Lesson 1: Stef Aupers – Digitalization and the challenges of AI
Three ways of digitalization:
- Distinction between traditional and digitalized media
50 years: three converging waves of digitalization and their influence on society:
Personal computer
Internet and social media
Artificial Intelligence: recent phenomenon
o AI: a social science perspective
First wave of digitalization: personal computer
The democratic promise of the PC…
IBM computer in the 1950’s => the ‘hacker ethic’ (freedom of information)
Prediction: we don’t need that many computers (10 to 12)
60 – 70s: Computers were only needed for governments
Referred to themselves as hackers because they had insights in computer technology
They were angry, formed a social movement, they didn’t want computers to be
reserved only for the government
“Bringing computers to the people” = goal -> hippies and hackers in Silicon Valley
Strived for democracy and individual freedom
Some people were pioneers in the development of computers
1975: Creation of the first personal computer: the Apple computer
1975 – 1985: mass-production, commercialization, development first personal
computer
Second wave of digitalization: internet and social media
The democratic promise of the internet
Computers were connected before: used in libraries, army,…
Interconnected system of computers can only be possible if we have personal
computers that are connected
Web 1.0: 1990’s: interconnected PC’s, visit website, not much interaction
Web 2.0: 2000’s: social media platforms (Facebook as pioneer) and User Generated
Content (USG) => very democratic idea
Time Magazine (2008) – Person of the Year: YOU
Internet = giant library of information, available for everyone
Internet and social media
Utopian social political climate where you can access the information you want
FaceBook -> blueprint social infrastructure internet
« Making the world more open and connected” (Mark Zuckerbuck)
From democratization to ‘‘surveillance capitalism”: our data is being stored, sold,
saved,…
Third wave of digitalization – Artificial Intelligence
, Very recent phenomenon in our everyday lives
John McCarthy -> mathematician / scientist
Introduction concept DartMouth Conference 1955: modest conference on what they
then first called Artificial Intelligence:
“… making a machine behave in ways that would be called intelligent if a human
were so behaving”
AI promises to make machine smarter in a cognitive sense = the essence of AI
Basic forms of AI:
Weak AI: AI that imitates our cognitive functioning but can only do one specific
function action (chess game lost against AI), only one function of our brains
Top Down AI: All instructions, you tell the computer to do certain instructions, no
interpretating, no learning
Strong AI: Variety of functions that we humans can do (ex. Social robots that speak,
can make art, have lots of information,…)
Bottom up AI: AI that is self learning, ‘evolutionary computing’, incorporating
knowledge
Now more Bottom up and strong AI than before (= holy grail)
The classical philosophical debate on AI:
Alan Turing VS John Searle:
When can we say a computers has, just like us, consciousness?
Different views
Is AI really intelligent? The Turing Test – Alan Turing
How can AI be considered intelligent? And if so, how can we decide that?
Computer/machine vs. human: we don’t know which one is the human or computer
If the conversation goes on for 8 minutes and we still don’t know who is which then
we can say we are fooled and the machine is actually intelligent
Is AI really intelligent? The Chinese Room experiment – John Searle
Opposite to Alan Turing: computers will never be intelligent
Somebody can give input, instructions letters, etc. who doesn’t know anything about
Chinese and make them write in Chinese and understand the language
The person in the room has instructions and know what to do, knows the syntax but
he doesn’t know anything about the meaning of these texts
If we give the right instructions the computers still won’t know what everything
means
Intelligence is about finding meaning in what you make and not just reproducing
things
Kevin Warwick: A third position in the debate
Argument that goes beyond the human-centric philosophical comparative analysis
that informs Searle and Turing
Both Turing and Searle are both wrong because they think about intelligence as
something only reserved for humans
Relativistic perspective on “intelligence”
, “Computers may well understand things in a different way to humans: animals
probably understand things in different ways to humans; some humans probably
understand some things in different ways to other humans. This doesn’t make one
intelligent and another not. It merely means that one is intelligent in a different way
to another”
Beyond philosophy of AI…
AI in society and everyday life
Outsourcing human labor to AI… control (ex. machines in factories, ChatGPT)
Communicating with AI, chatbots, social robots
AI: challenges for the social sciences
Communication Science
Traditional view: Humans communicate with humans through a medium (film, radio,
internet, social media,…)
AI challenge: humans communicate with the medium (chatbots, social robots,
personal agents, NPCs)
Peter & Kuhnz
“Empirical research within the computers-are-social-actors paradigm has solidly
demonstrated that humans being treat computers, and media more generally, as
social actors and eventually as if they were human”
Lindgren & Holmström
“technological and human actors must be seen as actors on equal terms”
AI: challenges for the social sciences
Bruno Latour:
Sociology has a bias: we are social animals and have social relationships with others
Traditional perspective: social relations between people – social networks consists of
people
New perspective: social relations between people and ‘things’; subjects and objects /
hybrid networks (ex. professor has a relationship with microphone)
Actor network theory -> objects are ‘actants’ and have, just like human actors,
influence on humans, social relations, organizations and (other) things
Social Network Analysis Actor-Network Theory: we also have connections with
items (ex. NPCs in digital games)
The Internet of Things
The Smartphone ‘talks’ to refrigerator, television, vacuum cleaner,… : (ex. if the
refrigerator is connected to your smartphone, it can tell you that you’re out of milk
and can tell the supermarket,…)
Relationships between humans and machines machine…
*objects have agency / being “social actors”
Ex. self-learning algorithm, smart cars and social robots
Her (2006) – relationships between humans and machines: science fiction or science
faction?
Coanda en Aupers (2021): Case-study Non Playing Characters in video games:
, o Players humanize the NPC
o Players have romantic-erotic relations with the NPC
o Impact on ‘human’ social relations of players
Beyond the anthropocentric paradigm:
o From humanism to ‘post-humanism’
Social robots
Human enhancement: humans become post-human entities
Article 1: A social science perspective on artificial intelligence: building blocks
for a research agenda
1. Towards a truly social science of AI
People have been grappling with the consequences of technology for centuries
o E.g. overpasses built so that people who rely on public transport couldn’t get
through => passage = reserved for elites and the white upper class
o Technologies are political ! => demand for a social science perspective
AI studies => studied by the same scientist who are engaged in creating the AI agents
themselves
o Interdisciplinary approach = advisable (“AI’s social sciences deficit”)
o Some social science research has shown that AI doesn’t make everybody’s life
easier or safer => can increase inequality, discrimination, and inflict harm
based on gender, race and class
1.1. Humans and machines in context
Need for a view where machines and humans constantly construct and reconstruct
the social word through dynamic interaction
Communication between humans and machines = socio-cultural rather than
technological process
o Conceiving AI agents as communicative agents
o Interaction and communication can no longer be seen as a human-only
process
AI = coded by humans thus encoded with human intentions
o Embody social values => makes them human dependent
‘cyborg’ perspective: past the increasing border between human and machine
1.2. AI agents as social actors
Social science of AI needs to approach human/AI relationship as complex and
multidimensional
AI => ongoing relationship with a surrounding social world
o Must study agents alongside other agents in their social and communicative
context
Actor-Network-Theory: agency of human and non-human agents is seen as equal
o Analytical approach that wants to move beyond anthropological, human-
centered, bias of traditional sociology
Social relations can emerge between all different kinds of entities (action of one
entity has an effect on the actions of another)
Three ways of digitalization:
- Distinction between traditional and digitalized media
50 years: three converging waves of digitalization and their influence on society:
Personal computer
Internet and social media
Artificial Intelligence: recent phenomenon
o AI: a social science perspective
First wave of digitalization: personal computer
The democratic promise of the PC…
IBM computer in the 1950’s => the ‘hacker ethic’ (freedom of information)
Prediction: we don’t need that many computers (10 to 12)
60 – 70s: Computers were only needed for governments
Referred to themselves as hackers because they had insights in computer technology
They were angry, formed a social movement, they didn’t want computers to be
reserved only for the government
“Bringing computers to the people” = goal -> hippies and hackers in Silicon Valley
Strived for democracy and individual freedom
Some people were pioneers in the development of computers
1975: Creation of the first personal computer: the Apple computer
1975 – 1985: mass-production, commercialization, development first personal
computer
Second wave of digitalization: internet and social media
The democratic promise of the internet
Computers were connected before: used in libraries, army,…
Interconnected system of computers can only be possible if we have personal
computers that are connected
Web 1.0: 1990’s: interconnected PC’s, visit website, not much interaction
Web 2.0: 2000’s: social media platforms (Facebook as pioneer) and User Generated
Content (USG) => very democratic idea
Time Magazine (2008) – Person of the Year: YOU
Internet = giant library of information, available for everyone
Internet and social media
Utopian social political climate where you can access the information you want
FaceBook -> blueprint social infrastructure internet
« Making the world more open and connected” (Mark Zuckerbuck)
From democratization to ‘‘surveillance capitalism”: our data is being stored, sold,
saved,…
Third wave of digitalization – Artificial Intelligence
, Very recent phenomenon in our everyday lives
John McCarthy -> mathematician / scientist
Introduction concept DartMouth Conference 1955: modest conference on what they
then first called Artificial Intelligence:
“… making a machine behave in ways that would be called intelligent if a human
were so behaving”
AI promises to make machine smarter in a cognitive sense = the essence of AI
Basic forms of AI:
Weak AI: AI that imitates our cognitive functioning but can only do one specific
function action (chess game lost against AI), only one function of our brains
Top Down AI: All instructions, you tell the computer to do certain instructions, no
interpretating, no learning
Strong AI: Variety of functions that we humans can do (ex. Social robots that speak,
can make art, have lots of information,…)
Bottom up AI: AI that is self learning, ‘evolutionary computing’, incorporating
knowledge
Now more Bottom up and strong AI than before (= holy grail)
The classical philosophical debate on AI:
Alan Turing VS John Searle:
When can we say a computers has, just like us, consciousness?
Different views
Is AI really intelligent? The Turing Test – Alan Turing
How can AI be considered intelligent? And if so, how can we decide that?
Computer/machine vs. human: we don’t know which one is the human or computer
If the conversation goes on for 8 minutes and we still don’t know who is which then
we can say we are fooled and the machine is actually intelligent
Is AI really intelligent? The Chinese Room experiment – John Searle
Opposite to Alan Turing: computers will never be intelligent
Somebody can give input, instructions letters, etc. who doesn’t know anything about
Chinese and make them write in Chinese and understand the language
The person in the room has instructions and know what to do, knows the syntax but
he doesn’t know anything about the meaning of these texts
If we give the right instructions the computers still won’t know what everything
means
Intelligence is about finding meaning in what you make and not just reproducing
things
Kevin Warwick: A third position in the debate
Argument that goes beyond the human-centric philosophical comparative analysis
that informs Searle and Turing
Both Turing and Searle are both wrong because they think about intelligence as
something only reserved for humans
Relativistic perspective on “intelligence”
, “Computers may well understand things in a different way to humans: animals
probably understand things in different ways to humans; some humans probably
understand some things in different ways to other humans. This doesn’t make one
intelligent and another not. It merely means that one is intelligent in a different way
to another”
Beyond philosophy of AI…
AI in society and everyday life
Outsourcing human labor to AI… control (ex. machines in factories, ChatGPT)
Communicating with AI, chatbots, social robots
AI: challenges for the social sciences
Communication Science
Traditional view: Humans communicate with humans through a medium (film, radio,
internet, social media,…)
AI challenge: humans communicate with the medium (chatbots, social robots,
personal agents, NPCs)
Peter & Kuhnz
“Empirical research within the computers-are-social-actors paradigm has solidly
demonstrated that humans being treat computers, and media more generally, as
social actors and eventually as if they were human”
Lindgren & Holmström
“technological and human actors must be seen as actors on equal terms”
AI: challenges for the social sciences
Bruno Latour:
Sociology has a bias: we are social animals and have social relationships with others
Traditional perspective: social relations between people – social networks consists of
people
New perspective: social relations between people and ‘things’; subjects and objects /
hybrid networks (ex. professor has a relationship with microphone)
Actor network theory -> objects are ‘actants’ and have, just like human actors,
influence on humans, social relations, organizations and (other) things
Social Network Analysis Actor-Network Theory: we also have connections with
items (ex. NPCs in digital games)
The Internet of Things
The Smartphone ‘talks’ to refrigerator, television, vacuum cleaner,… : (ex. if the
refrigerator is connected to your smartphone, it can tell you that you’re out of milk
and can tell the supermarket,…)
Relationships between humans and machines machine…
*objects have agency / being “social actors”
Ex. self-learning algorithm, smart cars and social robots
Her (2006) – relationships between humans and machines: science fiction or science
faction?
Coanda en Aupers (2021): Case-study Non Playing Characters in video games:
, o Players humanize the NPC
o Players have romantic-erotic relations with the NPC
o Impact on ‘human’ social relations of players
Beyond the anthropocentric paradigm:
o From humanism to ‘post-humanism’
Social robots
Human enhancement: humans become post-human entities
Article 1: A social science perspective on artificial intelligence: building blocks
for a research agenda
1. Towards a truly social science of AI
People have been grappling with the consequences of technology for centuries
o E.g. overpasses built so that people who rely on public transport couldn’t get
through => passage = reserved for elites and the white upper class
o Technologies are political ! => demand for a social science perspective
AI studies => studied by the same scientist who are engaged in creating the AI agents
themselves
o Interdisciplinary approach = advisable (“AI’s social sciences deficit”)
o Some social science research has shown that AI doesn’t make everybody’s life
easier or safer => can increase inequality, discrimination, and inflict harm
based on gender, race and class
1.1. Humans and machines in context
Need for a view where machines and humans constantly construct and reconstruct
the social word through dynamic interaction
Communication between humans and machines = socio-cultural rather than
technological process
o Conceiving AI agents as communicative agents
o Interaction and communication can no longer be seen as a human-only
process
AI = coded by humans thus encoded with human intentions
o Embody social values => makes them human dependent
‘cyborg’ perspective: past the increasing border between human and machine
1.2. AI agents as social actors
Social science of AI needs to approach human/AI relationship as complex and
multidimensional
AI => ongoing relationship with a surrounding social world
o Must study agents alongside other agents in their social and communicative
context
Actor-Network-Theory: agency of human and non-human agents is seen as equal
o Analytical approach that wants to move beyond anthropological, human-
centered, bias of traditional sociology
Social relations can emerge between all different kinds of entities (action of one
entity has an effect on the actions of another)
