Many people’s first introduction to A.I. (artificial intelligence) art was “deep dreaming”; silly, trippy, folding, color filled, images with dog and frog faces everywhere. It was almost like watching a computer’s thoughts as it tried to understand and recognize objects and patterns in imagery. It wasn’t creating original works, just displaying the non-existent patterns it had seen in a reference work you had fed it. Hence why people named it deep dreaming. It genuinely had no practical application and little input from the user as to the final generated output.

A.I. Art — Then vs Now

When people hear A.I. art is winning art shows, a naturally understandable abhorrent reaction occurs when “deep dreams” are their baseline understanding of A.I. art. It would not sound fair to the reigning masters of their crafts if people are just pressing a button and winning art shows without any input on the image generated; but that would be a quick and false assumption.

A.I. art tools have reached a maturity point. It is no longer just for trippy incoherent scenes of wild colors. These new A.I.’s have been trained on millions of works of art spanning all fields of human creativity including: illustration, logo design, typography, photography, 3D renderings, even fashion design. MidJourney’s interface is designed for users to input intricately crafted prompts that effect every aspect of the output from subjects, composition, lighting, medium, color pallet, weighting, and style cues. Users of these tools now have full creative control over the input prompts and can ask for anything they can imagine in whatever style they want. The process of learning how to craft these intricate prompts, using ones’ artistic taste to curate and iterate results, feels like learning any other art tool. MidJourney is not the artist, it’s a new tool to be learned and used by artists.

Still some might argue you are not the artist if you did not spend enough time on the work, or didn't produce the work directly by your own hands. Marcel Duchamp’s “Urinal” proved that wrong over 100 years ago. A “ready-made” urinal bought off the shelf and shown in a gallery setting. Art is about vision, and ideas placed into a context for expressive communication; not about the duration of work that went into its creation. In fact, throughout history many artists have had their works produced for them by others under their guidance.

Having your work produced for you isn't really a new system for the art world. Artists since the renaissance have had teams working under them producing their works. The Artists direct the creation of their ideas through their understudies, giving feedback along the way, before placing their signature on the final work. In modern creative fields, Art Directors often dictate their visions to designers and interns who produce the work through iterations in a feedback loop. Let me give you some examples and you will see why these A.I. tools turn any competent designer (in any field of design) into an Art Director with a team of A.I. designers and interns working underneath them iterating their ideas.

I prompted MidJourney to create the metallic armored ancient aliens I’ve been dreaming about.

Whether early on the client wants multiple sketched options or more flushed out 3D renders, you can use MidJourney to iterate out concepts quickly. Less time spent on initial concepts/options helps us quickly find our or the clients vision; allowing opportunity for more time to be spent on the final deliverable or giving us extra time to take on new projects we wouldn't have been able to.

I prompted MidJourney to create the metallic Armored Ancient Aliens I’ve been dreaming about. I got 13 really good characters out of it that I could choose to flush out or model further. I could also have asked for basic sketches but with the right prompts I can jump straight to mock ups that look 90% flushed out. Any of these could be shown to clients to choose from before moving in a final direction. Without MidJourney these characters would have taken 5–10 hours a piece to model, here I created a lot of 13 in under 2 hours total.

I prompted MidJourney to generate beauty looks based on the Chinese New Year.

As in other fields of design A.I. could be used for concept exploration. I generated beauty looks based on the Chinese New Year. I expected these would be great starting points; reference material for a make-up artist on an actual shoot. However, the results exceeded my expectations. These images have not been retouched, and with a little editing would be ready to go to print and skip the shoot all together. So why not push it a little further?

A photoshoot with Victorias Secret model Taylor Hill generated entirely with prompts using A.I.

When referencing celebrities, actors, musicians, models, or TV personalities, the A.I. has enough reference material to convincingly create their likeness. I asked A.I. to give me photos of Taylor Hill, with lighting, outfit, lens choice, and make up of my preference and it delivered. This is just the beginning of this technology. In the very near future people will be able to digitally capture or scan a persons’ likeness and phone-in any photography or acting work they want from that person using only a computer. Soon models and actors may not ever have to show up to a production ever again. Large photo or video productions may not be required for every application anymore.

Food Photography generated entirely with prompts via MidJounrey

One more thing…. Food Photography. In this field of photography many tricks are already used and accepted. Nothing is as it seems. Anything is done to make an appealing image representing the delectable for sale. Fake dew drop? Check. Coloring fruit with furniture crayons? Check. Replacing edible liquids with car fluids? Check.

It turns out enough good food photography has already been achieved and referenced that the A.I. can pretty much create the perfect image of whatever highly specific dish you envision, at whatever angle you want, on whatever surface you want it presented. If my agency had this tool in 2012 we may never have produced the photoshoot and shot Dole Foods 184 products; we honestly could have used A.I. to create the imagery for us with a much cheaper cost.

Lace bra concepts, whole outfits, and sneaker designs - MidJourney can do it all.

The final product in the world of fashion is usually a wearable garment. A.I. art will therefore usually be a good process tool here but not as frequently a tool for final output. In the case of digital garments for avatars, or 3D printed garments, A.I. generated outputs could evolve to produce the final product. In its current state, it can easily help you generate ideas for anything from lace tops to modern sneaker designs.

MidJourney tackles modern logotypes, icons, or classic hand drawn logos. Limited only by your imagination and articulation.

MidJourney tackles modern logotypes, icons, or classic hand drawn logos. Logo design with MidJourney is limited only by your imagination and articulation. Again, someone with a background in Art Direction or possessing an art education, is going to have an easier time getting started with this tool. The ability to use the correct terminology and articulate your desired results is paramount. Art Directors have already spent years refining their ability to articulate their ideas working closely with designers to get to the final product they had envisioned. The ability to create now lies more in vision, and communication, then physical manual tools of creation.

I literally cannnot wait to see all the new emerging typographic design

It’s become very popular in recent years to use custom typography in advertising and design. Often letters are individually modeled and rendered, or recreated by hand out of food, wood, or stone. MidJourney will be incredibly powerful here. Allowing people on tighter budgets and timeframes to use much more advanced typography in their designs. I have a feeling we will see many new type-faces created with help of these A.I. art tools.

I have lived to see new technologies emerge in art. I’ve seen the cycle before. The backlash, acceptance, and finally incorporation of the new tool into its niche use. I have witnessed this cycle with Photoshop, DSLR video, and cell phone cameras.

With Photoshop’s introduction people argued digital art was not real art. Now most art incorporates a digital workflow.

When DSLR’s first gained video features, professional video work went from requiring large production teams with deep pockets, to run and gun crews on a budget. There was backlash it would cheapen video work, ruin jobs for those in the industry. Instead it lowered the barrier of entry and allowed many new artists into the field. It increased competition and increased the over-all quality of work on any budget.

When cell phones gained competent cameras, in much the same way, the field of photography was changed forever. Lower cost of entry. More artists. People can now use their cell phones for graduations, engagement, and holiday cards. Cell phone photo and video are quality enough to be incorporated into a professional workflow. A competent professional with a modern cell phone can often deliver better results than an amateur with a DSLR. It gave people with talent and artistic vision a new cheaper tool to allow them to compete in a workplace that was filled with photographers that had spent tens of thousands on equipment. It leveled the playing field and made it about skill, not capital.

Now we have, yet again, a new shift in the industry. A new tool to learn to incorporate into our workflows.

With proper use of these A.I. tools any designer becomes an Art Director with a team of A.I. designers and interns working underneath them iterating their ideas.

Written by Creative Director - Harry Tiffany V www.instagram.com/harrytiffanyv/www.chasingtrees.agency

A weekly resourceful newsletter that helps designers stay in the know, be productive, and think more critically about their work. Take a look.