#44 AI in Architecture: Disrupting the AEC Industry with Tim Fu

[@StephenCoorlas]

Great talk - Love this conversation and Tim's approach/perspective on the future of AI integration in Architectural Practice. Agree with just about everything said here 🤘

[@schamon13]

Architects as gardeners - cultivating the soil - sowing our imaginative intent - discovering the fruit that emerges from the medium .

[@cichym]

In 45:22, when u saying about "movent of architet with no computer" it is like listening to music on vinyl. In times where have vast of digital equipment, and streaming services with access to all human music, Vinyls are still having good. Perhaps it is this analog approach.

[@priyanshisoni5856]

It's an amazing discussion. Coming from Engineering background with passion for Architecture, would love to know your view on how design automation in the field of Architecture will influence Structural Engineering. How can engineers keep up with the innovation to create buildings to support Architects' vision?

[@vitalperfection437]

future will the full of dumbest people . we are making lazy people who was proudly saying i designed it by asking A.I . their is no designer future.. what about fundamental 3d or design. what about if u have no internet and u have a task in 3d or design and u don't know manually 3d and design. because u always asking A.I . to design.

[@olugbengaadegun1753]

You have a point. But you must' balance the two. Be hard working and be up to date Technologically

[@svart-rav8072]

Lost me already when he described Evolutionary algorithms as "another verison of AI" at around 5:05 ...
That is so fundamentally wrong and shows how shallow the understanding of the underlying technology is.
It's actually shocking, that someone like him is talking about this technology as an "expert" in our profession

[@PauloSilva-dz6uh]

I put your concern over the chatGPT and this is what I've got: "Evolutionary algorithms (EAs) are not another version of artificial intelligence (AI) per se, but rather a subfield or a category of AI algorithms. EAs are a family of optimization algorithms inspired by the process of biological evolution.
In general, AI refers to the broad field of creating intelligent machines or systems that can perform tasks that typically require human intelligence. This field encompasses various subfields and techniques, including machine learning, natural language processing, computer vision, expert systems, and more.
Evolutionary algorithms, on the other hand, are a specific class of algorithms that simulate the process of natural selection and evolution to solve optimization problems. They are based on the idea of maintaining a population of candidate solutions, applying evolutionary operators such as selection, crossover, and mutation, and iteratively improving the solutions over generations. EAs have been successfully applied to a wide range of optimization problems in engineering, finance, biology, and other domains.
While EAs are a powerful tool for optimization, they are not the only approach to AI. AI encompasses a much broader range of techniques and methods, including but not limited to deep learning, reinforcement learning, expert systems, and symbolic reasoning. Each of these approaches has its own strengths and applications within the field of AI."

[@AidanNewsome]

They're very related, but if you'd like to define what AI is, be by guest 🙂

[@svart-rav8072]

Every innovation that contributed to the recent boom of "AI" (a term most machine/deep learning engineers wouldn't even use) we have seen in the recent years came from neural networks and even more precisely from deep learning.
Evolutionary algorithms have nothing to do with either of those and throwing so mindlessly in the mix when talking about conditional latent diffusers like SD or midjourney... is.. well... at least careless or maybe even naive
I think it shows very vividly the quite strong disconnection of users in the architectural scene to the actual technologies
And I would claim a discussion about what "AI" is, is quite useless.
There is a myriad of different algorithms and architectures revolving around differnt problems - and choosing, building and training the right setup correctly for the connected problem is the art of that

[@PauloSilva-dz6uh]

@@svart-rav8072 Thanks for your thoughts. What kind of discussion do you propose ?

[@AidanNewsome]

I'm not sure there's a discussion worth to be had with you anyways