What do you think should be done to get Archicad a bigger market share ?

[Anonymous]

I personally find ARCHICAD way more powerful, easy to use, faster and all in all innovative than other solutions like Revit. However since Revit is backed up by Autodesk and long years of Autocad dominance, ARCHICAD isn’t leading in terms of market share, what do you think GRAPHISOFT should do better? Do you think it’s enough for them to just push for a better product ? Or is there something else you would do ?

[DGSketcher]

@Podolsky: I don't disagree that there may be an element of inevitability to your vision, but I am an Architectural Technologist and I see not only the technical requirements of a project but also that Architects work hard at their Art to produce interesting spaces, and there in lies the difficulty for AI. An algorithm can never replace the human capacity for original and creative thought, especially where the design requires an empathy with not only the individuals that will use the building but also the environment it sits in. AI may be able to generate something visually sympathetic in a dense urban setting with plenty of constraints, but what happens when it has no point of reference and how far do you trust it on selecting finishes? Assuming AI became fully autonomous then is there an inevitability that it will see a cube as the ideal representation of a buildable structure with maximum volume to surface area, is that a future we want?

AI's future lies in building code compliance and potentially core structural design, but it will always be reliant on a human exchange / interface. There was a bit of a clue to this in AC24 with the introduction of SAM. Unfortunately from an architectural perspective it hasn't been well implemented. If the primary aim was purely to get architectural designers to think about their structural lines then I would be 100% behind it, I have seen too many designs where load lines from roof to ground have not been properly considered. Regrettably I feel the fundamental bugs and terminology in SAM have been counter productive to encouraging its take up, which just serves to underline why AI's future in construction is going to be a long and difficult road.

[Podolsky]

@DGSketcher: I like to hear how architects always saying: "I'm so artistic person and my talent cannot be replaced by any machine". And then he is reviewing modern magazines, searching where to steal another cool idea for his running project. Unfortunately most of the architects are not so artistic. Most of the time they make absolutely typical buildings - following national building regulations and styles. And sometimes - when client is ready to pay - do something original and artistic (but not always buildable!). It's also important to understand - many architects are able to provide construction drawings only when building is completed. So, it's a big difference - have an idea how build the building and actually build it.

We also shall not to forget - how many useless drafters are sitting in the offices and cannot produce error free good looking documentation. How it usually happens - or architect is selling to client full of errors, drawn somehow set of drawings - and client then worry how to build all that. Or, if people care a little bit about quality of project - it goes through seven circles of hell of reviewing, marking-up, changing, searching for errors etc. This endless process also includes scandals with client from time to time with lawyers involved (because client wants to build it as soon as possible), boring meetings with lost meaning - what this meeting is about... Up to 40% building material waste. And at the end we are getting property overpriced 10 times all around the globe and conversation about global warming and CO.

Let save the planet from global warming, CO, poverty and homelessness and create solid and stable AI that will make projects faster and better than current generation of architects.

[jl_lt]

i agree with mr. podolski that it will happen eventually (maybe not exactly as he describes, but close); when it will happen, i have no idea. but check out what these guys are doing:. www.digitalbluefoam.com.

on the other hand, while processes might get increacingly more efficient and faster, i dont necesarily agree that all this technology will make for better architecture if we are not careful

[Podolsky]

Why it's necessary to be careful, when it's already all is very bad with modern buildings. Because of price. If our parents were able to earn and buy an apartment / house during their working career, then today this possibility is closed for many of working people. Approximate price of small apartment in London today is 1 million pounds. So question - can an architect, who is working in London with salary £30,000 per year, buy an apartment? It's 35 years of his salaries. Real price of apartment in London is about £100,000. But because so many useless people are involved into process of making a project and selling it (including architects, who like to speak about art but cannot bring on site correct detail drawing) - the price is jumping to a million - we need to pay salaries to whom who constantly making mistakes in the project and to whom who correcting them. And also to pay to people who is "supervising" process (but in reality who is supervising how they are receiving percentages from the deals).

Architecture and construction is the same since Rome Empire. The only material, that brought new shapes in modern architecture, was reinforced concrete. But concrete also was known since Rome Empire. In total nothing really changed since that time - same brick, windows, columns, beams, roof, doors.

Most of the buildings in this world - boxes. Yes, we all live in boxes with rectangular windows. Shape and functions of the buildings are controlled by building regulations and local authorities (fire safety for example). Architecture is controlled by quite strict formulas, that coming from physical laws of gravity, heat and light distribution. Architecture can be transformed into pure mathematic, i.e. architectural rules can be passed to AI and AI can produce projects. Remotely, on server, and communicate with us as Siri: "I have 25 options of ground floor plan".

I think AI is absolutely real - and there is no point to close eyes to it or ignore it. Coming back to the main question of this post: to get a bigger market share, Graphisoft shall actively develop AI solutions on ArchiCAD platform. Maybe now not as mainstream, more like "Lab", but this is essential. I'm absolutely sure - very very soon AutoDesk will come to us with some very interesting AI solution as a game changer.

When I open ArchiCAD - I cannot say to it: "Please draw the wall, 2 meters length, 3 meters height, made from brick". But if I could - even if I could type it via console - it already would be very beneficial. It's like first step of human on the Moon.

[Nader Belal]

Let's not fool ourselves ... personally, if I had an AI capable of designing buildings, I would not hire an architect to design my buildings for what ever the reason.

You will tell me that AI can't replicate style, arts astetics, ... etc

I would tell you that right now in our profession, money talks, always, so if those things pays off, it would be considered, if not, it wouldn't.

Besides, right now there are thousands of AIs, that learn by feeding data, pics, drawings, models, etc ... all I have to do is feed this to my AI, and comment on the outcomes that it generates

[DGSketcher]

I'm not so sure AI is compatible with eliminating poverty and homelessness especially when you see posts on an architectural forum implying it would allow you to shut down inefficient architectural practices. In the UK the effects of AI are already clearly demonstrated in the diminished number of shops, rising unemployment, food banks and consequential "desperation" crime created by a move to on-line shopping no doubt driven by Amazon and FBs AI systems. Be careful what you wish for, the future lies in sustainable communities not a race to the bottom replacing jobs with automation. If you have workers making mistakes ask yourself why? Was it perhaps because you have had opportunities to learn that were denied to them because you were so busy clicking your mouse you didn't think to ask why they are struggling and help them to develop. People learn if you teach them, if you leave them in a vacuum don't complain when they don't understand. This thread feels like it is loosing touch with humanity.

Archicad should be an assistive tool not a business replacement technology.

[jl_lt]

Podolsky wrote:
Why it's necessary to be careful, when it's already all is very bad with modern buildings.

When I open ArchiCAD - I cannot say to it: "Please draw the wall, 2 meters length, 3 meters height, made from brick". But if I could - even if I could type it via console - it already would be very beneficial. It's like first step of human on the Moon.

Im not following you Mr Podolsky. Technology has certainly improved since the start of the 20th century but the projects we see now are not much better and in many cases worse than what was done BY HAND, INK AND PAPER decades ago. Why would you expect that more technology would make things better? or worse, why instead of being part of the solution would you sit and wait for someone to design a software to make automatic designs for you?

Why is it necesary to be careful? we, as a profession have already given up most of the control of our build environment to contractors, consultants and real estate developers. The problems you mention about prices have nothing to do with design technology but with economic pressure, global rise in costs, and because we have gotten out of the way (what was affordable 50 years ago is not affordable now because there are less resources). So you want to further give away control to AI? yes, we should be careful with that. If most real estate developers dont really care about architecture or built environment, what makes us think an AI will?

I think technology changes, but society and Architects are extremely slow to change our ways of living our own environment, so we want to use new technology to keep designing anachronistic buildings. We have to evolve too.

As a side note, you say you want to open Archicad and type for a wall 2 meters lenght and 3 meters high. You can already do that in Archicad! what is exactly that you want? the computer do all the design for you and you just watching as an achitectural voyeur?

[Podolsky]

Maybe I need to explain more, what I want.

1. I want more tools in my hands to develop automatic solutions, that working inside of ArchiCAD. If we would get interface Grasshopper-like that would give us full control of ArchiCAD project file - without meeting limitations from time to time, how it happens now if you decided to model something extraordinary and use full BIM capacities. Then maybe will grow the whole community where ArchiCAD users can discuss their algorithms and collect all that into large international AI database. I guess this is why Dynamo is in Revit and one day would be able control file on such a level.
We have Grasshopper connection, of course, but Grasshopper cannot make annotations, place dimensions and issue drawings.

2. Why the price of estate grew up 10 times, when quality of construction the same, even when we started to use computers? One of the answers - we are not using computers right. In time of hand drawn drawings and bills calculated on calculator nobody was able to say - I cannot draw it this way, because my pen cannot. But we facing to such a problem with computers. So, what will happen if we start to use computers in AEC on the level how it's designed today? First at all we will find complete financial mess - corruption of masses and government's. We know how Amazon, Google and Apple changed the whole industry of shops and logistic. It is possible to change completely construction industry - when digital revolution will finally enter the industry (currently AEC digital solutions are dancing tango with construction).

3. Humans are doing mistakes in sizes. Most common mistake - something somewhere is not fitting or clashes. And here there is nothing to do with teaching people. it's a geometry that person must learn in school, not in the office. Machine is not doing such mistakes.

4. When I'm saying - I want to say to ArchiCAD (or type it in console) do this and this, but cannot, means that I cannot do it. I cannot replace mouse operation with command. Software does not have console. But, AutoCAD, by the way, does. With AutoCAD it is possible to speak on AutoLisp.

[DGSketcher]

Podolsky wrote:
3. Humans are doing mistakes in sizes. Most common mistake - something somewhere is not fitting or clashes. And here there is nothing to do with teaching people. it's a geometry that person must learn in school, not in the office. Machine is not doing such mistakes.

Sorry but this really annoys me. We all make mistakes but this is minimised when you teach people of the consequences of their decision good or bad. From lead Architect to site labourer, the construction industry has to teach incomers how things work in the real world. Leaving school doesn't make you a specialist in any field, neither does coming out of higher education with a degree. If your staff have never been on a construction site how can they appreciate how that simple line on a piece of paper (or tablet) relates to the trades on site trying to make it work. If you have no time to support your staff with training, experience & guidance then you personally are compounding their failures and that of the industry. This forum is founded on the principle of providing positive support to people asking for help, in time they start answering other peoples questions, that is how teaching works in the real world. Stop hiding behind the need for AI and start showing some respect to those around you trying their best to contribute to society.

PS The price of property in London is inflated because you also contributed to that problem by choosing it as a destination. In basic economics it is called Sale & Demand, the more people that want a piece of a limited commodity the more the price goes up.

[Podolsky]

@DGSketcher: Really? Stop closing your eyes and avoid the huge problem we are all facing. Mistakes are doing employed workers and workers from sub-contractors. Who shall we judge, if lift suppliers, after receiving IFC model send us the model of lift and drawings, where top of the lift obviously clashes with roof steel purlins. And if nobody is checking every corner in the model, then such a clash will be discovered on site only during installation.
Maybe then we need to change lift supplier? Or organise them webinars?

I personally think, that an idea of federated models in BIM (how it's described in new BIM standards) will never work. Because it's about to merge corrupted models, review them, mark up clashes, send comments, organise various meetings and receive next revision of corrupted models and search for errors again. Until data is becoming error free - contractor will finish the building on site.

And about estate price. Do you think that prices are growing because of growth of human population? Then what, after 20 years, when population will increase in size more, the estate will be overpriced 20 times? And it's OK?
And how the terrible waste of building materials (up to 40%) is related to that?