Algorithms
We build them; they shape us

Technical advances promise a great deal then often deliver something completely different. Such was the case of the washing machine at the beginning of the twentieth century, and it may be the case with artificial intelligence today.

By Lorena Jaume-Palasí

At the beginning of the twentieth century, the new technology of the washing machine was anticipated to be a timesaver for women that would integrate them into the labour market. It was declared a technology of liberation, an expression of people’s will to take control of and shape their own futures. The washing machine came about as a result of deeper socio-economic changes: First, cotton revolutionised the labour market in many ways in the nineteenth century – including for women.

New materials – new challenges 

The previously most commonly used textiles – such as wool or linen – had to be beaten and scrubbed. The introduction of cotton textiles in the clothing industry created new cleaning requirements and challenges. The majority of households did not have running water. Washing clothes was not just women’s work. Women needed the help of servants or other family members – usually men – who split the firewood and carried large washtubs into the house. On average, one or two full working days per household were spent cleaning the laundry.
 
The new socio-economic developments may have promoted the integration of women into the world of work outside of their own households, but at the same time they continued social asymmetries and discrimination. The introduction of the washing machine was meant to save time and give women more time to do other things. The washing machine, it was promised, would reduce health risks for women caused by caustic soaps and substances. 

Future forecasts

Positive scenarios argue similarly today for the future of working with artificial intelligence: Algorithms will save time. Through the assistance of automated programs and machines, people will have more time to devote themselves to thinking and higher tasks. By now, countless studies have been done on the future of work, society and democracy in the age of technologies such as artificial intelligence. The main focus of their forecasts is on calculating the hours worked that can be saved with algorithmic systems which, depending on the job profile, means cutting jobs.
 
The limitless range of information that can be digitally produced and distributed – its increased rate of dissemination and the potential social effects – are also the focus of hypotheses. They speculate on what the job profiles of people who would work alongside programs or machines would look like. They reflect on the opportunities and challenges that algorithmic systems enable with their mathematisation and automation of the world. Efficiency and the crossing of geographical, linguistic or physical boundaries are the focus of both the incentives for the development of new services and of criticism. The narratives about algorithmic systems alternate between dystopia and utopia. For the most part, the focus on these crossings of boundaries is the cause of false expectations.
 
Indeed, the washing machine saved people time. By combining electricity and running water, it was possible for the machine to replace exactly those activities in the washing process that were usually carried out by men in the household. Thus, laundry became an exclusively female activity. The commercialisation strategy for washing machines made the washing of laundry again a predominantly female household chore.

 The washing machine set hygienic standards – and brought about new demands 

The washing machine demonstrates how the question of the time spent in the household can be considered relationally. The washing machine no longer led women into the world of work and, from a relational perspective, it is not even certain how much time it actually saved. In retrospect, the washing machine – its design and commercialisation strategy – was an expression of a social concept that put women in the home rather than the world of work. The washing machine led to higher household cleanliness standards and brought with it new tasks and standards for household activities.
 
The circumstances of the post-war period are also significant for the widespread commercial expansion of the washing machine. The men returned from the war and the integration of large numbers of women in universities and the world of work that was observed during the war, at least in the West, declined. With its design and the decisions made for its standardisation, the washing machine reinforced conventional social ideas about the role of women. 

Algorithms: Automation of social processes 

We can assess the algorithmic systems behind more complex software programs in a similar way. They are used to partially or fully automate social processes. The providers decide in advance what data formats and possible options for action exist in a process and the software executes one of these options depending on its interaction with the user. They mathematise social processes and convert them into programming codes. They are therefore an expression of social concepts and entail political decisions. Similar to the decisions about the design of the washing machine, the algorithm contains many covert assumptions about the world and the users of such services. At the same time, algorithms often reveal social mechanisms to us that we have only reflected on to a limited extent. These technologies are based on and serve social needs: We build them; they shape us. Artificial intelligence is an expression of our culture and our will to shape the future.
 
If the past demonstrates anything about the effects of the introduction of technologies, it is that they alone do not bring about any socio-economic changes per se, but are also or primarily an expression and means of power and power opportunities. Their development is less coincidental than an expression of the inherent political character of design and standards of escalating technologies.
 
A look at societal asymmetries and traditional images, on which legal possibilities and restrictions and socio-economic needs are based, will be crucial for understanding how algorithmic systems are designed and embedded in society. Artificial intelligence is an opportunity to use mathematics to map and understand the world differently, to automate social processes and thus to organise them differently. Saving time and crossing boundaries alone will not change the society of the future. The space for change will depend on the legal, social and economic practices and visions that shape us as a society.