may not be aware that we are biased; introduce math test and describe it as hard: men out perform women; if nothing is said: there is equal performance

search engine bias: ads for exec jobs are displayed less to women; searching for doctors shows mostly men; 3 black teenagers vs 3 white teenagers

AI & the Justice System

parole sentencing recommendations

predictive policing: predicting crimes, offenders, victims

tends to reinforce racist status quo

Bias and Robotics

Military robots can make life or death decisions- Failures of computer vision on dark skin make fatal errors more likely Driverless cars

Medical robots- Medicine requires value judgments- Example: triage

Explainable AI

is artificial intelligence that is programmed to describe its purpose, rationale and decision-making process in a way that can be understood by the average person.

Bias in ML datasets

Reporting bias- Example: book reviews are strongly positive and negative

Automation bias-People believe results from an algorithm more, even if it's not merited

Implicit bias- Assumptions based on your own assumptions that don't hold more generally

confirmation bias- Unconsciously affirm your own beliefs Experimenter's bias- Keep refining what you're building until it gives the answer you expected

selection bias- data selected is not representative

in group bias- bias in favor of people like you

out group bias- stereotyping people not in your group

FISA

Foreign Intelligence Surveillance Act

-Watergate: nixon spied on opponents

-lead to fisa

Without a court order: President (through attorney general) may authorize surveillance for one year, as long as there is "no substantial likelihood" that it will include communications to which a US person is party

With a court order (from FISA court) -If US person is party, need a court order within 72 hours of starting surveillance Must demonstrate probable cause; Target must be affiliated with a foreign power

Patriot Act

This law passed after 9/11

Gives federal government and intelligence agencies more authority to monitor communications

Gives secretary of the treasury greater powers to regulate banks To prevent money laundering

Makes it more difficult for terrorists to enter the US

Defines new crimes and penalties for terrorist activities

-warrantless search

-sneak and peak, tracking online info

NSLs

National Security Letters

Subpoena of personal records without probable cause

Originally aimed at espionage suspects

Now can be used on anyone, even if not suspected of criminal activity

Can prohibit recipient of request for info from revealing this (gag order)

Gag order ruled unconstitutional by lower court in 2007

FBI drops NSL request in that case, rather than appealing

section 215 patriot act

FBI can collect records of business, medical, education, library, and religious institutions

Don't need to demonstrate probable cause Institution can't reveal existence of warrant, or tell anyone what info they gave the government

Brandon Mayfield

A lawyer and convert of Muslim, was held for two weeks before the Spanish investigation revealed that the print instead belonged to an Algerian

-alleged perpetrator in Madrid train bombing

Bush v Gore case study

Hanging chads- Votes not counted if didn't punchall the way through

Butterfly ballots - Bad design

- Lots of votes registered for Buchanan in a strongly democratic county

Difference changed the outcome of the election

Florida database error

Some voters incorrectly identified as felons with no right to vote Also enough to change election result

2020 election case study

Highest voter turnout since 1900

Biden received 81 million votes, most votes ever cast for a presidential candidate

Record number of votes cast early orby mail (because of pandemic)

Republicans claimed widespread voter fraud

Attorney General William Barr: no evidence of widespread fraud or irregularities

Federal agencies - most secure election in American history 63 lawsuits in several states dismissed or withdrawn

Dominion and Smartmatic Voting Systems lawsuits

Promoting election integrity

Paper records or voter verified paper backups

audits (inspection)

Risk-limiting audits (RLAs) should be done by hand

evoting

Pros: accessibility

Richer people have more access

Voter privacy

Same system authenticates voter and records voten

Possible vote selling

Vulnerability to denial of service attacks

Vulnerable to security problems on home PC

Vulnerable to hackers

Experts argue there is no way to make evoting safe

Net Neutrality

The principle that Internet service providers should enable access to all content and applications regardless of the source, and without favoring or blocking particular products or websites

Zero rating

Free net connection, just for this one service

facebook free basics

wikipedia zero

Free Market vs consumer protection on internet access

consumer protection would be in favor of net neutrality, free market would be in favor of companies against net newutrality

Automation

pros: produces items quicker, reduces costs

cons:loss of jobs (manufacturing, pharmacy, law firms, teaching assistants

changing relationship with the workplace

longer work weeks

bringing more work home with emails, cell phones

organizations used to be more hierarchal but are now flatter and more connected

telework: improved with covid. increased productivity, improved child care expenses, less gas emissions, more distractions, difficulty scheduling meetings

The gig economy: service workers who complete short-term jobsfor clients, usually relying on a company to connect them (uber, door dash)

Globalization

pros:

Can increase everyone's standard of living

Each country has a specialty

Increases competition-Better prices for consumers

People in poorer countries deserve jobs too

cons:

American workers shouldn't have to compete with foreign workers who do not receive decent pay and working conditions

Globalization increases the rate of US job loss overseas Foreign workers also hurt by globalization

no unions for foreign workers

Race vs. Ethnicity

Race is understood by most people as a mixture of physical, behavioral andcultural attributes. Ethnicity recognizes differences between people mostly on the basis of language and shared culture

Technologies with racial bias

mage recognition returns a photo of a Black person on a search for gorilla

Apple Watch doesn't work with darker skin

Asians seen as "blinking"

In video games: not much representation, non whites are usually evil

fighting technological racial bias

Hire more diverse designers and developers

Include diverse groups in testing and user research

Be aware of your own privilege