AI and communicative action

At their best, developments in cities are informed by consultation with large numbers of residents and stakeholders. In his summary of the state of planning theory in 2005, Phil Allmendinger wrote:

“Recent revisions to UK planning process predicated on inclusion and consultation perhaps add an addition to this mix, that of stakeholder values and desires which are expressed at both the local and regional level” [104].

Consultation implies dialogue. Planning theory draws theoretical weight from the writing of the sociologist Jürgen Habermas, particularly his influential book The Theory of Communicative Action.

Part of my motivation for pursuing conversational AI (chatGPT, Bart, etc) is to investigate how this technology intervenes in what is already a conversational domain, namely the formation of our cities, including its agents and decision making processes.

Urban rationality

The concept of rationality provides the benchmark for urban decision making. In turn, that grounds decision making in “reality.” But for Habermas, rationality is already subject to conversational interactions within communities. Here is such an assertion by Habermas, which he identifies as the phenomenological position. The phenomenologist does not

“simply begin with the ontological presupposition of an objective world; he makes this a problem by inquiring into the conditions under which the unity of an objective world is constituted for the members of a community.”

That’s not a relativist position, as if to say that reality is whatever any individual or group deems it to be. The phenomenologist’s position is grounded in communities engaged in particular actions.

To assert that London in reality is host to secret agencies for managing interplanetary aliens (Men in Black) would only be of practical use in so far as any significant actions might follow from that belief. Hence the term “communicative action.” In more prosaic terms, Habermas says of communicative communities:

“Through this communicative practice they assure themselves at the same time of their common liferelations, of an intersubjectively shared lifeworld. This lifeworld is bounded by the totality of interpretations presupposed by the members as background knowledge.”

That’s a hermeneutical position to which many professionals and scholars subscribe. I’ve written about that in several places, including blog posts. See posts tagged “hermeneutics“.

That AI might perform as an actor or agent in such communicative practices is certainly of consequence. Much of the recent planning literature subsumes AI within discussions about the idea of the “smart city.” But it’s also worth thinking about what AI as agent in conversations about the city, in prospect or “reality,” might do to urban communicative practice.

Bibliography

• Allmendinger, P. (2005). “Applying Lacanian insight and a dash of Derridean deconstruction to planning’s ‘dark side’.” Planning Theory 4(1): pp. 87-112.
• Allmendinger, P. (2009). Planning Theory: Palgrave Macmillan.
• Allmendinger, P. (2021). Forgotten City. Bristol: Bristol University Press.
• Habermas, J. (1987). The Theory of Communicative Action. Cambridge, UK: Polity Press.
• Peng, Z.-R., K.-F. Lu, et al. (2023). “The Pathway of Urban Planning AI: From Planning Support to Plan-Making.” Journal of Planning Education and Research: pp. 1-17.

Note

• Featured image is from Dall-e, prompted by: Cross section of a secret agency in London for managing interplanetary aliens in the style of the movie Men in Black.