Intellectual Property
Copyright
protects creative works
Trademark
brand name
Patents
protects idea
Biases
Experimenter’s Bias
Automation Bias
tendency to believe automation more
Reporting Bias
people are more likely to look for data that supports their hypothesis, so more positive finding data
Coverage Bias
not representing properly (only white people or smtn)
Non-response Bias
like not responding to a survey, some groups of people may be more likely to do it
Sampling Bias
when some people are more likely to be sampled than others (no longer random / representative)
Confused on difference with coverage
In-group Bias
bias towards people in your group
Out-group homogeneity bias
stereotype people outside group / treat all of them as one or large groups rather than unique individuals
Confirmation Bias
Bias Application
Bias in justice system
AI would enforce current racist standards, especially through attempts at predictive policing
Bias in ML
bad training data → bad outputs biased training data → biased output Military robots have power over life and death Health robots need value judgement
Net Neutrality
Internet service providers must allow access to the internet as a whole
Why?
Ensure fair competition Equal access No discrimination Freedom of expression
We are already not neutral
large companies can be faster already, have advantages, etc
Limits competition but reflects global capitalism
Competition as a solution (free market)
Consumers can switch but it can be time consuming or expensive
Consumer protection approach
Protect consumer free speech, prevent discrimination
Zero-rating
No internet service costs for some services by some providers Threatens net neutrality because it incentivizes the use of some apps instead of others to reduce costs