CMN170V Final

Big data machine learning

"Big" doesn't need to know why: Machine Learning (ML) Ex: Translation Algorithms having a hard time translating languages Old way of translating: teaching the computers grammatical rules and vocabulary Google uses equivalent documents that are on the internet to analyze to find patterns and translate it 🡪 highly effective Machine learning algorithms trumps personality judgement

Social network analysis (Network Structure)

"Small World Phenomena": (1) small average path length ~ 4.7 degrees of separation six degrees of separate: the theory that everyone and everything is six or fewer nodes/ steps away Experiment: Letter U.S.: 5.5 5 degrees of Kevin Bacon Homophily: people are who similar hang together Social network metrics

e-social science

"computational social science" Simulations of downtown Los Angeles Behavior of where people will flee during an emergency Portland light-rail industry Simulated 1.6 million residents of Portland to optimize light-rail infrastructure

Technology determinism vs. social construction of technology

**Social online networks can be used to invade the privacy sphere of people. This statement is an example of: social construction Private sector and public sector are responsible for the development of technology. Mostly public sector funding Too risky for private sector; they'll do research and inventions afterwards and bring it to the market **What happened around 1925 to the number of beer brewing firms in the U.S.? Public sector intervention left its signature om the course of history during these years

USA Number 1!...?

**where in the field do you think the U.S. is in terms of ranking countries by their freedom of the press? 🡪 top 30% US doesn't have diverse media landscape We have media with homogeneous messages

Media concentration

- Horizontal ownership: competitors merging - Vertical ownership: buy a previous supplier or client - Cross/Circular ownership Execute production chain until the sell the product to consumers. Huge conglomerates - can work two ways: Previous competitors can join (horizontal ownership) Suppliers and providers can join (vertical ownership)

Digital footprint

- Information is always physical - it needs a medium to be realized - Information is inevitable - You have to actively tell it to be deleted - Very difficult to delete your own digital footprint // sometimes impossible - You don't own your digital footprint

Timeless time

- Professor Castells - The social practice that aims at negating sequence to install ourselves in perennial simultaneity and simultaneous ubiquity --processing information is a synchronous (in real- time // live ) and asynchronous matter (a post that you can read at any time) ** refers to the fact that : the combination of synchronous and asynchronous information processing allows to negate sequence

Network Structure

- internet allows different digital data to connect - How computers connect to send information - When computers used to have phones and literally call each other - WWW Using the internet to access the world wide web -Personal networks, professional networks, etc. the Linked in Maps examples

Poly-directionality

- one to one - one to many (printing, radio, TV, speech) - many to one (elections, applause) - many to many (meetings, gatherings) - you can choose which one best fits your communication need

Death of distance

- the development of time zones because before information couldn't travel faster than the sun goes around the earth - distance no longer causes an issue for communication

What are the characteristics of digitalization?

-Digital footprint -Timeless Time -Death of Distance -Poly-directionality -Network structure -Space of flows -Network Externalities - Economies of scale -Media richness selection -Exposure selection -Algorithmification: A.I

Millenium development goals

1. Eradicate extreme poverty and hunger 2. Achieve universal primary education 3. Promote gender equality and empower women 4. Reduce child mortality 5. Improve maternal health 6. Combat HIV/AIDS and other diseases 7. 8. Global partnership for development

Layers of the cube

1st Layer of Cube: Access: What kind of technology? 🡪 infrastructure and generic services Usage: which characteristics? 🡪 capabilities, skills, culture 2nd Layer of Cube: How? Effective adoption/ impact: who? 🡪 e-government, e-health, e-education 3rd Layer of Cube: Policy to narrow the digital divide People focus on the communication aspect of the digital divide. **Most studies of the digital divide do not track the number of storage- or computational devices but focus on telecom exclusively. Why is this ever more justified? Ever more storage- and computational capacity is taking place "in the cloud" and telecom is the portal to access this centralized capacity

Global Left of Right Wing

2007 international myth buster Canada, a socialist country, spent more money on the public sector than the US under George W. Bush, a conservative president Truth: Canada actually only spent 20% of their expenditures on the public sector and the US spend 22% Right wing European governments were spending up to 40% of expenditures on public sector Left wing European socialist/communist governments were spending 20-25% of expenditures on public sector, like the US 2012 international myth buster Truth: US policy stayed generally the same Other socialist/communist countries spent a lot more money on the private sector (40%) to decrease inequality other left wing countries decreased expenditure always look at the numbers! There is not a clear correlation between left/right wing and expenditure on the public sector **Who has bigger governments (as relative % share of the total economy)? Bigger; richer countries have governments that are twice the size of governments in poor countries

Income and happiness

According to the graph, there is a positive correlation between income and happiness. Happiness is spread out for low income countries and happiness is very similar (high) for high income countries. The trend line is vertical and creates a 90-degree angle and goes straight across Income isn't the best income to measure happiness

Aid

Aid The Monterrey Consensus (2002): 0.7% of GDP Rich countries would give 0.7% of their GDP to official development projects Only 4-5 countries actually fulfilling agreement **With regard to global Official Development Assistance ("Aid"), developing countries have given significantly less than they promised Money comes back to rich countries with interest rates. ** The problem with Official Development Assistance (ODA = "international Aid") is that developing countries pay several times more money as they receive in debt services to the same countries they receive aid from

Human-machine merger

Algorithms are allowing computers to make decisions Call center, flight check-in, travel booking, shopping counter Online dating

Big data digital footprint

Allows us to observe social micro pattern with bird's eye view on society **in some less developed African countries, how many children do you think have a birth certificate? Less than half. Therefore, we don't know they exist, unless they are walking around with some type of mobile device! Google Maps and Location history Can track your movement Human mobility traces are very unique Helps us learn a lot about individuals and their behavior patterns

Uncertainty of trajectory: exponential technological process

At each doubling step, you are making as much progress that you have since the beginning Exponential growth: 1,2,4,8,16, 32 Social change is a little slower than technological change Technological singularity: technology rivals and complements human intelligence High level of uncertainty **Exponential technological progress complicates the design and implementation of policies and strategies because: it questions the power of even short-term predictions

When is access cheap enough to connect the world?

Basic global statistics ** 40% of the world population lives with less than US$ 60 per month (US$ 2 per day) Average costs: $80 per month on mobile phones $40 per month on broadband $50 for computers PEOPLE SPEND BETWEEN 5-10% OF THEIR INCOME ON ICT **Considering global income levels and average ICT expenditure % of available income, the poorest 20 % of the world have as much available to spend on ICT per month: LESS THAN 5 USD

How digitalization impacts the empirical and theoretical parts of science

Big data: based on what happened in the past Theoretical Computer simulations What could happen in the future in theory (not 100% sure/ under the assumption it's similar to the past's history

Scale through individualization

Bottom-up scale: big data modularization of products Learning relationships (Netflix, Amazon, supermarket, etc.) Top down scale: versioning of products Tailor- made (newspapers, phone apps, etc.) Quality and quantity Time (real-time, asynchrony)

Unpredictability of side effects: creative destruction 🡪 private- public sector alliance

Broad input, but specialized input by experts Mix between private and public sector since they are both involved private sector is often involved with the development of technology, so they have a better idea of where technology is headed the public sector know where it normatively wants to guide the story working together can reduce the uncertainty **is there a difference between "international", "transnational", "multinational", and "supranational"? Yes, each different prefix means something different.

Why Policies and Strategies can never be 100% effective

Can policy suggestions be 100% effective? No. All models are wrong, even if they are useful. The only working model of the universe is the universe itself. Can't take EVERY LITTLE THING and detail into consideration in your model. They are always a probabilistic gamble **No model is ever able to make completely accurate predictions because we are never able to obtain and to model all the details involved in reality When we can't know for sure, we might as well rely on what's probable. This doesn't mean that the "not probable" won't happen. We can only try to reduce the uncertainty that we are facing.

All pervasiveness: general purpose technology

Can use technology for many different reasons So it's difficult to make policy for it since it's so broad and has so many uses Who is responsible for a "general purpose technology"? Who leads this policy agenda is important ICT budget dedicates 15% of budget to broadband infrastructure Decentralized policy planning and budgeting **The fact that ICT is a general-purpose technology leads to the fact that it is positive that the budget dedicated to public sector ICT policies: highly centralized

Dangers of big data

Characteristics of Big Data: 1. digital footprint 2. n=N 🡪 volume! 3. Data-fusion: big data footprint is always incomplete, so we always try to fill holes with a big variety of data 🡪 variety! 4. in real time 🡪 velocity! 5. Machine learning

Simulations for quantitative context-dependency

Common critique of simulations: "they're just simulations" You need mathematical formulas and codes to predict behaviors Mathematical equations versus codes **Computer simulations are NOT less powerful or less insightful than formal mathematical models (such as analytical differential equations) even though they are not quantitative.

Data fusion

Complementing unstructured messy data Images, text, video, audio Fill the holes of missing data to make conclusions **Data fusion makes use of complementary data sources to fill out missing data points, including different kinds of sources, such as images, audio, and video. Decentralizing and analyzing different data and bringing them together. Split them up and process them in parallels and bring them back together again 🡪 data fusion

Why do we need computer simulations

Computer simulations about what could happen in the future Data (from the past) has problems with changing futures Data 50 million most common search terms to predict the spread of the seasonal flu Models 450 million mathematical correlation models on search terms and flu outbreaks Identification of 45 search terms to predict flu outbreaks better than traditional models Data takes too long to be analyzed sometimes to make accurate predictions about the future

How companies can classify consumers

Consumers' financial vulnerability: Social influencer Rural and barely making it Ethnic second-city strugglers Retiring on empty: singles Tough start: Young single parents Credit crunched: city families Military personnel Elderly opportunity seekers "Oldies but goodies"

Mass customization

Cost leadership: same product by very cheap Differentiation strategy: customized products, more expensive Digitalization allows you to obtain individualization through scale. Data mining and big data (Amazon, Netflix, etc.) Scale through individualization

Unpredictability of side effects: creative destruction

Creates and destroys Impossible to predict all of the side effects People didn't want cars on the street because the cars would disturb the horses on the streets 🡪 but then people thought it'd be a good idea to use cars to get rid of the horse poop on the road 🡪 etc.

Internationality of digital networks: space of flows

Cyberspace does not recognize international borders So how do we make a policy agenda for it? Global policy agenda? Internet is governed by a private sector non-profit corporation in southern California : ICANN Internet corporation for Assigned Names and Number Some things are better being governed on a regional level, rather than on a global level to create more specific digital action plans (can even go to national level, municipal level, companies, etc.) Policies on the bigger levels influence policies on the smaller levels **The organization that governs the global internet is : a private sector, non- profit organization

Real time Big Data

Data footprint is often available in real time Algorithms listen into phone conversation to listen to your personality in real-time Listened to the kind of works you used and how they came out of your mouth Adapted this technology to match you with like-minded people Results for company: Customer service doubles from 47% to 92% Call average goes from 10 minutes to 5 minutes Stores this data for future inquires to immediately match you

Diffusion of Innovations through Social Networks

Diffusion of Innovation Knowledge stage Persuasion stage Decision stage Implementation stage Confirmation stage (recognize benefits, use it in your daily life, and promote it) ** The Communication Scholar Everett Rogers described the process of "Diffusion of Innovations" in different stages: knowledge, persuasion, decision implementation, and confirmation

Economies of scale

Digital products almost have infinite economies of Scale. Good thing for time and money. **In digital costs, most cost are related to fixed costs.

How to define the digital divide

Divides are inevitable since it's a diffusion process through social networks. 4 Perspective to Analyze the Digital Divide 1. What kind of technology (phones, internet, broadband, storage devices, etc.) 2. Who is the subject (country, organization, individual, etc.) 3. Which attributes matter (income, capacity, type of ownership, geography, etc.) 4. How are they connected (access, usage, effective adaption) * We said that we can take four complementary perspectives on the digital divide. The following is not one of them: "when does the subject connect"

Global context of trade and industrialism in terms of "flat earth"

Economically, the world is not flat. Although free trade brings a lot of benefit, it can also hurt domestic producers who are forced to compete with lower foreign prices. It is hard for infant industries to enter the markets.

Computational Social Science

Embracing the effects digitalization has on empirical work and theoretical work e-social science **Digitalization affects both aspects of science: Empirical data work (observations of the real-world) and theoretical modelling (hypothetical constructs) 🡪 TRUE Time to model all life on earth to get a complete picture of the ecosystem Obtain data to parameterize is the hardest part Motion activated cameras Validate model Humans patterns are tracked with mobile devices, social networks, food bought with credit cards We a lot of data to analyze in the social sciences, more than biologists, etc. Draws scientists to social sciences to explore digital footprint

Intro to e-science

Empirical: observe the world as it is Collect data about the world and then analyze the data/ observation in order to obtain some more fundamental truth 🡪 inductive reasoning Charles Darwin with natural selection Theoretical: start from axioms and ideas/ logical conclusions Albert Einstein with the speed of light 🡪 deductive reasoning

Algorithmification

Epitome of digital divergence: storage, communication, computation Algorithmification decides what type of information is stored, communicated, and computed

What is Big data

Five characteristics of Big Data that are most often present: 1. Digital footprint (produced anyway for free) 2. n=N (no sampling, but potential bias) | n= sample | N= the universe | now we don't need to sample 3. Data fusions is unconstructed and incomplete a. using complementary data sources to make up for missing pieces 4. Available in real time (dynamic) 5. Machine learning algorithms are now really affective (no need for theory)

Space of flows

Geographical space doesn't matter in the digital world, network space matters. Space of Flow is a combination of: Death of Distance Poly-directionality Network Structure Space is defined by where you are in the network and who you communicate with in the network.

Global historical context of "flat earth"

Global Flattener and Unflattener ** We will look at the global context in which digitalization takes place from two perspectives: Large- scale systemic & smaller-scale activist The Big (macro) systemic level: Historical / Colonialism How it all started Started equally in Africa 🡪 spread out all over the globe and different cultures and technologies and resources were developed and found 🡪 continued to explore and invent and find new and better trade routes, etc. : colonialism Lead to inequality, people used different means for living, etc. ** A "human slave" nowadays is: more than 100 times cheaper today than 150 years ago when "slavery" was formally abolished Historically, the world is not flat because we all started from one place but then dispersed into different cultures, ways of living, and opportunities (colonialism).

Global Flatteners and Unflatteners

Global Flatteners Free trade Digital revolution / information MDG's philanthropists Unflatteners Social mobility- not free to work or live in any other country Taking a vacation is different Not a free labor market (Adam Smith) Severe inequality Partial global governance ** The flow of workers HAS NOT been globalized

Internationality of digital networks: space of flows 🡪 international coordination

Governments need to coordinate Different e-government solutions are interoperable with other e-governments

Many definitions of the digital divide

How digital innovation are diffused through social networks ** The reason why some people say that the digital divide is closing, and others say that it is widening might be because There are many different perspectives. **Following our framework on the digital divide, we can look at it in terms of WHO, with WHICH characteristics, connects HOW to WHAT

From Digital Access to Real Impact

How to connect? How sophisticated is ICT? Different kinds of access At home, in public, at work, internet access facilities, etc. **In contrary to developed countries, in many developing countries many access the internet through public access centers Effective adoption = impact **Please distinguish among three different levels of sophistication in the digital divide: when access is achieved; when people actually use ICT; when ICT leads to real impact.

Big data doesn't sample

In statistics, the "n" stands for a small sample of the large university of all subjects, N. The data on how frequently you call and the length of these phone calls, we can know a lot about you (age, gender, etc.).

All pervasiveness: general purpose technology 🡪 decentralized agenda

Involve all kinds of different sectors since they are all affected Let each one write a contribution/ contribute Decentralized agenda= Broad input, but specialized input **The challenges arising from the fact that ICT are all-pervasiveness can be confronted by decentralizing the agenda and involving experts from many different sectors

Digital gender divide confusion

Negative correlation between women and use of technology. Women are more natural and more efficient communicators. **If one is not sure if the correlation is true or confounded, one can control on the basis of other variables and see if the effect continues to exist Weird correlation/ fake correlation between women and ICT Confounding variables Women with same level of employment, income, and education as men use ICT more than men Women just on average earn less money and have less education **The reason why women have less access to ICT than men is because women are discriminated against in society

Human development index

Not only looking at income as an indicator Dimensions (and indicators): health (life expectancy), education (expected years of schooling and mean years in school), living standards (GNI per capita) Missing indicator: political freedom indicator, democracy Why? Countries can't even agree on the right level of transparency and what democracy really is. it is difficult to measure. ** The HDI is based on 3 distinct dimensions: health, education, and economic power The Human Development Index is based on 3 dimensions: health (life expectancy at birth), education (mean years in school and years of school expected), and income (Gross National Income per capita)

Global ICT developments

Number of subscriptions of countries is increasing Digital divide is decreasing in terms of quantitative characteristics, but there is a difference in quality Ex: A mobile phone in a developing country versus one in a developed country

e-Science for politics and policies

Obama 2012 campaign -Used data fusion -Ranked voters -Ran voter simulations to predict voting behavior

Flat earth comparison

One argument: due to digitalization, the world is flat. This increased flow of information, every individual can participate in globalization. We are unified by equal opportunities due to digitalization. The iron rule: "If it can be done, it will be done" If you don't do it, it will be done to you. ** When the T. Friedman uses the metaphor "that the world is flat", due to digitalization, he means it as: an expression for viewing the world as a level playing field in terms of trade and communication Counterargument #1: the world is not really flat because there is not equal access and therefore, not equal opportunities (global divide) Counterargument #2: this "equality" is not a good thing. ** Two common counter arguments to the claim that digitalization lead to a flat world full of positive opportunities are: Digital inequality and global downsides of digitalization Counterargument #3: the global context in which the digital revolution happens The Big (macro) systemic level Historical (colonialism) Economical (industrialism) Social (mobility) Political (governance) The small (micro) activist level Aid, projects, and resource allocation Individuals and individual governments

Network externalities example

Phone example: x nodes = x^2 - x degrees 1 node: 0 degrees 2 nodes: 2 degrees 3 nodes: 6 degrees 4 nodes: 12 degrees 5 nodes: 20 degrees Exponentially more valuable to be in bigger social networks

Positive and negative feedback

Positive Feedback: Feeds a behavior or action Put oil in to the fire to blow it up or put water in the fire to cool it down More leads to more Less leads to less Negative feedback: Stabilize overtime Keep fire burning More leads to less Less leads to more (decrease price to sell more) **Regulation that aims at creating the right balance between information sharing on social online networks and privacy protection is an example of: Negative feedback

Digital pixels

Products evolve to people's needs and preferences 🡪 Adopters of new innovations: bell curve 1. Innovators / creators with unique skills to execute this idea 2. Early adopters 3. Early majority (most percentage of individuals in it) 4. Late majority 5. Laggards ** In his theory on the "Diffusion of Innovations", the Communication Scholar Everett Rogers distinguishes among 5 different kind of adopters: innovators, early adopters, early majority, late majority, and laggards

Digital divide

Progress of technology and Diffusion into society ** Typically, periods of intense technological diffusion are accompanied by slowed down technological progress based on continuous innovation Digital divide: new dimension of inequality (information and technology inequality) that has to do with the diffusion of technology

Two dimensions of the digital divide

Quantity Number of subscriptions person Digital divide has decreased Quality Bandwidth Digital divide has increased Digital divide increases or decreases depending on what kind of digital innovations are diffusing through social networks

All models are wrong: limits of science

Representations of reality are always incomplete; they are just models! Models are never 1 to 1. Different models can explain the same thing from different perspectives. "All models are wrong, but some are useful" **people say that all models are wrong because models necessarily only capture part of reality. The Butterfly Effect 🡪 a flap of a butterfly in china creates a hurricane in Florida To understand reality, you need to use many models. Digital computational social science gets us a little bit closer to predicting reality! Big data requires big visions for big change

Philanthropy

Rich people such as Bill Gates pledging to donate Aim at developing hands on technologies to help end poverty (irrigation systems, anti-Malaria nets) Improves basic health, bringing power, providing clean water, boosting agriculture, investing in education) ** In what is known as "The Giving Pledge" some of the world's wealthiest individuals and families have committed to dedicate the majority of their wealth to philanthropy ( over 50%) Sometimes these voluntary efforts are not as successful. The Global Digital Solidarity Fund Places 1% on private sector corporations Proposes a new financing mechanism, mobilized new partners, promotes South-South cooperation Private sector didn't want to pay tax Looked to see if governments would volunteer and pledge

Network data fusion

Social network Analysis 🡪 analyze TRADITIONAL database of attributes and network databases of links together to get a very interesting and complete picture of how society actually works Data fusion helps us learn about who you are and who you hang out with through social networks

Diffusion wave through social networks

TWO WAYS TO MEASURE DIFFUSION Exponentially growth of diffusion of cumulative adopters (S-shaped) Bell curve of new adopters ** There are two complementary ways we can track a diffusion process. One results in an inverted U-shaped "hump", the other in a S-shaped diffusion curve. What is correct: The S-shaped curve tracks the total number of people who have already adopted since the beginning Social Networks: the channels through which innovations (and other diseases spread) Nodes/ vertices and edges/ties ** With a denser network, the diffusion process will faster than before. ** When the network is too sparse, it might be that some parts of the social network are disconnected and are never reached by the innovation. ** The shape of the S-shaped diffusion curves become steeper in later years. General tendency: increased social connectedness leads to increased speed of diffusion ** in the diffusion of innovation in social networks, the network structure in general affects the diffusion pattern There is also diffusion of legal innovations/ laws.

Different Perspectives of a Globalized World

The 80 % of the global population that lives in lower income countries has 70 % of the world's ICT subscriptions

Global Governace

The Big (macro) systemic level: Political (governance) United Nations Organization League of Nations (founded after WWI 🡪 failed) United Nations was founded after WWII **balance of power among the 3 powers of the State are: Legislation makes law (represents the people), executive branch implements laws (Police force, military, the president/commander in chief) , judicial branch interprets laws and checks the executive branch really implements what the legislative branch says United nations Has legislative branch - general assembly that represents all of the countries Has judicial system -- International Court of Justice Problems : not set up well and dysfunctional Does good work too NO executive branch - partial and incomplete global governance Closest thing we have is the UN Security Council People feel like deep reforms are necessary to ensure social economic growth and to prevent another world war **Who are the five members of the UN Security Council? US, UK, France, Russia, China ** The current global governance structure in form of the United Nations is characterized by: a productive legislative, a quirky judicial, and an unbalanced executive

Social mobility

The Big (macro) systemic level: Social (mobility) ** Can people move around more freely now around the globe than 200 years ago? Technologically, yes. We have planes and trains and cars, but we also have a lot of laws and visas that make it more difficult to actually move to another place Adam Smith's definition of "free markets": the mobility of labor should be free and allowed Free labor markets: assure that the right people are there for the right job, while less qualified ones are replaced **Institutional regulations like the ones around the "green card" go against the ideal of free markets of Adam Smith The Four Freedoms of the European Union 1. Free movement of capital 2. Free movement of goods 3. Free movement of services 4. Free movement of people *NOT GLOBALIZED OUTSIDE OF THE EUROPEAN UNION

MDG's and SDG's

The Small Activist Level Millennium Development Goals (MDG's): global policy agenda What we want to work on 1990-2015 Millennium Development Goals (MDG) 1. Eradicate extreme poverty and hunger 2. Achieve universal primary education 3. Promote gender equality and empower women 4. Reduce child mortality 5. Improve maternal health 6. Combat HIV/AIDS and other diseases 7. Target 8. Global partnership for development **The MDG's are goals state leaders have set for the world's countries to be achieved by 2015

What's the difference between the Internet and the WWW? (Network structure)

The internet is the physical infrastructure (the network of cables and radio waves) and the WWW is the generic service. The internet is used for email, file transfer, and conferencing.

Red queen effect

The technological frontier is moving a target ** Economists use the term "red queen effect" to describe: a process where the goal is constantly moving ahead Number subscriptions = number of people Information capacity = economic capacity ** It is increasingly becoming useless to measure the digital divide in terms of the number of subscriptions, because everybody has at least a mobile phone nowadays. However: differences in terms of telecommunication capacity continue to depend on economic possibilities

Transaction cost theory

Transaction Cost Reduction Internet Banking is cheaper for banks, instead of paying employees. Cost reduction= more productivity Digitalization leads to bigger companies typically, but it depends on the product or services their selling!

What is the antidote to confront the particularity of the extraordinarily high uncertainty in digital development?

Uncertainty of trajectory: exponential technological process 🡪 short-term flexibility All pervasiveness: general purpose technology 🡪 decentralized agenda Unpredictability of side effects: creative destruction 🡪 private- public sector alliance Internationality of digital networks: space of flows 🡪 international coordination

Antidotes to confront the particularity of high uncertainty in digital development

Uncertainty of trajectory: exponential technological process 🡪 short-term flexibility It's important to know where you want to go in the long-term 10 years is too long, so much can happen Short term plan: 2-year action plan ** One way to confront the uncertainty arising from exponential technological progress is to Combine a long-term vision with flexible short-term action plans ** In only 2 years, policy makers in Latin America considered a considerable share of policy priorities to be obsolete. How many? About 1 in 4 priorities.

Proxies vs reality

We confuse them. Data proxies versus realities Confusing the foot with the footprint. Metadata: data about data NSA kills terrorists based on data footprint.

Development of what kind of human development

What are the most important pillars and indicators to look at? John Rawls: freedom through justice Development is freedom Therefore, we need justice to obtain it. The original positive behind the veil of ignorance ** Rawls says that justice would be realized if a society would follow the following procedure: society is designed by everybody without knowing one's individual position in society 🡪 freedom 🡪 development Amartya Sen Political freedoms, economic facilities, social opportunities, transparency guarantees, and protective securities These instrumental roles sometimes needed to be traded to obtain the other Multidimensional components of development

What is human development?

What is a developed society? Generally we think of rich countries, economic power, GDP, purchasing power, a lot of income Joseph Stiglitz, economist video Thoughts shape behavior If we strive for high GDP, we will create policies that will increase GDP **The Nobel Prize winning economist Joseph Stiglitz is worried about using GDP as the common metric of development because: how we describe success is what we strive for

Network externalities

When the value of a product to one user depends on how many other users there are, economists say that this product exhibits network externalities, or network effects The value of digital products: have Positive Externalities: the more people who use it, the more positive it is Traditionally, economists see things as having negative externalities: the more people use it, the less valuable it is and less there is left for you. ** What is true with regard to positive and negative network externalities: positive network externalities means the more people, the more benefit for everyone

Crowd sourcing

When you source your resources from the crowd Poly- directionality (review concept^) Death of Distance

Social media metrics (network structure)

Who is the center of the network?: Network Centrality Degree centrality: most connected/ The center is the node with the most connections (the most popular kid on the block) Closeness Centrality: closest to all (reach all others more quickly, few steps with all others) - sum of all the steps to all the other nodes Betweenness Centrality: on the paths connecting all ("gatekeeper") - sum of shortest paths through node/ all shortest paths Eigenvector Centrality: friends of friends (it's all about who you know) Proportional to sum of neighbor's centrality If your friends have a lot of friends, you're good. If not, you know no one.

What determines the digital divide

Who is the subject of connectivity? ** The digital divide is NOT always measured as the divide between people 3. Which attributes matter? Why do some connect more or earlier on then others? More income More education Spurious correlations: analyze them separately ** The only reason why people with more education also have more connectivity is because they also have higher incomes 🡪 false, we don't know! Spurious correlation. Can integrate more variables What determines who connects early or later? The network structure, which is not captured by just looking at attributes. ** According the analysis we've just seen, the two most influential variables in the explanation of the digital divide are : education and income

Schelling's Segregation model

You can see segregation between neighborhoods in society ** the rule each individual is given in this set-up is: if you are not happy (not at least 50% of your neighbors are like you), then move to the next free space and reevaluate if you're happy now. **after running the model, which percentage of the neighbors of each individual will be from a. similar color? About 50% (same as the individual's preference) **When changing preferences from 75% to 76%, we reach a lower level of segregation. **So called "tipping points", "phase-transitions", or "non-linearities" refer in this case here to the fact that: small changes in the behavior, attitudes of individuals can make big differences in the collective behavior of society In conclusion of this model: Human intuition often fails. Total is different than the sum of its parts We can detect phase transitions/ tipping points (taking even, little steps and waiting to see where the big change finally happens, the final tipping point)

Can we end global poverty by tomorrow?

`**how many people are there in the world? 🡪 7.53 billion people! ** How much money would we need per day to have one additional US$ for everybody worldwide living in poverty? $2.8 billion FX market trades $2,000,000,000 billion/year How much money we'd need to end poverty tomorrow: (1,000 billion/year) / ($2,000,000,000 billion/year) = 0.05% **The amount of taxes collected by the U.S. government (from all levels) amounts to roughly what % of the U.S. economy? About 25%

Media richness selection

ability to hand multiple cues simultaneously, facilitate rapid feedback, establish a personal focus, utilize natural language High to medium to low media richness: face to face dialogue, video chat, telephone, voicemail, text message, fax, email, letter Depends on complexity of information that you are communicating Likelihood of being hired is highly affected by context and medium that you communicate with when being evaluated

The ultimate theoretical limit to the power of Big Data for making predictions in a changing world consists in the fact that:

data is always from the past and cannot predict unprecedented future dynamics

Whistleblowing

disclosure from an insider/employee in a government agency or private enterprise to the public of those of authority of mismanagement, corruption, illegality, or some other wrong doing Wikileaks: where people can send whistleblower insides Whistleblowing or giving away secrets? 🡪 thin line Digital information leaves a footprint`

Big data analytics does not merely refer to the volume of data. In class we identified other characteristics. The following is not one of them:

machine learning makes data fusion unnecessary

Exposure selection

makes it technically possible for you to control how much of your digital footprint is exposed - "selectable" degree of privacy/ identity Facebook default privacy settings have gotten more relaxed and public as time has gone on

Hadoop and Map Reduce are two pioneering Big Data solutions that made the processing of vast amounts of data possible for the first time. As explained in one of the videos, based on Google search technology, Map Reduce...

maps large datasets across multiple servers and then gets it back as a summary in the reduced stage

During the 2012 Presidential campaign, the team of President Obama

spent $ 1,000 million and used Facebook apps that convinced almost 80 % of targeted undecided voters to vote for Obama

Analyzing online search engine entries, researchers from Google were able to predict the spread of the winter flu in the United States almost in real-time. They achieved this by analyzing:

the 50 million most common search terms and correlating them in 450 million models against actual flu cases from the past

Surveillance

the digital panopticon " The idea of a guard being able to watch all prisoners at all times in the center of a prison 1984, a novel where the government and private corporations spied on their society Big Brother

Through the combination of real-world data sources and computer simulations, researchers were able to show that in the case of an outbreak of smallpox in the city of Portland, the first and second important elements to detain the spread of the outbreak were:

the speed and delay in the response