Together with a group of designers, researchers and journalists we are working in a publication on the Application of AI for Planning and Climate Adaptation (SCAPE magazine).
While diving into the topic, we have started wondering: how will less profitable and more activist fields like landscape architecture or nature conservation be able to develop their own AI systems? And how would be the best approach to make them not only efficient but also to work within the same values of openness, collaboration and sustainability that we share, and we do not see in current available models.
Inspiring initiatives in other fields make us think that there is another way around Big Tech corporations, and we would like to understand the developer perspective on it.
We are happy to hear any opinions, discussions, strategic advices, development tips or any other remark shared that you think is essential for developing, deploying and maintaining such an open source AI system for Landscape Architecture.
For context, as Landscape Architects, our work is quite broad, from designing green public spaces for cities, to developing city level planning focused on greener, walkable and climate adaptive neighborhoods, to larger regional plans focused on nature and floodplain restoration.
In the field of landscape architecture the emergence of the computer and internet changed the profession, and not always for good. We can see the risks of ai, pushing landscape architects to more generic design, quick visual output, efficiency, low cost, etcetera. At the same time we see the opportunity of integrating ever improving climate models, ecology mapping, better understanding how to manipulate the landscape to optimize biodiversity and climate adaptivity. But what about the things that are hard to digitalise? Word to mouth stories, soft values, local culture, local history, seasonality, atmosphere, etcetera? Exactly because landscape architecture is not a very large/profitable market, it’s not likely commercial companies will jump on this. We think it’s worth developing/training an AI for local soft values—run on a solar/hydro powered datacenter. With universities—but we’d need a larger community to make it work.
Thank you in advance for any answer – we will link to this post and fully cite you in the magazine for all the information shared,
And hopefully we can build a collective view on this,
[Question] How to build open source AI models for landscape architecture?
Hello everyone,
Together with a group of designers, researchers and journalists we are working in a publication on the Application of AI for Planning and Climate Adaptation (SCAPE magazine).
While diving into the topic, we have started wondering: how will less profitable and more activist fields like landscape architecture or nature conservation be able to develop their own AI systems? And how would be the best approach to make them not only efficient but also to work within the same values of openness, collaboration and sustainability that we share, and we do not see in current available models.
Inspiring initiatives in other fields make us think that there is another way around Big Tech corporations, and we would like to understand the developer perspective on it.
We are happy to hear any opinions, discussions, strategic advices, development tips or any other remark shared that you think is essential for developing, deploying and maintaining such an open source AI system for Landscape Architecture.
For context, as Landscape Architects, our work is quite broad, from designing green public spaces for cities, to developing city level planning focused on greener, walkable and climate adaptive neighborhoods, to larger regional plans focused on nature and floodplain restoration.
In the field of landscape architecture the emergence of the computer and internet changed the profession, and not always for good. We can see the risks of ai, pushing landscape architects to more generic design, quick visual output, efficiency, low cost, etcetera. At the same time we see the opportunity of integrating ever improving climate models, ecology mapping, better understanding how to manipulate the landscape to optimize biodiversity and climate adaptivity. But what about the things that are hard to digitalise? Word to mouth stories, soft values, local culture, local history, seasonality, atmosphere, etcetera? Exactly because landscape architecture is not a very large/profitable market, it’s not likely commercial companies will jump on this. We think it’s worth developing/training an AI for local soft values—run on a solar/hydro powered datacenter. With universities—but we’d need a larger community to make it work.
Thank you in advance for any answer – we will link to this post and fully cite you in the magazine for all the information shared,
And hopefully we can build a collective view on this,
Best,
Simon