Alessandro "Aledelpho" Delfino is an AI artist that has seen and done it all: he's worked as an illustrator, designer, and graphic design professor. Now, he's the founder of an exciting startup in AI. In this exclusive interview with Realm AI, he shares his story into the magical world of AI and the many things he's learned along the way.
Realm AI: Could you please introduce yourself and share how you developed an interest in AI art?
Aledelpho: Sure! I'm Alessandro Delfino, I got an artistic degree in the Academy of Fine Arts of Brera, in Milan, Italy and I've been an illustrator, designer and then Graphic Design professor.
Let's just stay that when AI generated art appeared quite from nowhere I jumped on that hype train to the point that I left my job and started to mainly focus on this subject. For me the technical representation of my dreams and ideas was the most time-consuming aspect of the whole creative process and as soon as I found this new technology... It was like a dream becoming reality!
Unleashed creativity without worrying about technical requirements and preparation... AI art is objectively nuts!
Realm AI: Impressive resume you've got there, Alessandro! And the fact that you decided to focus on AI generated Art shows that you see a future in it which brings me to my next question - what do you think the future holds for AI Arts? I know you don't think it is just a passing trend like many do, correct?
Aledelpho: Well, I don't really see AI Art as something that can be left behind.
As one of my ex professors said: "You don't go back to the horse when you have a car" - It can't be a trend, if people get something useful out of it, with a much smaller effort than before. Traditional art will remain, but I believe it's going to be more of a hobby, something to express yourself in a personal way. But for me, AI art is going to become the norm of the professional creative sector and I'm very optimistic about its development and future.
It's surely hard to guess what the future could hold for AI Arts, since they're evolving so quickly - but I believe we're going to see some interesting new approaches in the next months.
Realm AI: I see what you are saying - I won't want to go back to the old ways either.
The next months are really going to be interesting - provided that nothing holds back the development we are seeing in the space.
Now, one of the causes for concern for people generating art through AI is copyright and intellectual property rights. Do you have any view on this? If yes, how do you navigate the complexities of ownership and attribution when collaborating with AI systems?
Aledelpho: You know, I was going to write a long reply to this complex question, but I think the thing we should all agree is:
The main lesson Art gave to the world, through time, is that it's not about the technical qualities of the artists - it's about the quality of their ideas, their creativity, their choices. Maybe we reached the point where an artist can be defined as such, just by using someone else's technical abilities and creating something new through that.
Said so: I believe we'll have to develop more "ethical" models, with time.
I'd be honestly happy to train a model on my manual artworks and as soon as I'll have the equipment to do that - I'll go for it!
(This is a vector Illustration I made last year)
This is cool!
Now that you brought up the talk about model; you caught our attention with your models - one of which has been downloaded 7k+ times. What inspired you to create your own model and how has the journey been so far concerning that?
Aledelpho: Well, I have to admit that, with time, I understood that "drawing" was not something that was giving me joy - and probably never was.
Now, I know that it could sound weird from someone that studied in art school and worked 10+ years in this sector, but ... I always liked to draw just for the joy and excitement that I was offering to the viewer. While I was drawing I was forcing myself to finish it because in my mind I was repeating: "I can't wait to show this to X - I have to do it!".
Creating models for me has become the next obvious step in that direction.
Since I liked to bring joy and happiness, freedom and creativity to other people, I now care more about building the ultimate "Comicbook" model or a fantastic "Object oriented" one, than drawing something myself.
I can finally share my "taste" in picture generation, fine tuning a model in order to help people share their fantasies and ideas with the world.
Realm AI: Oh, I see. On behalf of everyone that has used your model, I say thank you! 🤝
Back to the question - how often do you try out new generative AI tools? And have you tried our app, Realm? If yes, what do you think about it? If not, would you be open to trying it out?
Aledelpho: I have to try a new generative tool - you never know what you're missing, if you don't explore what's hidden in the fog around you, right? (not literally)
I've always been more interested in softwares that gave me the possibility to generate as many pictures as needed, milking my poor GPU instead of some external servers. I'm forced to do this because in order to generate a final Checkpoint I have to create a lot of pictures and compare the different results in order to choose what combination is working better before to try and mix a new model - and I have to do this A LOT of times in order to shape a final version.
I usually end up with 2-6k pictures generated and that ... could be expensive on some external providers.
Said so, I've heard good things about Realm and even though I've not tried it yet, I sure will in the near future in order to see how it works! I'm curious and I think people should always take time to try new tools!
Realm AI: Of course, people should always take the time to try new tools. It is the best way to know what works for you.
Thank you for the kind words about Realm. Let me know what you think after testing it.
Are there any ongoing projects or collaborations that you're currently working on and would like to share with our readers? We'd love to hear about any exciting ventures on your artistic horizon.
Aledelpho: Of course, I'll gladly give you my feedback on it!
Well, about the projects, in fact there is a pretty big project I'm working on. I'm one of the founders of an Italian startup that currently aims at creating the ultimate tool for Stable Diffusion.
We are already working on it and the first beta should be released in the next month - if you're interested in knowing more about it, I'll gladly share more information about it as soon as I'm able to!
Let's just say that we have a pretty nice team and I'm so excited to share the result of these six months of hard work with the world!
Realm AI: Sure, I'd like to know more. I can't wait to see what you are building.
I have two more questions for you before I let you go.
Aledelpho: I'm ready! Shoot! 😄
Realm AI: The first one is: can you share three of your favorite AI generated art with us? Let's see your magic!
Aledelpho: You know, honestly in the last few months I've been so busy with my project and with my models that... I pretty much never had time to develop pictures for my own amusement. I think I'll have to share some of the pictures I've made for my models! (take into account that these pictures are coming from the noise - they are not img2img which means that these are the first rounds of something that can be improved a lot!)
I had to make these pictures in that way (just txt2img) in order to help people, re-creating them with their own softwares.
Realm AI: Final question is a two part question - do you have any parting advice for upcoming AI artists? And Where do you envision yourself in the next few years within the dynamic field of AI art?
Aledelpho: Well, my biggest advice is a technical one.
Start from your imagination: The best pictures are always made by starting from a very clear idea and then letting the AI inspire you through the whole process. The artist needs to feel like an Art director with a lot of robot artists at his/her service, drawing and shaping his/her ideas!
Ai Artists should never stop at the first good result: img2img is an important feature that can be used to enhance a picture over and over again! It's like being George Lucas in the pre-production of the Star Wars movies, looking at a lot of concepts made by the artists and choosing in what direction you want to move your artwork.
Don't post everything you make: the internet is full of Ai generated pictures and the more "mass" you create, the less your work will be seen as valuable by your clients.
Keep experimenting with different keywords, different subjects, different models... There is a lot to be learned by testing different techniques.
In regards to myself... I can see me working hard in order to improve this technology - I've been "bullied" a lot in the past for my love of "digital art" (you know, in a fine art academy it's quite easy to be seen as a "Mudblood" by the "Real artists" that paints with oil and sculpt real marble) but I never stopped believing in new techniques and new ways of making arts.
If Michelangelo Buonarroti (Caravaggio) or Leonardo DaVinci would still be alive, they would use these new technologies.
Realm AI: I could use some of this advice. Thank you
I wish you best of luck in your future endeavors. I hope to see more of your work on the server in the near future. Can you share your social media platform with us in case some of our readers want to follow you?
Aledelpho: (This is something I've done with Photoshop, way before the AI revolution)
If people are interested, they can contact me on these instagram accounts - I like to give tips and suggestions to future artists (I was a professor because I liked to teach, more than I cared about the paycheck), so, you're welcome to write me if you want:
Personal Instagram: https://www.instagram.com/aledelpho/