Midjourney Architectural Design (Interior / Exterior)

[@dragonwisdom]

If you are reading this please remember that you are meant for greatness! 😊... Keep working hard everyday and you will achieve your goals. You got this! 😊

[@StephenCoorlas]

Great positive message 🤘🏼

[@christopherstorey721]

To anyone freaked out by this. I have gone through phases of extreme fear in the last few weeks after seeing videos like this. But I've come back to earth and feel pretty grounded now. After examining my day to day work I've come to the conclusion Architecture is actually pretty safe from an AI take over. These images tell an extremely superficial story of the design process. They look interesting and will be useful for things like video games. However they completely ignore the conceptual complexity of architecture; as it relates to multiple frameworks, such as context, contracts, stakeholders, constructability, sustainability, clients, budget, bylaws, building codes, functionality, efficiency and last but not least sculptural meaning or art. Responsiveness to all these shifting conceptual frameworks is what makes good architecture. Furthermore these images imply a shallow view of architecture as art, which just doesn't reflect the reality of practice of any art. Where architecture is art, it is necessary that it be human concept, derived from human experience. Otherwise it is meaningless to humans and not art. It is that simple.

[@StephenCoorlas]

With no disrespect, I think you’re looking at it wrong. It’s not about architecture being safe from this or any new technology, it’s about how the technology will be adapted by our profession. There’s no doubt that the architectural practice is multifaceted, but it’s likely you’ve been using technologies that have replaced laborious, or repetitive tasks during documentation and/or design. Even the simple task of copy/pasting a storefront mullion, or dragging a curtainwall system with automated mullion placement could be considered a shallow overlook of a meaningful design moment. Not to mention automated schedules, tables, keynoting, and Revit has some great generative design tools for duct layouts and other MEP tasks. The point is, you can decide to see “art” in a lot of tasks, and if you consider automating design as a significant loss for the practice of architecture then I think our profession has been hanging on by a weak thread to idea that architecture is an art form that must be guided only by human emotion. This technology will relinquish us from the shallow task of design, and allow us to focus on more meaningful and purposeful agendas of efficiency, performance, and overall healthiness of the architecture, in terms of human physical and emotional health, and the healthiness of our earth.

[@mikeshiva2118]

Thank you Chris i feel so reliefed after reading this

[@blaply3421]

​@@StephenCoorlas I agree. I'm someone who has been interested in 3D design/concept art for a long time in various fields (mainly entertainment). Although it seems AI can and will be able to do (at least) concept art better than humans, instead of being scared, I see opportunity in this, even in a more threatened field than architecture. I think AI should be utilized as a tool/assistant which saves creators time and lets us focus on more important things like art direction, meaning, etc... There are way too many dumb and time-consuming robotic tasks but also creative ones that could be done with the help of an AI. In a way, it's just a tool that should be utilized similarly to how computers and also generative and procedural creation processes were already utilized in the past. I know AI is a double-edged sword but we have to learn how to utilize it the best way instead of being afraid of it.

[@nihilistarchitect]

I love the part you mention something on the lines of : if you are not into 3d visualizations then it’s going to be a bit difficult to you to get results… I don’t see where the difficulty is at all, anyone pretty much can do it. Some might do it better but everyone who can type and put into words an intention can do this and I guess that’s my problem with these toys. Also I find it extremely interesting how you defend the idea of using your own references as input , then to let an AI, which is basically a black box you don’t understand , that has references from all images it studied online , come up with a result . I myself don’t know what do think of this but I like to ask the question. As for saying that the final results are beautiful, is like saying a pair of shoes are beautiful. No mention on confort, durability, materials and construction are made. As with parametric architecture , that didn’t take into consideration the power of computing to optimize design for the most part , this is purely form, aesthetic vaine pursuits , which is just in pair with the imagery infused Instagram world we live in. Meaning and context are in danger here. Wanna be creative and truly innovative? ASK questions, study, think, analyse. Architecture is so much more than being beautiful.

[@StephenCoorlas]

Yes, good you were listening. Feel free to explore my other videos where I discuss the philosophy of universal truth and how Plato attempted to define beauty in relation to forms and particulates. The concept of beauty can, in fact, hold deeper and more meaningful associations when used in context or relation to past experiences, which is why beauty is considered subjective; but can also, in some instances, be a parameter as an objective descriptor if enough data supports it. A lot of people need help learning to be more specific or descriptive of the type of images they want to produce. Yes, I encourage everyone to use their own images as the primary input source for generative content. At some level, every idea is a remix of existing data, but reducing the creative process to clicking "submit" vs filtering the data through your own mind and hand, based on your experience and your perspective, I believe adds a level of authorship and copyright security to the generated content. Let's go deeper.

[@discgolfsantafe]

Hi, Stephen. Joseph here. As a California licensed architect & general contractor (and past architecture instructor at Otis Parsons in LA), I've been intrigued for the past year by AI for writing, content generation, and the possibilities of design. The video was great because it apparently accurately portrays the evolution of your work from hand-generated models, thru, probably Autodesk, maybe something else, then into the murky world of MidJourney. I really appreciate the integrity you bring to the design process, particularly in advising (young?) designers to generate work based on their independently-created designs. Yes! You are no doubt closer to the current construction technologies than I, but I would love to see some of the more dreamlike MidJourney designs manifest into actual creations/creatures - a thing I have a hard time comprehending... at least today.

[@StephenCoorlas]

Hi Joseph, thanks for your thoughtful comment. It has been an interesting journey for me, and because this is all so new I’m doing my best to responsibly guide especially younger architects as they begin to explore a new technology that is bound to replace many of the nuanced design process I’ve had to battle through. I think it’s a very exciting technology that will relinquish our profession to focus on even more valuable agendas. Construction technologies are also advancing quickly. I have my eye on 3D printing and I’m hoping to encourage others to begin exploring how to bring AI designs to life with 3D printing and other digital fabrication methods.

[@discgolfsantafe]

@@StephenCoorlas Sounds good. Be interested to know what your "more valuable agendas" might be.

[@StephenCoorlas]

@@discgolfsantafe Mainly more efficient build processes that are earth-friendly, but also healthier designs, both physically and mentally for humans. And also better functioning designs that are less dependent on manufactured systems, utilities, and trendy finishes.

[@discgolfsantafe]

@@StephenCoorlas Wonderful. Good luck - I really like the direction you're taking and sense the necessity during these trying times! I'd like to bring some sort of affordable housing with gardens for the homeless to New Mexico, and have been looking into gov't programs that would support something like that. (No luck yet) I've seen people in CA using SIPs to make the building process easier and repeatable for less skilled folks. I may be naive, but I believe that if people can be part of design/building development (actual hands-on), they would be more invested in their living areas and feel grounded enough to positively change their lives. Architecture in action! Yay!

[@ShustovR]

Thanks! There is so little of this on the web for architects and designers and these are amazing results. + your sketches are beautiful! Subscribed! I'm waiting for new videos.

[@StephenCoorlas]

Thanks for the kind words 🙏 I'm hoping to encourage others to pursue these ideas!

[@VarunKumar-fn4fm]

I consider your subscribers (including myself) lucky to be the first to come across this amazing stuff. Great job!

[@StephenCoorlas]

Thanks so much - I really appreciate that!

[@adetibakayode1332]

YES !! VERY MATURE AND HELPFUL !

[@StephenCoorlas]

Glad this was helpful! 🤘

[@twitterglobalarmy]

While Midjourney certainly has some interesting features, I find myself gravitating towards Blue Willow more and more. The fact that it's completely free

[@StephenCoorlas]

Yup, it’s a great alternative, and the quality is getting better!

[@CryptobdSchool]

@@StephenCoorlas Yes, Blue Willow is growing fast. BlueWillow V3 is now live and deployed. Have you tried with V3?

[@StephenCoorlas]

@@CryptobdSchool Yes, and it's impressive. Similar to Midjourney's V3 or maybe even V4. It's definitely progressing quickly. I think it will continue to be a competitor to Midjourney especially if it remains free.

[@GRUMPNUGS]

Great video, I'm an archi student currently working on a project where I can visualize the scenery and atmosphere I want and I've got the cad model about half complete and looking for inspiration on how to push it further but hit a bit of a block. Learned some things from watching your workflow and I've used my current renders and a specific description of what I'm aiming for and it's given some great results. Gave me ideas on how to keep pushing my project, Thanks.

[@StephenCoorlas]

That's awesome to hear - really glad it was helpful for you 🤘Thanks for tuning in and hope you find the other content useful as well.

[@dailybusiness8162]

I'm sad when i see this. Im an architect and 3D guy myself, been for a while. I remember the hard work we put in, developing concepts, using the tools we had, parametrical or not, being passionate and proud of our creative work. Now this seems of no value. Im definitely not copying something I have not created, becoming the tool of an AI. Yes, call me outdated and the future will proof you right. But I really see a serious disconnection.

[@StephenCoorlas]

I actually agree - There will be a disconnect unless we get out ahead and teach students/young aspiring architects how to use these tools responsibly. There is also the argument that we are products of our environment and our "creativity" is basically the same process that the AI is using by blended inspiration images/concepts to develop new themes. I think we do have a responsibility to use these tool to generate thoughtful/mindful concepts, and that certainly cannot get lost in the process or the technology. Thanks for chiming in to share your perspective. I'm an architect/3D myself, and I see this as a tool that I want to better understand to help my efforts, not replace the value and meaning of our efforts.

[@dailybusiness8162]

@@StephenCoorlas Thanks. Maybe I feel a bit overwhelmed with the speed and actually the high level of AI results. It's going to have an impact. Not only in the creative process of how something looks, but as well how concepts are made, how floorplans are layouted, probably even how things are built, physically. Impossible to predict the future - but I totally agree with you: we need to investigate the possibilities of AI in architecture, to develop technology and to develop ourselves, finding the point of coexistence. Responsibility totally makes sense and we will see - what can be done, will be done. Subbed to your channel, I will curiously follow your videos 👍

[@margretbar]

Its a tool like anything else, the same sentiment was expressed over photography, sampling, etc. Rock'n'Roll is the devils music.

[@StephenCoorlas]

@@dailybusiness8162 Sounds like a reasonable and level-headed approach. I'm in the same boat. My curiosity has led me here, and I want to be a voice of reason before we lose potential great architects to only producing eye-candy architecture 🤘

[@StephenCoorlas]

@@margretbar Yes - exactly. We must embrace it. It's here to stay!

[@nasserahmad4653]

Great tutorial,I been waiting for this video, thanks

[@StephenCoorlas]

Glad it was helpful - we'll do more of these in the future 🤘

[@giri.goyo_yt]

Excellent demo and tutorial. Did you ever release the prompt doc? Would love to see!

[@StephenCoorlas]

I've had several people asking, so I will release it very soon.

[@ingridleibhan8853]

So inspiring Thanks a lot

[@StephenCoorlas]

Glad you thought so!

[@shanechu7654]

Love the tutorial!

[@StephenCoorlas]

Thanks 🙏🏼 Hope it was helpful!

[@rodfarid2056]

Thanks Stephen!

[@StephenCoorlas]

My Pleasure 🤘

[@atajqtsawa3293]

you can setup v5 as default, no typing is needed. /settings

[@StephenCoorlas]

Thanks - yes, this option is now available

[@Skyhook77]

super cool, thx!

[@StephenCoorlas]

Thank you! 🤘

[@keith6371]

I am an AI researcher as well as a founder for an AI startup. I believe AI will encourage human creativity rather than diminish it. AI, like most creative people, are slaves to fashion. Most people simply follow the current most popular trend rather than being truly innovative, this is true whether in fashion or architecture design or art. Avant grade designs are often dismissed as being too out there, but in the age of AI, avant- grade designs will be the most valuable. More people will be encouraged, indeed demanded, to be more original and bold. Young artists will be allowed to be more creative rather than slave away at repetitive non creative works. As they say, this is the best of times, this is the worst of times

[@StephenCoorlas]

I do believe Avant-garde explorations will become more commonplace, although I'm not necessarily an advocate for this to become the primary driver in our society. I believe AI will relinquish us from the repetitive tasks to focus on other more critical agendas related to our health and sustainability of our planet.

[@fynali]

Btw, the inset of the speaker could be reduced to a decently sized small talking head in the corner. As it is currently, it's blocking a large part of the screen!

[@StephenCoorlas]

Agreed. This video was a bit of an experiment for me.

[@audiogus2651]

Love the video! I make environments for games and AI tools are now pretty much mandatory for me to use. Hope to work on creative forms like this one day.

[@StephenCoorlas]

That's actually really interesting to hear - It's definitely being integrated with a lot of visual/creative fields. Architecture seems a bit slower, but it's catching on!

[@thebest3702]

What Ai tools are you using for your environments?

[@patiencegough4288]

Just use /blend rather than uploading each individual image, seems like a long workflow. Also if you type /settings then you can set it to always use V5 so you don't have to add the command to the end of the prompt. Otherwise, great video.

[@StephenCoorlas]

Thanks for the heads up. I've been experimenting with /blend and will be doing another video on that shortly. Also, yes, you can set V5 to the base mode in settings.

[@DanFrederiksen]

It's very impressive what it can do. Some of the buildings would be expensive to make though. Maybe do the images ahead of time, then it's more engaging. And this video is 720p 60Hz. Don't do that :) At least 1080p, for quality visuals 4k is easily worth it. More than 30fps is waste. For your visualization, try using more realistic materials and lighting. Use a renderer that has credible GI. I use vray in max but there are other ways. The medical center looks like a paper model in a dream like absence of light which runs the risk of looking very different from the real thing. Foggy is not a parameter in reality either. Architecture is a design for reality.

[@StephenCoorlas]

Thanks for the input. Yes, there are many ways to approach generating these images, and for many purposes. I'll need to look into my recording parameters so thanks for that.

[@themrgumbatron]

fascinating!

[@StephenCoorlas]

Thank you! Hope this was useful!

[@1Waym]

very interesting thanks

[@StephenCoorlas]

🤘🏼

[@diarmaidupton]

Minor feedback man - please kill the reverb on the mic. Whether is an applied effect or the room itself. Great content 👍

[@StephenCoorlas]

Ah, it’s the room. Just finished building it out, but definitely planning to add some sound panels. Appreciate the constructive criticism 🤘🏼

[@audiogus2651]

@@StephenCoorlas a transient shaper plugin on the audio can help with reverb. (Ps i like the reverb, sounds like a sermon :) )

[@StephenCoorlas]

@@audiogus2651 Ha! I actually thought the reverb was OK too 🤘

[@ANGEL-fg4hv]

You know you can just drag the image into the prompt without opening browser right. Even a right click copy image would work.....hecc even a cimple Control + C , Control + V.

[@StephenCoorlas]

Thanks for the heads up - I don't think I've actually tried that. Sounds much simpler!

[@MikeKleinsteuber]

With respect V5 prefers proper flowing English descriptions rather than the V4, short, comma delimited words, so it may be better for you to re-write your prompts either manually or use chatGPT4 to do them more efficiently. Also, why not change V5 in your /Settings if you're gonna use v5 all the time.

[@StephenCoorlas]

That’s definitely another way to prompt. Yes you can now switch to V5 in the settings 🤘🏼

[@mamatethiam55]

Its a great video, thanks for all the effort. How can i download the prompt pdf ?

[@StephenCoorlas]

Hey - Thanks for watching. If you follow along I'll be posting it to the community in a couple weeks. It's being formatted for everyone's use.

[@davidmullins1706]

Love this tutorial, New to midjourney is there any way can you share a list of prompts you use?

[@StephenCoorlas]

I will share it in an upcoming video. Stay tuned 🤘🏼

[@atmaatma6867]

Love the Video , How is the design process from the overall exterior of the house to the interior homogenized if only Midjourney is used? because prompt creation in MID creates 3D images. Based on 3D images, architecture is still working on converting to 2D drawings (matching 3D images) as a function to use and CS .

[@StephenCoorlas]

Great question, we’re not quite there yet in terms of extracting 2D plans into 3D imagery. But you can certainly use other tools such as Stable Diffusions ControlNet for image consistency across design options. This is really best used as a concept tool for inspiration at this point.

[@lcominghome]

Could you please find a way to reduce the echo in your sound? Thanks

[@StephenCoorlas]

Working on it

[@festusmwathi5064]

Does this tool need heavy pc like graphics card,ram??

[@StephenCoorlas]

No - It's a web program. You do not run this directly from your PC, only through your internet browser.

[@user-qo6nl1kq8u]

hey coach..this is cool video lesson which is very impressive. btw, do you know how to keep the 1st images are stay as it is without any modification or altered by midjourney when merging 2 images together for new midjourney created images ? for example, if you do want to see your building facade with any alternate modification that midjourney created on those facade, then your original facade images has to be untouched and stayed same without any new modification..try to find the way how to preserve those 1 st base images as a footprint images, but midjourney always change it to new creation all the time even 1st image are aways alterered and touched drastically...so there is the way to preserve it's 1st images not touched by the midjourney? very curious to see the answer for that..thanks always. you are my coach ..

[@StephenCoorlas]

Hey, I haven’t seen anything for Midjourney that can do that. Stable Diffusion released something called ControlNet, which does exactly that. I’ll be doing a video on that in the near future.

[@sirrodneyffing1]

Can't wait to see how local authority planning officers deal with this: Newsflash..they wont have a f*ing clue.

[@StephenCoorlas]

Nope - This is going to hit everyone like a tidal wave.

[@aruwamalik]

Hello. My image files doesnt give me the option to open in another browser to enable me copy the image link Help

[@StephenCoorlas]

It should if you right click on the image. Otherwise I’m not sure howls to help.

[@VVVWVVVVWVVVWVV]

is it possible to only re render of picture i upload in midjouney?

[@StephenCoorlas]

Hey - not yet. Stable diffusion has a program called controlnet on their website that allows for re-rendering your image with different effect, but what you've described is what everyone is looking forward to!

[@LouisGedo]

👋

[@StephenCoorlas]

🤘

[@NathanEllery]

I want to be able to upload a basic image that I've created from a 3D package and then AI to enhance that image. I need something from the real world that I can present to my clients, not some wild imaginative piece that has no relevance except wow factor. If I could use my design as a reference image and AI sticks to my structure and only enhances the image, plants, materials, lighting reflections etc that would be a useful outcome. This seems like the most useful and very achievable idea but I'm not clear on if any AI tools can do it yet.

[@StephenCoorlas]

Hi Nathan, I completely agree with you. Fortunately there are tools that can produce what you are describing. So far DALL-E's inpainting tool can be used to erase part of the image and then reproduce new content in place of the erased area based on a text prompt. Stable Diffusion is also working on an add-in called ControlNet that basically locks the composition and geometry in the image and then allows you to generate options on top of the image. Lastly, Adobe just released beta testing on their Firefly suite which looks like it will also have the enhancement features you are describing. Definitely worth looking into.

[@FalcoStarlight]

If you can't beat them, join them... If you can't beat AI architecture, join AI architecture... Peace ❤

[@StephenCoorlas]

There’s a lot of really great potential for these tools, but yes, it certainly seems there is no avoiding the integration of AI in architecture 🤘🏼

[@aceheart5828]

It has aesthetic and creativity. But very frivolous and far too random to be useful. It feels like it deeply lacks function, intent, purpose or interest in how the space will be used. The forms don't stem from building methodology. It is inefficient and would have a high building cost. Modernism began with 'The Machine Age', manufacturing, tectonics and how it came together as a building. It worked because it made sense from the component, up. It prioritized efficiency, comfort & function. It shunned unnecessary craft or decoration and favoured simplicity. It costed less (only spending if it improved the space) It reformed the entire industry. AI begins at the extremity of design fantasy with no boundaries. Leaving you stunned with its generous hand in using all of your budget on a few random strokes.... For it to ever become anything. programmers need to integrate it into the architects work flow, leaving all decisions in the hands of the architect in a delicate interactive process.

[@StephenCoorlas]

I agree - This is certainly an infatuation phase in producing fanciful concept designs, but regardless, the experimentation during this phase will naturally settle into workflows that service the technology - for instance, 3D printing. Yes - Modernism was a philosophy based on efficiency and machined tectonics, but it took decades for manufacturers to catch up to this philosophy, and offer complete architectural systems such as modular ACM panels, storefront/curtainwall systems, and other manufactured building products. This AI infatuation is not unlike the modernism paradigm shift. When the architectural discipline adapted modernism from the machine era it was after Frank Gilbreth's Time & Motion studies based on Scientific Management theory which experimented with many odd and peculiar potential applications for efficient workflow implementations. I believe after the AI infatuation phase, more functional and efficient uses for AI will override the playfulness, and we will be emerged in front-to-back-end AI integrated workflows that conceptualize based on psychology and functional uses of space, then manage and catalog building information, and share the data with digital fabrication processes. This is the promise of AI integrated architecture. What you see here is just the door opening.

[@chrisschuler4119]

There are already softwares that can develop multiple layout options for developments in seconds...a job that would probably take days for architects...now AI doing the concept and rendering...soon architects won't exist anymore.

[@StephenCoorlas]

I don’t think architects “won’t exist” but their role and tasks will definitely change

[@catallaxy]

Well, it looks nice but, who is going to do the working drawings, eh? And you're going to need multi-millionaires for clients. It is also obvious that these results are not being driven by any kind of conceptual design approach. This is merely a new digital variation on Image Transfer Design which, may be o.k. for interior decoration or marketing but, not serious architecture.

[@StephenCoorlas]

Only a matter of time before Revit and other documentation software will be able to apply the images as geometry in your model. Yes - very conceptual for now, but will soon become integrated in designers workflows. Check out Veras by EvolveLab.

[@catallaxy]

@@StephenCoorlas This is not conceptual design. It is more like throwing cottage cheese, Hungarian paprika, capers and butter in a blender to get Liptauer spread. The absence of purpose is noticeable. Without deliberateness, there is no design. I am sure that Midjourney will be good for something like creating backdrops for renderings but, a machine has no motivation or consciousness. It will never know what to do or why. It is not a question of programming, it is a question of art.

[@StephenCoorlas]

@@catallaxy We're all programmed at some level. So even your interpretations of art are merely a reflection of your experiences surrounding the topic of art, sometimes merged with unassociated topics. This is what Midjourney does. and if you guide it well, you can create a vision. Yes, this is a video showing process, not intent. I will have more content on intent in the future - If that's more interesting to you.

[@architect1580]

It is unfortunate that all the architecture that AI generates today is similar to the designs of the architect Zaha Hadid, or she was ahead of time, or she had already developed this technology, so today all the facades and interiors are similar to the great work of this architect. of his video there is nothing salvageable or interesting among his images, it is repetitive and similar to what a secretary can design while having lunch with the help of AI...

[@StephenCoorlas]

I actually agree with this - The monotony of curvilinear design is based on the data sets of images these AI models were trained on, and we will only be able to produce or rather reproduce content that has already been made. I encourage everyone to continue sketching, dreaming, and researching their own ways to develop architecture, and their own design language. To be fair, Zaha did not invent curvilinear architecture, but she certainly popularized it in a branded fashion.

[@davvidu3465]

Few potted plants in a white hall??? That's not a design, this is nonsense 😂😂😂 a 5 yeard old can do better😂

[@StephenCoorlas]

Even a 5 year old knows “Design” is subjective.