CB

The future of architecture in data-ism

In Homo Deus: A Brief History of Tomorrow, Noah Yuval Harari extends the history of the human race, explored in his previous book Sapiens, into a speculative future involving humanity, nature and technology. Harari‚Äôs thinking centres around a common ground introduced up by data that organisms and computing machines produce intelligence through algorithms. Harari uses a modern scientific theory that living things exist as biochemical data processors, while computers process digital data, making them comparable in experiencing and decision making. This equivalence in detecting and processing the world provides Harai‚Äô with a possible future scenario where the biochemical cannot process or absorb the scale and speed data flow required by ‚Äúsuper intelligent‚ÄĚ digital algorithm. While Harari concedes the assumption that life is reducible to data flow and decision-making fall short in explaining consciousness, the assumption sets up a scenario where humans understand and identify through their contribution to data flow in the ‚ÄúInternet of Everything‚ÄĚ (Harari 2018). Harari‚Äôs last chapter specifically explores the ‚Äúreligion‚ÄĚ of data-ism which considers humanity‚Äôs future in a world organised by contribution to the data flow. While Harari‚Äôs explores the impact on nature, technology and humanity, there is an excuse to consider architectures role in data-ism, and the potential influence of data-ism on architecture as a way of thinking?

Two distinct possibilities appear to exist, both concerning the actual and perceived role of humans in the universe. The first is for humans to de-centralise themselves in the world and become more aware of the benefits from caring for nature and the natural environment, rather than manipulating and mistreating them. The other scenario is for the replacement of humans at the centre by algorithms that become more intelligent than humans, resulting in humanity becoming manipulated rather than the manipulator. My understanding of how technology progresses as a response to human problems, and the fact that intelligent algorithms can make decisions but do not have independent intentions, leads me to have faith in scenario one.

The main reason for siding with scenario one is through understanding means that intelligent algorithms influence the physical world. The majority of algorithms exert agency through information, examples of this agency appear in the fields of investing, music and art. However, algorithms have trouble influencing the material world directly because they do not have bodies. The counter-argument to this statement is the existence of robots which do affect the physical, but algorithms do not produce robots, humans do. While an algorithm can influence the physical world through a robot, the human-defined purpose of the robot determines the outcome. The critical aspect to point out is that despite much materialist theorising and Hollywood speculation, algorithms cannot build robots and are always limited to the ability of their interface with the material world. The interface is a human construct. Some predict a future where intelligent algorithms begin to create material interfaces by depositing matter through 3d printing, but the 3d printer is a human-designed interface. I am not claiming that algorithms may one day produce material interfaces, just that they will require human help to do it. Therefore we must keep our eye on the ball to not create the terminator - I know that was what you were thinking.

The real threat from algorithms is through human behavioural manipulation via information. Political interests realised a long time ago that semantic information manipulates humans. In this contemporary ‚Äúpost-truth‚ÄĚ era propaganda can come from both humans and algorithms (bots), with many humans unaware of its real origin. The outcomes are stark, algorithms help elect seemingly unelectable leaders, algorithms promote crazy, illogical responses to threatening human viruses, and algorithms even coax human behaviour in physical space. So while superior algorithms seem physically harmless to humans, the real threat is from manipulation.

So to return to the question - what is architecture’s role in data-ism? If Harari’s future is correct and humans become manipulated toward maintaining a data flow critical to intelligent algorithms, then the built environment could shift in its role from supporting human functions to determining them, but how would this happen? We know that algorithms cannot shape material space through direct action, so it requires agents able to do its bidding, hello humans! If an algorithm knows the arrangement of material space will produce data for its upkeep, via distributed sensors in the built environment, it could convince a human to construct it for them. Construction requires visual communication, which is part of the traditional architect’s toolkit, but this requirement for visible information changes when materials shape and organise through digital fabrication? In this process, data potentially passes from algorithm to fabrication machine, then into material parts and coordinated through the logic of a jigsaw. Rather than relying on a level of construction knowledge and skill, as architects do, the algorithm looks for anyone with hands to connect modular and discrete elements into easily assembled wholes. In the case of large scale 3d printing, human labour need not apply as matter can construct through algorithmic decisions, although, the guilty fabrication machine would rely on humans for existence.

If this scenario seems far fetched look to Toronto Quayside where Sidewalk Labs hope intelligent algorithms will direct flexible construction systems and planning rules promoting renovation to organise the urban environment to maximise rental returns and human happiness. Therefore, while algorithms have no direct influence on the built environment, they do not worry as they have the means to manipulate a specific chemical in biochemical algorithms, dopamine. Those who are involved in imagining and maintaining the built environment must have a heightened awareness of information sources influencing decision making to be confident that they are not inadvertently puppets for the new algorithmic overlords.

Harari, Yuval Noah, 2016, Homo Deus: A Brief History of Tomorrow, Random House