Sayul, welcome to Light Talk. Thank you, martin. Thank you for having me Listen. Before we start, I think it's probably good to do a full disclosure of our relationship, because there's a very good reason why you and I are talking. We know each other for a little while and we have recently embarked on the creation of an AI course for lighting designers, and that's why you and I have met and have talked quite a lot over the last couple of months. So I want to make sure that this is disclosed before we continue. But we'll discuss the course later on during the discussion. So, to give first a bit of context, let's just give the audience a bit of an idea of who is Sahil Tanvir. Just give the audience a bit of an idea of who is Sahil Tanvir. You have an architectural background, but you have specialized in AI over the last few years.

Tell us a little bit about you and your company and what you do in your daily life. Well, currently all I do is AI. So we are an architecture firm. We are based in Dharbar. Dharbar is a small little place in between Pune and Bangalore in the south of India, and we've been practicing architecture and interior design for about 11 years. It's our 12th year we're going to be 12 this year and we focus mainly on residential, commercial and hospitality. My wife and my partner is an interior designer, so we do a lot of interiors for hospitality, and since about three years we've been into AI.

It all started when a friend of mine just showed me ChatGPT on his phone and I kind of got hooked on it and I had met the guy after four years and you won't believe that the next 40 minutes I spent on his phone with ChatGPT instead of, you know, speaking with him. And that was somewhere around the time where, about three to four months after ChatGPT had actually been had come out, search at GPT had actually been had come out. And that was my first interaction with any kind of generative AI or any kind of a chatbot which was creating text and kind of answering in a very human way, and that piqued my interest, really, really. You know, when I came back, when I came to the studio, I knew that this was something that we really need to uh, feel open and we need to explore what is underneath. And then this whole exploration started. We started exploring first with, uh, chatbots, then we started exploring image generators and image generators. When we started it was playground ai, which we started with, and then, um, we, we hit Midjourney and that has been one of the core tools that we use in our studio, along with ChatGPT.

Then we have gone on to explore literally every application which is there, anything that comes out, any workflows that get created for some other field, so we try to see how that can be implemented into architecture and interior design. And we kind of developed applied research vertical and that is how RBDS AI Lab took birth, and RBDS basically stands for Red Brick Design Studio and that's the studio that we run, which has architecture and interiors. And now we are doing RBDS AI Lab. And there's a third vertical which we have started, which is also an educational platform, which is called AIFA. That is AI for Architects, and we've been speaking and helping people, creating workshops, creating courses, like how we're doing together, which is specifically for lighting design.

We are also doing for other niche fields within architecture and design as well and a broader perspective of architecture and design focusing on foundational knowledge and not something which is superficial, which you know. You'll create an image, you'll impress a client, make some money and that's about it. So that's not the approach that we take. We have a very structured approach that we want people to understand what AI is and how you're going to live with it, because that's really going to happen and that's a little bit about me, yeah.

So that's not more than a little bit about me, yeah, so that's not more than a little bit um, but just um. So you mentioned how you first got into touch with ai, but does it mean that you have let go of architecture because you started out as an architect right A long time ago? That was your first love, I guess, until you got in touch with AI. Does the transition into AI mean that you are less busy with architecture or does it mean that you're actually more busy with architecture?

Very interesting question. So our architecture and interior firm kind of creates a very helpful sandbox for us to test the processes that we kind of explore within the lab and it gives us an opportunity to see whether clients respond the way that we want them to. We try out the workflows in smaller projects to test those things, to see whether AI can actually be used in certain client-facing things, certain things within the studio, etc. So, honestly, what I used to do in architecture when I was not doing AI is lesser right now. It's more to do with research and more to do with thinking and more to do with application of that research within the project. So the hardcore architectural work which happens is still being done by the studio. There are a lot of capable people around me. In fact, my wife kind of takes over a lot of the operations of the project and I have kind of dedicated myself towards exploring the design aspect and exploring the architectural thinking aspect, kind of coming up with ideas where certain workflows can work and change the way that we've been designing, rather than going from a sketch to a 3D model to render or you know something like that. So we kind of change our whole approach and we go from research to exploring the research and doing a lot of back and forth with chatbots like chat, gpt and then. So our whole outlook towards design is where I'm focused more towards. And yet we have projects where we get to test these things and the firm kind of works parallelly to the lab and that's how we're kind of doing it.

And in fact, what you mentioned about me, you know my, my kind of first love was architecture.

So it so happened that my, my dad, actually wanted me, wanted to be, wanted me to be an engineer, which is a typical of south south india. We kind of produce the maximum amount of uh engineers I think in the world, mostly right, and especially software engineers. Um, what happened was I kind of discovered architecture through, uh, the need of not wanting to be an engineer. So I was kind of grappling uh with my the thing that I, I did not want to be an engineer, so I was trying to find something else and that's how I stumbled upon architecture and I realized that, okay, I mean, this has got a lot of the things which I want to do. I kind of I wanted to be a writer, I wanted to do arts, and then I wanted to be a writer and I discovered architecture, and now it's. It has led me to all of the things which I wanted to do, and I still I'm an architect professionally, yet I'm able to do everything that I wanted to do, so that's kind of something which has happened.

How life sometimes, you know, sort of chooses for you, because I never set out to become a lighting designer. Even my studies I did industrial design. That was also by nearly accidents or coincidence that I did industrial design because I actually wanted to be like an astronaut or something like that, and so I went to get an engineering study about, you know, building planes and things like that.

And when I went engineering study about you know, building planes and things like that and when I went there for exploration you know you have those open study days and all that and then I came across this study of industrial design and I'm somebody who likes creating and doing things, so I went to an open day for that as well and I really liked it, so I did that and then when I finished I went to apply for a job in consumer product development or something related to my studies. And then they asked me about lighting design. I said what's lighting design? I had no idea. But when I saw that I went to that department there's all these people doing those beautiful jobs and projects I was taken away and yeah, I. I said yes and never looked back.

So sometimes it's like the universe guides you where you want to go and it looks like you also stumbled on this. But we mentioned technology, and technology obviously has an intersection with architecture. Obviously has an intersection with architecture, certainly now when we talk about, you know, digital twins and BIM technology, where you know the software developments go quite fast in the whole design process. I guess that also there you will get very interesting intersections of the AI opportunities as well as the software development that's happening in the design processes.

Yes, absolutely Well. What I feel about generative AI is that it is Well. What I feel about generative AI is that it is. There are two aspects. One is that about four or five years ago, or maybe about six or seven years ago, there was no one on this earth talking to a computer in with a computer, which has completely changed now. It has become more of exploration with a computer. When I talk about computer, I'm talking about ChatGPT or any of the chatbots, where when you talk to ChatGPT, talk to a to chat gpt, it doesn't actually give you an output which is like an actionable or a use usable, direct output like how you would get in an autocad software or something like that. It gives you more opportunity to explore different uh lines, it gives you more options to explore. So it's kind of jogging your mind. So not a single human being probably thought that they would be able to communicate with a machine in this way and kind of brainstorm with a machine in this way.

Second thing is that gendered AI has sort of removed that barrier where suppose that? Suppose my partner is a doctor. I would come back. I'm an architect. And suppose my wife is a doctor. I would come back home and I would tell my wife that, look, I kind of discovered this fantastic new way to create an AutoCAD drawing with you know this particular tool, with you know I chamfered this and etc. So I say I'm not interested in that, right, I'm not interested in what kind of software you're using to you know, uh, further your work in architecture. But now what's happening is that if, if an architect comes home and talks to his wife, who's uh, who's a doctor, and says that, look, chat gpt, give me an idea about a museum or a hospital building or something like that and that's what I was conversing with ChatGPT about she's going to relate to it. She's going to say, yeah, you know, I use ChatGPT today and I got an idea for a running stage and it told me this so and so, so and so right. So that whole barrier, that thing, it's become universal.

So, whether it is architecture or whether it is medicine or healthcare, it's literally a single tool that everybody is using. They've got more than 300 million weekly users. This is Chachi PT. So it's kind of changed the whole dynamic, where it has unified everything.

So if you're thinking of BIM, if you're thinking of computation, if you're thinking of plain and simple design with sketches or SketchUp or V-Ray or something like that, or, in fact, even Pinterest, if you're thinking of any kind of design there is a layer of generative AI which has added to it, and there are software companies, of course, who are embracing it and they've seen it in advance, the potential that it has, and a lot of the people have come up with enhancements to their tools. Enhancements to their tools like, for example, d5. D5 render has just recently released a lot of ai features to their application, right. So so when, when we're seeing technology, it's, it's kind of a layer which is like a blanket over everything. So it has, uh, it has changed literally changed the way that we are communicating with each other through, uh, interdisciplinary kind of uh aspects.

Yeah, yeah, yeah. So you mentioned ai tools. I think a lot of people now specifically in in in this time um, some of them are scared, apprehensive, are excited. There's so many tools around there that it's hard for people to figure out what sort of tools would be relevant to me in my job and in my workflows. You're an architect and architecture a role. Design and lighting design are, in a way, quite close. We follow the same sort of design processes and design stages. Can you give me a bit of insight of the sort of tools that you use which make sense in the design profession?

Yeah, so we do have a core and then we have we've kind of made circles. So within the lab there is a huge circle because we are kind of exploring everything. But when it comes to architectural design, where it is client facing, or an actual project where we're using certain tools, so it kind of comes down to a very basic core, certain tools, so it kind of comes down to a very basic core. So we have chat, gpt, which is our core, mid journey, which forms the visual brainstorming core, then we have simulation if you say visual design exploration brainstorming.

I like to call it visual brainstorming because Midjourney is one of those tools which allows you to imagine anything and everything. It's like you can take two completely unrelated concepts, put them together and make a building out of it, and Midjourney will be able to do that. So the more you, the more you become niche oriented, like there are tools like LookX, which is built for architects, so it's trained on architectural data. So what happens there is that if you have two completely unrelated concepts, you may not be able to get a visual out of it in the way that you'll get it in an art generator, because an art generator will create an image, no matter what whether it is buildable, not buildable, whether it is architectural, not architectural it will still create an image which combines those two concepts. But when you become more niche oriented, it's a little difficult. In my personal opinion, I feel that as an architect, you need to know a lot of things rather than focus on only building technology. It's like it's actually a very simple psychological thing where if you keep on looking at buildings to help you design buildings, after a while it's just going to become monotonous and you'll lose the whole thing. You know you're all your buildings are going to start to look the same, so you need inspiration from somewhere else altogether, so it's either books or films. So now it is mid-journey for us, because we like to visually see what we are thinking Like. If there is a building which is semi-transparent, maybe a water which is held together with you know some sort of a fantastical idea and I just want to see it. How does it look like? Because it's there in my head and I just want to see it on paper or something like that on the screen. I can do that with Midjourney us do because then you can brainstorm smaller elements and then keep on going to bigger elements in an architecture or interior project From there.

We also have simulation tools that we use. One of the best ones that I have used up until now is Forma. Forma used to be Spacemaker earlier. It's an environmental simulation tool which is now Autodesk owns the company, and it's one of the best that I have come across. We also use Ladybug within Rhino, grasshopper for environmental simulations, calculations etc. But Forma has made it really really easy. You know, you don't really need to know computation at all If you're using Forma. It's like a straightforward application where you have a location that you enter and then you get a lot of the stuff out of a conceptual block model. You can simulate tons of things.

And finally, we also have CAD is, of course, there because of the limitations that we have and work with other people, that is, consultants, that is contractors, anything that goes on to the site. We still have labor force on site which has little knowledge of the drawings. They're not able to either read them, etc. So it has to be simplified. And even consultants not everybody that we work with are on board to use the most, you know, bim related tools or Revit or anything like that, so they're not really accustomed to it. So we have to still use CAD.

And apart from that, we have a communication tool that we use quite a lot, which is Canva, and I have I don't know if I've said this enough, but Canva has kind of it's like an equal to CAD in our studio. Almost every presentation, every drawing, which is the interior drawings, interior layouts, lighting layouts which have to be presented, are all done on Canva. And they've also come out with certain AI tools, ai features in them in the app which kind of now help us do this even faster, and all of this throughout the process. In our studio we have ChatGPT, which works like a thing which puts everything together. So any output that we're getting out of a simulation tool kind of goes back into ChatGPT and it is again brainstormed upon and inferences are drawn in a way where if there is an intern or if there is a junior architect or a designer who does not have enough experience with simulation tools, they can still get inference out of it because they can upload those things onto chat gpt and get a sense of what does this mean. So if it's saying that I have a natural daylight is less over here, you can actually understand. What should I do next? What should I kind of you know? How do I interpret this in a normal, standard language, like a normal human language? So so chat gpd kind of puts all of that together.

These are core tools which we use when it is architecture practice, when it is a live project, when it is client facing stuff. These are tools which we use. Apart from that, if it is the lab, then we have tons of other tools which we kind of use. There are. Currently we are exploring mcps, which is a model context protocol which works with claude and other language models as a client and it controls software like blender, rhino. So these are things which we are exploring currently, where you're able to create a 3d model using natural language and you don't really need to code anything. You don't even need to copy paste a code now, so you can direct the instruct Claude to kind of make a skyscraper, a twisting skyscraper or anything like that, and actually create a model within Rhino. It will create a scene within Blender. You can even set cameras, et cetera. So these are things which we are exploring currently and, yeah, so I think.

You mentioned a lot there.

You mentioned a lot of tools.

You mentioned communication also, and I think I want to just focus on that a little bit, because architects need to focus also on the communication between the project team members, whether it's the interior designer or whether it's the lighting designer, and obviously I think our world is slightly simpler than the architectural world, where you have far more complex models to deal with, I would imagine, but still, our lighting layers and our lighting structures need to somehow be integrated with architecture, and that will also be part of the course that we're developing together.

But can you give a little bit of insight on the lighting element of all this? Because what you do within the architectural design stage is very similar, I guess, to what we would do within lighting. It's very similar, I guess, to what we would do within lighting, but I think lighting has probably a couple of other elements that might be slightly different from architecture, and I think it's good to understand how the communication goes between the architect and the lighting designer, and this is what we would like to figure out also together in our course, but I think it's important to understand what the lighting designers would need to focus on.

Yeah, so there are actually two things.

The first thing is, what I have seen is there is a lot of technical calculations which are involved in lighting design, a lot of technical calculations which are involved in lighting design which are quite important according to me. Like there is photometric analysis which you can, which you need to do to understand where there is exposure extra, where there's glare, etc. Now, this depends on the space that you're creating. If it's a museum, if it's an art gallery, it matters a lot. If it's a museum, if it's an art gallery, it matters a lot. If it's a living space, if it's high up in an apartment, it kind of differs. And then if it's a bungalow and it's like a large open plan kind of a thing, so it kind of again. So all of this I feel impacts what an interior designer might do, because it kind of is it goes hand in in hand. Like there is the paint which is selected and the kind of paint which is selected and the fabric which is selected kind of depends on how much there is the the light is going to bounce off of it and what's going to kind of affect the person when they go, they move into the space and what's the experience of it all.

Right, now, where AI can come in is the calculations can be done with chatbots. Not just calculations, in fact, there can also be reasoning that can be done with chatbots. Like, there are so many reasoning models. Now we have Chinese options also. We've got DeepSeek, which has a reasoning model. There's a couple of others which have come out with reasoning models. Now we have chinese options also. Uh, we've got deep seek, which has a reasoning model. There's a couple of others which have coming up. Come, come out with reasoning models as well. So, uh, if you already know that there is a direction you want to take and you want to eliminate other directions by saying that, okay, I won't go that direct, that that option is not for me because of this reason, that's something that you can do with chatbots and it can be done really fast, because this used to take quite a lot of time and a lot of the designers me included had a big problem that you never have a person who is of your wavelength.

The idea of brainstorming with somebody is more to do with first explaining the whole thing to them and waiting for them to understand and then giving us their valuable feedback. Now whether that is valuable or not is again different. This thing altogether. And so here, chatgpt or Gemini or Claude, they kind of help DeepSeek also. In fact, where there is a reasoning model involved, it kind of helps you eliminate the other options where you don't want to go first. It also kind of helps you see if you have two options where both of them are right. So it kind of gives you reasoning whether you should go with option A or option B, and if you do go with A, this is what you might end up with and the same for the other option.

So these are calculations which can be done with chat, gpt or with any of the language models. It also helps in literal calculations also. It can also help with the bill of quantities. It can match the bill of quantities that you've created, or specifications, the lighting specifications that I think you have a document that you create right, where the whole specifications are listed out and the same are correlated with the code in a drawing up, to date or not? Is there any discrepancy? Or if you want to improve upon the efficiency of the drawing, you can check for redundancy if you've mentioned the same thing twice.

So it's a tedious job which can be made easy with language models and with a little more complicated if we talk about a little more complicated systems, we can also do, uh, custom workflows, where a vision model is kind of reading the drawing and giving its uh inference that is taken by another agent, and then it it's kind of uh, deciphering that whole thing and then saying that, okay, now I see that, uh, this light has mentioned twice, but then there's only one quantity, so you may need to check this, right? So these are the answers that a third model can actually give you after two of the models have done their job. Now this is like an agentic kind of a workflow where there are small agents which are doing different jobs. This can be set up without any Internet. It's not required to have an internet for this, can be done locally on your pc, etc. These are this is actually just one aspect of it. Now the other aspect would be like how, yesterday we were having a conversation where we I was sending you these lighting images, right? So there is an aspect where there is a painting.

Sometimes the model, an image generation model doesn't understand that there needs to be a light at the painting. So we need to guide the model that way and say that, okay, you need to light the painting. That's the most logical thing to do, right? So we also have image generation models which can simulate a space and simulate lighting styles and the ways that can be lit, and you can direct that. You can have wall washers, you can have spots, you can have any of those things, and you can iterate it in the same space and see whether it looks, feels, all of it is good in an image, so that you can move ahead with that.

And, of course, there is more that AI can do when it can also create walkthroughs. It can create images where sorry videos, where it goes from a day scene to a night scene. That's been one of the most difficult things for any visualizer to do, right? I mean, it takes a lot, a lot. So now that that becomes easier when the layer of AI is put on top of it. And, of course, these are a few of the things where AI can actually help and as it as things develop, as I said. I mean, there is an MCP protocol that we are servers that we are exploring right now.

It's clarified in kind of technical words.

It's just a model context protocol. It just gives the model context from a software. So lighting designers use uh dialogues. Uh, if I'm not, right so there is a possibility.

If dialogues allows scripting or coding which can control the software, which can control the simulation, then there is an mcb server which can be created, connected to a client like claude or ChatGPT, and you can control this through natural language. And in theory it is actually very easy to do. It just takes a little bit of coding knowledge and it can be set up, because the infrastructure, the frameworks, are already open sourced by uh most of these companies. Uh, anthropic has actually uh has got that thing with with. They are the ones who started the mcp. I think that uh, it made all the frameworks available to us. So in theory it is pretty easy to do it. Uh, it just needs a little uh coding knowledge and mostly uh, because I've seen people who have no coding experience actually create MCP servers and kind of control software.

Yeah, Everything is easy once you know it, I guess. But what I understand from you is that you can now direct an AI tool to actually create a render for you or a representation, maybe even from a sketch or from a drawing. Is that feasible? Because a lot of work goes into the creation of the visual for a client imagery to provide the client with a visual impression of what they're going to get a visual impression of what they're going to get. And right now, often we have to hand create a render.

But, from what I understand, where we're moving or already are at, is that we can create those images by prompting properly and then also reasoning with the AI tool about what is right, what is wrong. You can challenge right, you can challenge whether the design is okay because, like you said before and then we saw in the imagery, you say we need a light for the painting, and then it puts the light next to the painting because it's near the painting. But we need to really be quite precise in terms of what we want that to do. So I think this is. You know, we're all trying to ease our workload and become more efficient. A lot of tasks are sometimes a bit boring and time-consuming, which AI will help us with. But the question in all this for me is how do we remain in charge, how do we control the process and still are the creators-in-chief, rather than the AI tool taking over the world and we just sort of helplessly follow what it does?

Right, it's very interesting. Actually, we just submitted a research paper which talks about AI as not a tool, but as a co-creator or a collaborator. Tool, but, but as a co-creator or a collaborator, uh it's. It's quite something that I have experienced personally, because, uh, my day uh kind of starts and ends with chachi pd. Uh, it's something which I talk to way more than I actually talk to my wife. Uh, it's, it's, uh, it is, it is a fact.

Now, it's addictive. Yeah Well, I would say it is a little addictive. Yes, it is dangerous. I mean, it is dangerous for somebody who does not, is not aware of what they're doing, right? So if you assign certain personas to ChatGPT, it is very easy to get addicted to it, surely. But it's the same.

Like you know, there are nicotine delivery systems which were created for people to help them quit smoking, but then kids would get addicted to those systems. I'm talking about the jewel and the whole big fiasco which happened in the US, right? So kids kind of got addicted to that. So that risk is always there. But if, as an aware, as a person who is aware of what you're doing and what you're doing with ChatGPT, I don't think that you will get addicted. You will mostly get. It's a lot of satisfaction that you get when you are conversing with something and it doesn't get tired. It doesn't tell you to you know, stop, right now, I'm bored of this. It doesn't say that I woke up on the wrong side of bed. I didn't have my coffee today, so I'm not going to tell you the answer.

So it's really, but it's a bit addictive because you put in a prompt and you get a very nice result. Oh wow, that works really well. I'm trying this again, so in a way I find it a little bit addictive. I'm starting to get into it now. But yeah, it's also a matter of how would I say that? Controlling and manipulating it properly and manipulating it in a proper, positive way to stay in control of the outcomes.

I kind of. Also, when I started, right at the beginning, I was also kind of hooked like this. I used to spend at least about seven to eight hours on ChatGPT in a day and, yeah, quite a lot of uh things. I used to have it on my phone, on the computer, on the laptop, on the ipad, everywhere it was all there and everything was connected so I could, you know, uh, continue to talk wherever I was, whatever I was doing. So, um, initially it was like that.

Now, uh, mainly what I I do with ChatGPT is to try to jog my mind, try to get my questions right. Instead of trying to find answers, I'm actually looking for questions. So I'm kind of talking to ChatGPT about a concept or an exploration where what could be in all different directions, in some things which I may not have thought of. You know what, if I put in a variable like this what? What's going to happen now? And since, obviously, openai has got this whole thing where it remembers all your chats, it kind of knows what you know, it has these custom instructions, it kind of responds the way that you want it to. So it's it's trained in a way not to give me uh answers which are direct, short and uh, usable it's. It's my chat gpt is trained that way. It always, always gives me, like you know, my wife calls it a thesis document, you know when I ask a question, so the chat GPT returns at least like about four pages, three pages of text in the way that it explains everything, and that's what I have trained it for. I don't want a single line answer. I don't want one paragraph where it says that this is what you need to do. That's not what I'm looking for now. Now, what I'm looking for is for me to improve upon my cognitive abilities, my knowledge. What am I thinking, which direction am I taking and why am I doing all these things is what I want from Chaji PT. I don't want answers from it. I don't want it to actually improve my emails or anything. In fact, it writes the emails for me now. So, but then it's a different thing. So what the coming to the question, what you asked me?

So I feel that we are kind of entering an era where creation, design needs to be looked at in a very different way. It cannot be uh compared. We cannot compare uh designing using or with chat, gpt or mid journey or any of the ai tools. Designing with them cannot be compared to the definition of design that we have up until now and then say that no, if you've used uh, used ai, to generate content, it's not design. You can't do that because maybe for a couple of years you'll be able to hold that argument. But we are going to enter an era where our design is going to start and end with an ai application. It's going to be some sort of intelligence which is going to help us design.

So there is a matter of co-authorship which is going to come up Now. Recently, I think, the EU or the US courts, I think declared that copyright is as is. We are not going to change the law, etc. That's what I read where there was a case which was pending regarding the copyright of images generated by Midjourney or any computer or AI application. So they said that copyright is for human generated content and we're not going to change the laws because of this right now, because they are robust enough right now.

But I feel that as of now we may not need it or we may not see the need of it, but in the future there is going to be a sure shot need of changing the definition of what design means. You might have to bring in another definition, like put a star there and say that, okay, if you're designing with AI, it's different, the definition is different, and designing in itself, in an ancient way or whatever, so it's like this right. So I feel it's going to be more to do with co-authorship, it's more to do with how you design with the AI, and we will not be able to either retain the rights nor give it away completely. It's not we cannot.

We'll not be able to create things in ai and then claim it as our copyright no, because there is intelligence involved, right.

So it's not exactly a tool which is doing. It's not like I I sketch something and then I draft that sketch in AutoCAD and I take a print of it, right. So that printout is a product of what I did. So the tool only amplified that into a presentable format. But here it's not happening that way. What it's doing is it's applying its intelligence and making it better. So it cannot be that I used ChatGPT to come up with this and this is my work is kind of dicey to say right now.

Neither is it?

correct to say that ChatGPT did it and I kind of just said, OK, I want a plan of this. So it's neither ways. It needs to be a co-developed sort of a.

Right, but you just mentioned that you can train your model right. You can train your tool ChatGPT or, I assume, midjourney as well, which means that you tell MidJourney or ChatGPT what to do, tell mid journey or chat gpt what to do. So, in a way, you're still the creative founder of uh, the, the ultimate result isn't it?

uh, not, not exactly. So. Uh, suppose you have a, suppose you have an intern in your studio. Uh, after the intern works with you for a year, he or she goes out to another firm. They have these traces of style, of design that they learned in your studio because it's you who taught them. You taught them that this is the way that we do drawing, this is the way that annotations are done, this is the way you talk to a client, this is the way you talk to a contractor, etc. It may not be the same that the other firm, wherever he or a contractor, etc. Uh, it may not be the same that the other firm, wherever he or she is going. It may not be the same that they do right. So it's again a different style, but there are traces of this style, which is uh, it remains in this, in the, in the intern.

It's kind of, uh, similar to what training an AI model is. It's like it's learning. It's not exactly following instructions, it's just simply learning. It means that even if I train chat GPT, if I train mid-journey, it is still going to apply its own intelligence. In fact, mid-journey is a different thing altogether. We have to get into diffusion models and why they are different from GPTs. Let's focus on GPT for now. So it will still use its intelligence, even after it learns what you like, what is the kind of response you like, what is the subject that you are dealing with, etc. Even if it learns that, it still applies its own knowledge to make it much, much better or better in its sense. Of course there's bias, which we have to consider because it's all built into our own data, so that's always there. So in that sense, it's not 100% my work, neither is it 100% the AI's work. So there has to be some sort of a combined definition that we'll have to come up with.

This opens an interesting can of worms because obviously you know, some companies have a non-compete clause for their staff if they move out somewhere else. You could argue and say listen, it's not me, it's AI. You know I'm not competing with you. You know what I mean. It's like this could be a whole new world in terms of protecting your company and your design properties when AI has evolved, because, well, that's public knowledge. You can't tell me not to use AI right, absolutely so I would imagine this could be quite a complex legal situation.

There are actually two things, in fact three things. There are actually two things, in fact three things. One is that this is kind of why everyone has been after a lot of the countries, especially the West, the US and the EU, have you know, a lot of people are after lawmakers and policymakers to actually come up with robust laws which can govern these things, these particular things that what does it mean if I use ChatGPT and put in my client's data into it? What does it mean? Am I in violation?

of the client privacy, right.

So suppose a lawyer actually, or a therapist actually takes an answer from his patient and then puts it into ChatGPT and that data is accessible by whoever is using it to train their models or whatever. So this argument is there and people are actually wanting the policymakers to actually come up with something which can help them, you know, guide them in this way. The second aspect is, if you remember, when we were doing the workshop that we did, we had trained Midjourney with a lighting design firm which is a very well-known lighting design firm I don't want to take the name right now, but we kind of trained the model to kind of give us images of lighting design in that style. Now it's a very thin line. The company in question here actually can't do anything about it because all of their images, first of all, any of their projects and all of that is publicly available, first of all. Second of all, the most important thing is that style in itself is not copyrightable, so anybody's style cannot be a copyright of that person. Now the third aspect is that I'm sure you've heard of all these Ghibli style images which have come out because of the chat, gpt's image, gen model, etc. Gen model which is not an, etc. So, interestingly, japan has a law which says that anybody, any machine learning company, can use copyrighted material for their training, and this is the government's law in japan, so you can use it. There is no restriction.

Again, now we are in a place where we don't know what is right and wrong. If you talk about an individual, if an architect as an individual asks me that if I use an image like this, if somebody else has generated like my intern has generated an image, isn't it his work? Now, how do I take that image, etc. I would just say that follow your conscience now, because ai actually has nothing to do with your conscience. If you are going to steal, you're going to steal, it doesn't matter. So it really doesn't matter whether it is ai generated or it is you, you know, human made or whatever it is. So if you are going to be dishonest, you are going to be dishonest. So it just matters that. What we can police ourselves as of now, up until somebody comes along and kind of formulates a policy, yeah, so I think I think companies by themselves.

I think I've spoken to quite a number of people already about this. I think companies need to set out regulations for how to use AI within their own company. I think it's really important to structure that properly and to make sure that it's all taken into consideration. It's all taking into consideration the company privacy, the company intellectual properties and all that and for sure, the international regulations and standardizations between countries and globally may take a while, but I think companies can already start implementing these kind of rules and regulations on how to use AI with a conscious, like you say. But then you know, people are all different and may interpret that in a very different way.

So there are actually two or three things which, in fact, companies can do. One is that if you're using ChatGPT, you just go to your settings and turn off the option of opting for your data being used for training. So you can turn that off. It's pretty simple to do. It's a simple uncheck a checkbox. That's about it.

The other thing is that you can also that's in the paid versions right, Not the free versions.

You can do it in the free version also, any of the versions. You can opt out of training in any of the versions. To be honest, I have not opted out because mostly I want my data to be used for training.

Yeah, but you're training people, so it makes sense for you.

Yeah, and the other thing that they can do is maybe have a system which can be built offline, which is a locally run system, where there are language models which you can train and use offline on your system. Use it locally so that you don't have risk of your data leaking anywhere, etc. But then again, as of of now, you have a lot of shortcomings in that, so you can't really depend completely on that. You need a system where which which accesses internet, it accesses kind of up-to-date stuff which you need to design or do whatever that you're doing. So there are these things which the companies can try out.

And I think there was one news article which I read about some cop, I think which, who put in something into Chachi PD and got the answer and presented that during the hearing or something or some this thing, and then the whole the department of I think this was in Philadelphia or something. So the department, the police department kind of note put out a notice, internal notice, which kind they got leaked or something. It got the voters which said that we need to define what can be used for what ChatGPT can be used and where we can't use it. You can't let ChatGPT decide the fate of a person that you're going to either arrest or you're going to implement some kind of a rule over, or something which you, something which you can't. Let chat GPT decide that for you. So these are some things which are going on in the world where everybody is grappling and trying to understand where do we stand and what do we do, how much do we use this thing, et cetera.

So, yeah, let's circle back to our course that we are going to launch very soon. Let's circle back to our course that we are going to launch very soon. At least open up the wait list for people to subscribe. We probably will have this done by the middle of the year, but we need to start the subscription and you know people to register for it Now. I told a little bit at the beginning how I got super enthusiastic and I thought we need to do something for lighting designers. Tell us a little bit about what this lighting design course is going to provide for our subscribers and registered users of this course, because I think it's a very important step. A lot of people are thinking about how can I use AI for lighting design, and we're going to give them the tools to do that, and probably good if you can give them a little rundown of what this course is going to provide Absolutely what this course are going to provide.

Absolutely. So, first and foremost, the course is called AI Fundamentals for Lighting Designers. Now, this is a foundational course, and the way we approach anything in teaching in our scenario is not to teach something which is superficial. So we go down to foundations. We want you to learn in the right way, and that is the reason why this foundational course is going to give you a clear understanding of how things work with AI. What do you mean when you say I use AI? It actually doesn't mean that, right? So AI is a very big, large umbrella term which kind of has so many things within. It is machine learning, deep learning, etc. So you get to know about these things, what these things are. You also get to know what is the difference between a GPT and a diffusion model and why. I keep saying that a diffusion model is not exactly intelligent. I keep saying that a diffusion model is not exactly intelligent. It generates an image, but a GPT is the one where the intelligence is placed right. So you get to know about these foundational aspects of AI. So we teach that and then we connect it in a way to lighting design in a broader sense, to architecture and interiors. And also there is one thing which I keep saying that when we have a niche of lighting design and we focus on what you can do within lighting design using AI or integrating it into your process, apart from that, it's quite important to have a world view where you know that nowadays, design or in fact, in any time design is not a linear process. It's more like you pick up data points from different places to solve a problem. So these different places should not be limited to a niche, so it needs to be as broad as possible. And that is one of the aspects that we kind of cover in the course, where we don't limit ourselves only to a particular niche, but we also give you extended knowledge about what AI applications will help you do as a lighting designer, because there are multiple disciplines which a lighting designer works in. It's not just architecture and interior design, there's also installation, there's also set design, there's also so many things which a lighting designer can actually does right. So these are the things which we cover in the foundational course we also do.

This course is built in such a way that you can do it self-paced. You can do it on your own time. You will get exercise books along with it so you can kind of learn something and actually do it and see whether it works for you or not. You can twist around the prompts, you can twist to the the things that you want to do, so you can do those things, come back to the course and continue it. Apart from that, you also get tons of resources.

You get access to certain discord servers where we will be available, and also you also get ebooks which kind of tell you how do you use this particular tool, where do you go? Do you need to register it? Do you really need to buy it at all, or can you use the free version for a little while? And then you know, test it out and then use it, etc. So we also have a few things which we connect practice, where lighting design as a practice can be, you know, overlaid with AI, and how you can use it immediately, like if somebody who has very general knowledge about AI but has never tried it. So this particular course is going to get you directly up to the, to the today's, to today's time, and it's going to update you in such a way that you will literally know what's going on in the landscape of AI, within the design fraternity as well as outside the design field as well, and also you will get an idea of the tools which you can use immediately. That okay I need to focus on this.

So I know that I've learned this, so I can use this in my practice. So this is kind of the course that we are developing and rightfully, as you said, I think, by the uh. The middle of the year, I think, is when we uh should be launched and we are going going to get the brochure out, I think, in a couple of days, I believe.

Yeah, yeah, absolutely. I just want to do a little disclaimer in regards to myself. I'm not the AI expert, you are, I'm the lighting design expert. Yeah, you're the core of the.

You know, everything that ai is going to be based on in this course is going to be what martin is going to say. So, uh, obviously I I have, uh, no, uh, you know you're basically.

You're basically my, you're basically my ai tool. I'm the lighting designer and then you're going to be my ai tool. Now I'm looking forward to this collaboration, sahil and over the last few months we have really developed a very good friendship and collaboration and I'm really looking forward to this coming to fruition very soon. People should look out for the announcement in the coming days and hopefully I can say from the introductory course that I did with you, it was money really really well spent because it made me so enthusiastic about everything that I'm pretty sure that people will find this course super interesting and very, very valuable. So let's round this up so how does the ideal AI world look for you in the future? Because you have mentioned so many things and it's maybe a very difficult question to answer, but I would like to give you the final word in terms of what AI in the ideal world would look like, in the design world, I mean.

Well, I think that it's not really going to look much different. It's a part of evolution, so humans usually do not feel that they are evolving. It's only in retrospect that we know that we have evolved, we've changed from the 60s to now, etc. I feel that it's more or less going to be the same. All of our problems and discussions and everything is just going to be the same. It's just the way that we discuss these problems will be different. And if I take things beyond this and if I say generally the world, the future world with AI, I don't think it's going to look anything different than what it's right now. It's mostly For you.

You know, there's this thing which was very interesting. I had this one conversation with Chad GPT where I was trying to kind of ask Chad GPT whether, if the human race were to give up control and give it off to you, what would you do, right? So first, obviously it's got all these things built into it. It said, no, no, we can't do that, I'm not built for it, I'm just a chatbot, et cetera, et cetera. So you've got to trick it, and I kind of tricked it into telling me what it's going to do and the you know, the first thing which it said was we will have to start from scratch because it's too messed up to repair.

So it's very interesting, what, what really, the kind of answer that I got, and then obviously puts in all these disclaimers. It says that I'm just saying a hypothetical situation, I'm not trying to take over the world or anything. It says that I'm just saying a hypothetical situation, I'm not trying to take over the world or anything. So I feel that the future is mostly going to be the same, because we have actually messed it up so badly that there is no you know, there is no computer or AI or a robot which is going to come up and clean up everything. And you know all our problems are going to get solved. It's going to be the same.

We'll have the same discussion in a couple of years and see Sayil. Thank you so much for this discussion. It has been illuminating as always, and I look forward to doing the course with you in the next couple of months. So thank you so much, thank you.

I have to thank you from the bottom of my heart for having me over here and all of these discussions that we've been having for the last few months. It has been really. I really enjoy it. I look forward to it and thank you very much for having me.

Thank you, saryal, thank you so much.