Forgive me if this topic is discussed elsewhere, but I've been following a couple of companies - Testfit and Google's Delve. It seems they are quickly evolving to become real alternatives to architects for multifamily/large development applications. It seems to me automation (generative design, AI), may happen faster than all of us are anticipating at least for the front end tasks of our profession.
With that in mind my thought is that high end residential work will in some ways always remain in the hands of the architect. Wealthy clients will always want something custom or unique.
Does that make high end residential a safer bet over the next 30 years?
These are just some thoughts, but would love to hear what other's are thinking the future holds. Perhaps I should just skip through all the bullshit and start learning to program in Python.
When Delve finally becomes commercially viable, it will be a great tool for multifamily firms when it comes to coming up with floor plan options, massing and other preliminary design work that currently takes a lot of man hours. That is, if the software is flexible enough to accomodate all the local variables that make this work such a time consuming task in the first place. Will architects pay to use integrate it into their BIM workflow or will Google's original (now downscaled) goal of selling it directly to developers, bypassing the architects altogether, win the day? I think most developers will still leave the work to architects but the danger is that the race to the bottom in terms of fees (and hours) will be exacerbated.
And high-end res - helping rich motherfuckers rape and pilllage the earth for their own egotistical glory - is about the least sustainable choice you can make for the planet.
one thing that surprised me when i first worked on a residential tower was how little time we spent on unit layouts. we had a principal who was good at that and once the tower geometry and target sizes were defined it took him all of about 4 hours to hand draft the standard units.
i don't doubt AI tools will become part of the architects workflow, but at this point they only expedite studies - they don't actually invent concepts or figure anything out. i think their adoption and evolution will be slower and less impactful than the routine business cycle impacts on the market.
it might even grow the market for architecture by letting design teams produce good work faster. consider that the development of large jet aircraft didn't mean we needed fewer pilots to fly the same amount of people as small propeller planes - it meant more people could afford to fly further. the market grew to fit.
that said, if you can do the work high end has always been a better niche market. the clients aren't price sensitive and tend not to run out of money when financing dries up.
This is a great example. So some firms take a much longer time figuring out layouts, whether its due to trying to balance a funky massing with logical interiors, weird local code idiosyncracies, or constantly changing client requests. If there are principals who can be counted on to do layouts efficiently, then Delve would have less application. For firms that rely on foreign manpower to spend countless OT to "perfect" a floorplan, Delve could come in handy - if its more cost effective than hiring a bunch of foreign grads who would swallow long hours in exchange for work permits!
Are there really firms out there where the building design and layout are farmed out to cheap interns and foreign contract-work staff? I've never worked in a place where the bulk of the space planning and design work wasn't handled directly by the PA or partner in charge.
'AI tools will become part of the architects workflow, but at this point they only expedite studies'
Kinda the big (and continuing) misconception. ML w/ programmed variables is hardly 'intelligent.' Dunno why everyone keeps referring to any of this as 'AI'....bit of a personal pet peeve.
The PA/Partner in Charge oversees the planning - but sometimes the actual burden of re-drawing and re-drawing floor plans (plus all the code compliance stuff) is hevaily borne by green card applicants. So its not farmed out per se as the employees are still in house, but its the kind of ardurous work that some non-Americans are stuck doing for years while waiting for their green cards.
Nov 18, 21 3:41 pm ·
·
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.
Thoughts on emerging automation software, where to direct focus for longevity.
Forgive me if this topic is discussed elsewhere, but I've been following a couple of companies - Testfit and Google's Delve. It seems they are quickly evolving to become real alternatives to architects for multifamily/large development applications. It seems to me automation (generative design, AI), may happen faster than all of us are anticipating at least for the front end tasks of our profession.
With that in mind my thought is that high end residential work will in some ways always remain in the hands of the architect. Wealthy clients will always want something custom or unique.
Does that make high end residential a safer bet over the next 30 years?
These are just some thoughts, but would love to hear what other's are thinking the future holds. Perhaps I should just skip through all the bullshit and start learning to program in Python.
A safer bet? Like crypto? High end residential is so fucking niche, you'd be better off creating NFTs.
When Delve finally becomes commercially viable, it will be a great tool for multifamily firms when it comes to coming up with floor plan options, massing and other preliminary design work that currently takes a lot of man hours. That is, if the software is flexible enough to accomodate all the local variables that make this work such a time consuming task in the first place. Will architects pay to use integrate it into their BIM workflow or will Google's original (now downscaled) goal of selling it directly to developers, bypassing the architects altogether, win the day? I think most developers will still leave the work to architects but the danger is that the race to the bottom in terms of fees (and hours) will be exacerbated.
Nothing is a safe bet. Especially architecture.
And high-end res - helping rich motherfuckers rape and pilllage the earth for their own egotistical glory - is about the least sustainable choice you can make for the planet.
don't you live in the hamptons?
one thing that surprised me when i first worked on a residential tower was how little time we spent on unit layouts. we had a principal who was good at that and once the tower geometry and target sizes were defined it took him all of about 4 hours to hand draft the standard units.
i don't doubt AI tools will become part of the architects workflow, but at this point they only expedite studies - they don't actually invent concepts or figure anything out. i think their adoption and evolution will be slower and less impactful than the routine business cycle impacts on the market.
it might even grow the market for architecture by letting design teams produce good work faster. consider that the development of large jet aircraft didn't mean we needed fewer pilots to fly the same amount of people as small propeller planes - it meant more people could afford to fly further. the market grew to fit.
that said, if you can do the work high end has always been a better niche market. the clients aren't price sensitive and tend not to run out of money when financing dries up.
This is a great example. So some firms take a much longer time figuring out layouts, whether its due to trying to balance a funky massing with logical interiors, weird local code idiosyncracies, or constantly changing client requests. If there are principals who can be counted on to do layouts efficiently, then Delve would have less application. For firms that rely on foreign manpower to spend countless OT to "perfect" a floorplan, Delve could come in handy - if its more cost effective than hiring a bunch of foreign grads who would swallow long hours in exchange for work permits!
Didn't mean to downvote, whoops...
Are there really firms out there where the building design and layout are farmed out to cheap interns and foreign contract-work staff? I've never worked in a place where the bulk of the space planning and design work wasn't handled directly by the PA or partner in charge.
'AI tools will become part of the architects workflow, but at this point they only expedite studies'
Kinda the big (and continuing) misconception. ML w/ programmed variables is hardly 'intelligent.' Dunno why everyone keeps referring to any of this as 'AI'....bit of a personal pet peeve.
The PA/Partner in Charge oversees the planning - but sometimes the actual burden of re-drawing and re-drawing floor plans (plus all the code compliance stuff) is hevaily borne by green card applicants. So its not farmed out per se as the employees are still in house, but its the kind of ardurous work that some non-Americans are stuck doing for years while waiting for their green cards.
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.