DATA 104 MIDTERM 1
Agency - ANS-Agency is the ability or capacity to act or exert power. Technology informs the
way in which people both perceive and exercise their capacity to exert some degree of
control over the sociotechnical relations in which they are enmeshed.
Exemplary cases:
- Algorithmic decision-making (COMPAS, Gig economy laborers)
- Real-time interplay of human and mechanical agency (autonomous vehicles, pilots)
- Patients tracking their health through data apps and devices
Google Walkout
Disruptive innovation and the tech entrepreneur agent
Questions to ask with this tool:
How is human agency delegated to a technology?
How does the delegation affect responsibility?
What kinds of agency are amplified and what kind is constrained by the technologyIn what
ways?
What kinds of social structures does the specific form of agency aim to challenge?
\Alternatives - ANS-It once was different. ... It's different somewhere else.
Cases we've seen:
Population thinking and statistics aren't automatic. They don't emerge in contexts where
states aren't trying to govern their resources (human and others).
The notion of "normality" sits at the core of early population studies and statistics -- which we
only notice because today we give much more attention to the long tails and the outliers.
The seemingly inevitable aspects of Silicon Valley growth didn't always hold and don't seem
apply everywhere. (In fact, they don't even get recognized and described until the 1990s.)
Self-tracking in Ben Franklin's day may look similar to self-tracking today -- but different
motives may underlie its practice.
Questions to ask to use this pointer:
Why were things different then or there? What does that comparison tell us about here and
now?
What might make it change? Change again?
How does this call into question notions of what's natural, inevitable, or automatic?
\Classification - ANS-Implicit and explicit social organization of beings and knowledge into
discrete categories governed by identifiable principles. Societies produce knowledge and do
work (e.g. with technologies) by sorting, ordering, and classifying phenomena in the world.
Classification systems inform social order and vice versa.
Exemplary cases:
- Bettina Wulff and Google autocomplete
- Risk and vulnerability assessment algorithms
- Content moderation; Designing ML for identifying and eliminating hate speech in social
media
Questions to ask with this tool:
How does the technology depend upon classification in order to work?
How does the technology classify? What/who does it classify?
, What are the categories of classification? What's the process used to classify? (e.g. Does it
count as the thing we're interested in? Do we count it? How?)
How were the categories determined/decided upon? By whom? With what purpose? Can
they be revised?
\Co-Production - ANS-Skill or knowledge in a domain. Technical expertise is usually
institutionalized and is variously valorized in societies (e.g. associated with establishment of
facts, trust, and authority). Technical experts often wield particular kinds of social power.
Exemplary cases:
- COMPAS algorithm and the expertise of the judge
- Value of transparency in algorithmic systems and necessary expertise to be able to engage
it
- Cumbrian sheep farmers
- Climate science
Risk management; work of regulatory agencies
Tech workplace, expertise and gender
Questions to ask with this tool:
Which aspects of expertise are shifting with technology? Which are staying the same?
Who is thought to be an expert? What is the social role of their expertise? Relationship to
"lay" or non-expert?
What is the role of the expert in democratic institutions? How can expertise and democracy
come into conflict with one another?
What are the skills socially-deemed necessary to become an expert?
What are the conditions of expertise necessary to participate in the public sphere?
What qualities of expertise are valorized (e.g. objectivity, compassion, leadership, etc.)?
\Contigency - ANS-It could have happened differently. ... It's not controllable or predictable.
Cases we've seen:
Today's approaches to responsible conduct of research emerged a context of biomedical
experimentation -- and seem to be showing their limits in the big data era.
"Choice architecture" and "nudge theory" emerged in a particular context -- the industrialized
Anglophone West in the 1990s.
An individual decision by Edward Snowden opens the floodgates on state surveillance. An
individual decision by Chris Wylie breaks open Cambridge Analytica, and an avalanche of
FB critique follows.
Questions to ask to use this pointer:
Did an individual's actions change the course of history? How do individual actions get
amplified by systematic dynamics ("floodgates," "avalanche")?
Is there a reason this particular thing happened at this moment? Is it really by chance?
If something originated in contingency, how might it need to be changed for different
circumstances?
\Dynamics - ANS-What forces are driving? ... What structures are shaping?
Cases we've seen:
Data science emerges as a new science in the force field between industry and academia, at
a moment when industry needs are pushing ahead of institutions of higher education.
Silicon Valley moves in waves, with financial gains from one era's technology seeding the
next.
Agency - ANS-Agency is the ability or capacity to act or exert power. Technology informs the
way in which people both perceive and exercise their capacity to exert some degree of
control over the sociotechnical relations in which they are enmeshed.
Exemplary cases:
- Algorithmic decision-making (COMPAS, Gig economy laborers)
- Real-time interplay of human and mechanical agency (autonomous vehicles, pilots)
- Patients tracking their health through data apps and devices
Google Walkout
Disruptive innovation and the tech entrepreneur agent
Questions to ask with this tool:
How is human agency delegated to a technology?
How does the delegation affect responsibility?
What kinds of agency are amplified and what kind is constrained by the technologyIn what
ways?
What kinds of social structures does the specific form of agency aim to challenge?
\Alternatives - ANS-It once was different. ... It's different somewhere else.
Cases we've seen:
Population thinking and statistics aren't automatic. They don't emerge in contexts where
states aren't trying to govern their resources (human and others).
The notion of "normality" sits at the core of early population studies and statistics -- which we
only notice because today we give much more attention to the long tails and the outliers.
The seemingly inevitable aspects of Silicon Valley growth didn't always hold and don't seem
apply everywhere. (In fact, they don't even get recognized and described until the 1990s.)
Self-tracking in Ben Franklin's day may look similar to self-tracking today -- but different
motives may underlie its practice.
Questions to ask to use this pointer:
Why were things different then or there? What does that comparison tell us about here and
now?
What might make it change? Change again?
How does this call into question notions of what's natural, inevitable, or automatic?
\Classification - ANS-Implicit and explicit social organization of beings and knowledge into
discrete categories governed by identifiable principles. Societies produce knowledge and do
work (e.g. with technologies) by sorting, ordering, and classifying phenomena in the world.
Classification systems inform social order and vice versa.
Exemplary cases:
- Bettina Wulff and Google autocomplete
- Risk and vulnerability assessment algorithms
- Content moderation; Designing ML for identifying and eliminating hate speech in social
media
Questions to ask with this tool:
How does the technology depend upon classification in order to work?
How does the technology classify? What/who does it classify?
, What are the categories of classification? What's the process used to classify? (e.g. Does it
count as the thing we're interested in? Do we count it? How?)
How were the categories determined/decided upon? By whom? With what purpose? Can
they be revised?
\Co-Production - ANS-Skill or knowledge in a domain. Technical expertise is usually
institutionalized and is variously valorized in societies (e.g. associated with establishment of
facts, trust, and authority). Technical experts often wield particular kinds of social power.
Exemplary cases:
- COMPAS algorithm and the expertise of the judge
- Value of transparency in algorithmic systems and necessary expertise to be able to engage
it
- Cumbrian sheep farmers
- Climate science
Risk management; work of regulatory agencies
Tech workplace, expertise and gender
Questions to ask with this tool:
Which aspects of expertise are shifting with technology? Which are staying the same?
Who is thought to be an expert? What is the social role of their expertise? Relationship to
"lay" or non-expert?
What is the role of the expert in democratic institutions? How can expertise and democracy
come into conflict with one another?
What are the skills socially-deemed necessary to become an expert?
What are the conditions of expertise necessary to participate in the public sphere?
What qualities of expertise are valorized (e.g. objectivity, compassion, leadership, etc.)?
\Contigency - ANS-It could have happened differently. ... It's not controllable or predictable.
Cases we've seen:
Today's approaches to responsible conduct of research emerged a context of biomedical
experimentation -- and seem to be showing their limits in the big data era.
"Choice architecture" and "nudge theory" emerged in a particular context -- the industrialized
Anglophone West in the 1990s.
An individual decision by Edward Snowden opens the floodgates on state surveillance. An
individual decision by Chris Wylie breaks open Cambridge Analytica, and an avalanche of
FB critique follows.
Questions to ask to use this pointer:
Did an individual's actions change the course of history? How do individual actions get
amplified by systematic dynamics ("floodgates," "avalanche")?
Is there a reason this particular thing happened at this moment? Is it really by chance?
If something originated in contingency, how might it need to be changed for different
circumstances?
\Dynamics - ANS-What forces are driving? ... What structures are shaping?
Cases we've seen:
Data science emerges as a new science in the force field between industry and academia, at
a moment when industry needs are pushing ahead of institutions of higher education.
Silicon Valley moves in waves, with financial gains from one era's technology seeding the
next.
