Ai is the answer, but what was the question? – –

The title of this review is a minor modification of a lecture by the brilliant architects Cedric Price entitled “If technology is the answer, what was the question?” The obvious exchange of technology for AI is suitable today, since artificial intelligence is apparently everywhere: Google search results are adopted by AI overviews. E -Mail software offers AI support. While I write this, almost every banner contains some mention of AI on my browser tabs on my browser tab. Even a heading in today's New York Times claims that “every AI used for everything” and at the same time asks: “Is that bad?” After years in which he was hidden in various technologies, the AI ​​is now in front and in the center, which makes people alike excited and impressive: how far will AI progress and what will it mean for me, my job? Chatgpt and other chatbots make writers quite nervous, while Midjourney and other diffusion models do the same with architects. Should architects worry about the effects of AI on their work, their career choice? This is one of many such questions that can be treated in time in time Architecture in the age of artificial intelligence.

Leach mentions the rhetorical question of Price on page 176 of his 320-page book in a chapter that should be of interest to every architect: “The future of the architecture office”. At this point in the book – in the seventh of nine chapters – the leaders will have recognized that Leach is a strong advocate of the AI, for technology and for the further development of the profession through the introduction of AI. Thirty years ago, architects were between hand drawing and computer -aided drawing (CAD) on the hinge. Ten years later, after the turn of the millennium, the building formation model (BIM) made themselves in architectural offices, while Grasshopper and other parametric software studios infiltrated studios infiltrated in architectural schools. While the trajectory is clear, the hug of these technologies now keeps most architects in the hands of architects. AI, on the other hand, shifts control over a machine that has been trained on more data than every single architect could ever digest in her life. While the images of Midjourney and other generative AI programs are exactly the case that pictures that are often controlled by the input of practicing architects should nevertheless accelerate AI technologies for architects for elimination.

Leach fans fans all fear of AI in the book. In the second chapter “A Short History of the Ki”, one of the progress he writes about, Alphago, a Deep Learn computer program, the name of which implies its purpose – is the game of Go – and his successor Alphago Zero, who taught himself to use data from human games, and without knowing the rules of Go! Chatgpt and other chatbots “scraping” data from the Internet was given a lot of attention to learn, but Alphago Zero has a similar, but admittedly narrower what was done by itself: she trained in three days by playing 4.9 million games from Go against itself. Alphago defeated Lee Sedol, a professional Go player in 2016, with the movements of the AI ​​player, who was so unexpected that Lee Alphago described it as creative -a term that is often reserved for humans. Three years later he withdrew from Professional Go and said he could never surpass a AI player. Here, too, writing on the wall is clear: career altering through an entity that has the potential than to be to be creative As it is intelligent.

The AI, which came unexpectedly and creatively in the field of architecture in the area of ​​architecture, occurred a few years ago, not long after Dall-E, a mechanical text-to-in-in-in-im-im-im-Date model, published in 2021, and the first edition of Leach's book was published. Midjourney's Open Beta was published in mid -2022, and shortly afterwards populated photo -realistic pictures of fantastic buildings Instagram. This trend was continued as updates for Dall-E, Midjourney, stable diffusion, lookx and other generative design applications burned out some of the quirks and reacted more to the needs of architects. Since more and more architects include these technologies, some of them are being called what Randy German refers to as “superuser”, as those within Foster + Partners, used R + D -Team, Code research group by Zaha Hadid Architects' Code and Design research group, and Coop Himmelb (L) from Deep Himmelb (l) au. These three companies and their respective “Superuser” are discussed in Liach's book in a chapter entitled “AI and Architecture” but outside of Foster architecturally intelligent.

The biggest obstacle for the AI, which is intelligent enough to take over the architects' jobs, is the jump from 2D to 3D. “This remains extremely difficult,” writes Leach, “not only because the required calculation performance would be extremely high, but also because there are still no tools available for dealing with the complexity of working with neural networks in 3D.” At least AI can create 2D images, but no 3D forms, even though startups concentrate on architecture and development that could realize such a shift. One is forma, which was founded in 2016 as a spacemaker and then acquired by Autodesk in 2020, and another is Xkool, who was founded in 2016 by two architects, one of them, Wanyu Er, a former grandma employee (Xkool = “Ex-Koolhaas”, get it?). Spacemaker/Formma was created to make codes and other data for developers so that AI would become an “invisible assistant” for architects, while Xkool was created by architects for architects so that they could “hand over the routine and repeat the computer to repetititive work”. In both cases, architects who use architects to use AI will “to maximize their return on capital”, Leach writes “and optimize the performance of their buildings” -the architects will cause the technology to be firmly included in their offices.

Every book about AI and in particular one about AI and architecture must make predictions. Excavation builds his own predictions in those of Toby Walsh who wrote Machines that think: the future of artificial intelligence In 2009 and Ray Kurzweil, which were published The old intelligent machine In 1990 and The age of spiritual machines: when computers exceed human intelligence 2009. The predictions of the latter “about technology in general and AI in particular”, writes Leach, “have proven to be remarkably precise.” While Kurzweil is related to “wireless networks [allowing] The easy release, of course […] And other creations ”, his mistakes are ridiculous, as on television until 2009 with“ cybernetic chauffeurs ”. LEACHE's own predictions form the practical concerns of architects and the efforts of a AI.