Connected by Climate

Connected by Climate

                  In A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming¹, Paul Edwards tackles the topic of global knowledge using the climate as his example. He begins by using the common phrase, “think globally, act locally” (Edwards 1). By this, Edwards means that our local actions impact the world and what’s happening around the world impacts us locally, in other words we are an “interconnected whole” (1).  To provide an example, Edward uses climate to address the complex task of identifying and sharing knowledge. It is fitting that he chooses “climate” to talk about people being connected since, according to the Oxford Dictionary Online¹, climate’s suffix “mate” can be defined as a as a noun that means, “a fellow member or joint occupant of a specified thing” and as a verb that means, “connect or be connected mechanically.”  Furthermore, as Edwards says toward the end of the first chapter, the weather data network “is arguably the oldest of all systems for producing globalist information” (Edwards 24). Climate is a great example to use because its suffix not only means connection, but the study of it is one of the first world-wide attempts to form unified knowledge.

            As Edwards discusses, knowledge formation does not begin with communicating information. Knowledge begins with assembling data, which includes knowing how to identify data and also having tools to accurately capture and assess that data. In the case of climate, each country uses different measurements, the tools they use change over time, and even where the data is connected changes. All this results in differences from data collected 150 years ago and even data collected 20 years ago (6).  This is where the concept of a “vast machine” is introduced. A vast machine is “a sociotechnical system that collects data, models physical processes, tests theories, and ultimately generates a widely shared understanding” (8).  Data and observations must be transformed into widely accepted knowledge which includes the political process, transmission of information via the media, and even an understanding of what counts as data.

                  Since we’ve mentioned that data varies, it is important to have “reanalysis,” which is a technique that helps standardize data previously collected, even when collected at different times by different methods. There is also the concept of “gateways” which “can join previously incompatible systems” (10). In the case of climate, this allows weather systems to interact with the ocean monitoring systems, seismographs, and more.  “Knowledge infrastructures comprise robust networks of people, artifacts, and institutions that generate, share, and maintain specific knowledge about the human and natural worlds” (17).  These knowledge infrastructures are sociotechnical because people don’t just add facts. They must assimilate the facts they have, which is akin to the basics of scientific knowledge. To do so, the information presented must be consistent with other things people know. Plus, the new information must be accepted within a community and the person providing the information must have trust and authority. Scientific knowledge is therefore communicated through many infrastructures and institutions including universities, libraries, and laboratories.  “The infrastructure is a production, communication, storage and maintenance web with both social and technical dimensions” (18). Infrastructures are largely invisible until they no longer work. For instance, one may not pay attention to the infrastructure of roads, highways, and traffic lights until the traffic lights don’t work and as a result, traffic no longer flows smoothly and the number of accidents increase.

            Globalist information occurs when the knowledge transcends into a political contect and nations begin working together to create change. “It may be driven by believes about what knowledge can offer to science or society” (25) and creates a world-wide infrastructure of knowledge formation and transmission. With countless examples related to climate, capturing and analyzing climate data, and organizations that study climate, A Vast Machine explains how knowledge is formed and how it is then communicated, not just locally, but globally, influencing people everywhere.

²Oxford Dictionary. Oxford University Press. Accessed at http://www.oxforddictionaries.com/definition/english/mate

¹Edwards, Paul. A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming. MIT Press: 2010.

Simon Says and Computers: Information is not Intelligence

Simon Says and Computers: Information is not Intelligence

Computers can only do what you tell them, making computers akin to participants in the popular children’s game “Simon Says.” In the children’s game, one person plays the role of Simon and commands participants to perform certain actions beginning with the phrase, “Simon Says.” Participants can only do what Simon commands. Similarly, with computers, the programmer is “Simon” and a computer’s response is limited to its understanding of what it is being instructed to do. Computers are not able to act outside of this realm, so the commands (programming) initially received are critical to computers output, functionality, and effectiveness.

In “The Stupidity of Computers,” David Auerbach¹  explores the intelligence of computers. Filled with numerous examples, this article takes readers on a journey of computer output. It begins by discussing how programmers input information so that a computer can understand and respond to queries. It then evolves into conversations about search engines, language and ontology, and popular information-generating and information-gathering websites. Finally, it concludes with greater implications of how computers impact government, social structure, and individuals’ identification and categorization of themselves and others. Ultimately, this article sheds light on the limitations of computer intelligence and the growing dominance of computer intelligence on human life.

Through a series of examples, Auerbach explains the growth of computer intelligence. For instance, a simple sentence such as “Highlight Japan on a map” requires a set of commands that could be as long as a paragraph. The computer can respond to this request only if the programmer properly input the necessary commands and also provided a map and a highlight feature.  Computer instructions must be precise and language matters. Ambiguous language brings about confusion and in searches, can bring about undesired result.  Initial search engines could only search for what an individual input and could not extract similar information. For instance, if an individual did a search for “outdoor parks in Alabama,” original search engines would only be able to generate search results that contained the four words in the original search. The search results would not generate responses that referenced “Ruffner Mountain in Birmingham” because this last phrase does not include any of the key words although Ruffner Mountain is an outdoor park in Alabama. This example shows that computers can only respond to the information given them, whereas a human would be able to respond with Ruffner Mountain if they were asked about outdoor parks. This inability to interpret or extrapolate answers challenged the effectiveness of internet searches, a problem that was solved by the search engine Google.

Eventually, Google found a solution by “using the topology of the web as opposed to its meaning.” Google began paying attention to which pages were linked together – in other words, they must have a connection to one another.  This helped provide more relevant results. That way, if someone types in “outdoor parks in Alabama,” anything that classifies itself as such would pop up. In addition, an individual might even see advertisements for products relevant to being in the outdoors such as an ad for Mountain High Outfitters, a store that provides outdoor gear. Today, Google not only provides more relevant results, but it will also try to “predict” what you are searching for by creating a drop down menu of choices as you input letters and words in the search bar. For instance, by simply typing the letter “a” in my Google search bar, the following pop up “Amazon, Alabama power, AOL, and Alagasco.” When I type in the entire word I’m looking for “Atlanta,” Google populates the dropdown menu with “Atlanta Braves, Atlanta, Atlanta Airport, and Atlanta Braves Schedule.” These are the most popular searches individuals are currently looking for when they type “Atlanta” into Google.  Google is able to return desirable responses because of how its programmers have set it up to recognize information and its relationship to other information and also because of metadata categorization. Individuals can aid how online information is categorized and the frequency in which it is returned in search results by providing “key words” that are often used in searches. On Twitter, the key words are proceeded by a hash tag symbol.

The latter half of the essay focuses on the importance of ontology, or assigning recognizable categories to metadata. This results in comparing and contrasting the effectiveness of several popular internet giants such as Ask (formerly Ask Jeeves), Yahoo, Wikipedia, Amazon, Facebook, Twitter, and more. Due to the massive amount of information online, the web is faced with the challenge of categorizing information. Amazon successfully does this by categorizing information in well-accepted categories such as “books, household appliances, toys.” Facebook successfully does this by having individuals select the categories to which they belong – people can input their religious group, college, favorite bands, favorite TV shows, favorite politicians, and more. These categories or ontology used by Amazon and Facebook create methods in which to group and analyze people. Auerbach cautions that these categories are limited to human’s determination of categories and that computers cannot automatically generate the categories.  Auerbach also cautions that the government and other places are making connections and conclusions between what people have liked. However, “the sheer number of uncontrolled variables at work makes it dangerous to take any of these conclusions at face value.” 

Other internet-based examples of the role internet has in shaping how we receive information include blogs, LinkedIn and Ning. LinkedIn.com uses the slogan, “World’s Largest Professional Network” and is essentially a place for individuals to house their professional and academic accomplishments online. People can link/connect/network with other individuals by searching individuals by profession, college, geographic location, skills and more. In an attempt to help individuals find other individuals relevant to them, LinkedIn also features “individuals who looked at this profile also looked at” and provides a list of other individuals frequently searched for tangentially with the person currently on their screen. Ning.com has the slogan “build and cultivate your own community.” Unlike the other social media websites that connect people who have mutual connections, Ning.com connections are limited to those within that particular Ning group. On Ning, one would be a part of multiple Ning groups that don’t intersect, which is unlike other social media. The information contained on these websites and other websites are being used to create categories, correlations, and connections of people and items that would otherwise not be connected.

Although this article does not go in-depth into any one subject that it mentions, it provides a comprehensive overview of the limitations and implications of computers continual integration in society.  The greater implications of this article are that computers and the internet are only as smart as the information and algorithms initially input into them. Computers and the internet are not able to think for themselves and are therefore just an advanced version of the children’s game Simon Says. Computers can only do what they are told although they are very good at responding and providing information, as long as they have access to the information requested. Information should not be confused with the higher calling of intelligence. Computers can help humans synthesize and analyze information. Computers cannot replace or create human intelligence. 

 

 

¹ Auerbach, David. “The Stupidity of Computers.” Machine Politics: Issue 13, Winter 2012. https://nplusonemag.com/issue-13/essays/stupidity-of-computers/.