Thomas Mclucas

The project is focused on Marttila, Greater Helsinki and looks at the Finnish type-house, specifically those manufactured by Puutalo Oy in the post-war years, as an example of the material consequences of standardisation and automation. The Marttila neighbourhood was inaugurated by the Finnish president in 1943. It was built as a residential area for injured war veterans after the Fenno-Soviet war, which ended in 1940. It was one of the first large-scale developments constructed by Puutalo Oy and it has since retained much of its suburban character.

Several of the smaller houses built by Puutalo Oy in Marttila are now difficult to recognize due to substantial alterations and additions. One might see a similar resistance to the standardisation of the smart home, whereby technologies are ‘domesticated’, through circumventing restrictions and undermining the intentions of the hardware and software designers.

Puutalo Oy used industrial techniques to standardise components allowing for homes to be made for Finland, Columbia, or the UK. The contextual adaptability of these homes is reminiscent of the computer-simulated urban dynamics of Jay Forrester in 1969, cited by Nicholas Negroponte, who wrote of a building “model which, with proper changes in parameters, is good for New York, Calcutta, a gold rush camp, or West Berlin.” Popular houses sold by Puutalo Oy in the 40s were the Rauhakoto (Abode of Peace), Metsäkoto (Forest home), and Syväaho (Remote meadow), each name describing something similar to the subtle ‘indicators shaped by site conditions, traditions, social setting, prior experiences, the whims of inhabitants’, which Negroponte deemed crucial to the responsiveness of a truly intelligent architecture produced by machine learning. 

Puutalo Oy’s homes were a means for advancing the welfare of the family, asserting a certain morality that equated the normative relationships in the home to positive societal outcomes, whereby gender identities were mutually inclusive to this idealised housing model. Similarly, AI is often engendering digital normativity, through the development of algorithmic standards which inform what should be considered normal in human actions in digital and physical space. Using objective observational data, predictive algorithms reduce the options available to their subjects to known precedents, thus eliminating the inherent subjectivity of the subject, and normalising them. AI acts to compress data so that it is consumable, however, the resolution lost in this compression is often without consent.

Developing methodologies of architectural masking and glitch, the project uses various methods of data-bending in order to create disturbances to undermine the normative and idealised vision of Puutalo Oy. The starting point was using the sauna as a typology to sit in opposition to this model housing. The sauna contaminates the image of Puutalo Oy and resists its normativity.

The original drawings are sourced from the Puutalo Oy archive at the Finnish Archives for Business Records (ELKA) and edited for consistency.

Aurora AI summary

The Finnish constitution guarantees citizens the right to the ‘sanctity of the home’. This aspect of the constitution is conflated with the privacy of information. Finland’s Ministry of Economic Affairs and Employment published a policy paper in 2017, the expressed ambition of which was ‘turning Finland into a leading country in the application of artificial intelligence’. This manifested itself in Finland’s state sponsored AI programme, AuroraAI.

AuroraAI directed users to potential public services. The programme was focused on the key events of different human transitions. The aim of the program was to give citizens access to personalized services based on personal data and aggregated data from the network.

These intimate interventions require personal data, the collection of which may seem to stand at philosophical odds to the rights to privacy as established by the country’s constitution. Indeed, the Finnish Constitutional Law Committee has proposed strengthening the Constitution, to ensure accountability and recognise ‘the ambiguity of algorithms in automated decision making’.

The speculative future envisaged shows Aurora making mistakes in adapting a Puutalo Oy house for different groups through literal mistranslation, misunderstanding and programmed ignorance. The scenes express effects on the actors who are excluded from or whose environment is normalised by Aurora.

The sauna is a type of heterotopia, where one must take certain actions to enter the space, as part of a purification. This both isolates them and makes them penetrable. They represent the values of the culture they exist in while simultaneously inverting and contesting them. As an analogy they allow us to think through the processes of undressing/scanning, washing/clearing and sweating/generating as a sequential spatial experience which both describes the traditional sauna or a system of digital de-simulation.

Appropriating the space of the sauna and associated paraphernalia acts as a method to interfere in data collection and control, to allow inhabitants to manipulate the flow and visibility of their personal data.

As a form of resistance to Aurora’s adaptation of the Puutalo house, the project proposes an alternative algorithm which aggregates the parts of a Puutalo Oy house to produce new typologies. The algorithm is first tested or ‘trained’ on a sauna. The sauna volumes are maintained but the Puutalo panels are designed around them.

The algorithm has become more complex over its development. After initially selecting the part names and number of parts, various constraints were added. Firstly the number of parts became proportional to their representation in the original ‘Päiväkoto’ Puutalo Oy design which dominates Marttila. Then the volumes of the ‘Päiväkoto’ house were generated and aggregated. These volume inputs could be changed to begin to accommodate different users. The aggregations of these volumes acted as surfaces along which parts could be arranged into walls of a consistent thickness. As these arrangements became more and more feasible the media used to represent them became more material, in scale models, scale installation and architectural drawing.

Every generated building is unique and one could stop at the first option generated or infinitely iterate the design. Through this position there is a negotiation between the algorithm and the individual, as the buildings take flexibility on the part of the user in occupying these new spaces while at the same time being completely bespoke.

The buildings shown could be the inputs of a sole occupant, who wants to live alone, a group of young people, maybe two couples who want to live together, or a multigenerational household.