Industrial Design Takes on AI

RIT Industrial Design MFA / Course Design by Juan Noguera / October 3 2022

Introduction

Juan Carlos Noguera

Assistant Professor of Industrial Design
Rochester Institute of Technology
jcnfaa@rit.edu

In late summer of 2022, it was impossible to escape a myriad of social media posts, ranting or raving about Artificial Intelligence (A.I.) image generation. Projects like Midjourney and OpenAI’s Dall-E had turned a key we cannot turn back: converting text prompts into images of ever increasing quality, seemingly out of thin air.

And opinions were very split: anxiety about how it may negatively affect the creative field, and excitement about the possibility of using AI as a creative tool, both seemed prevalent. My colleagues and friends in the world of Industrial Design didn’t yet know what to think. I decided it was best to tackle the subject head-on and to start the conversation of “How may AI tools impact the every day workflows of industrial designers?”.

I worked with an amazing group of Industrial Design MFA students here at RIT, who were excited to jump straight into exploring this question. The prompt was simple: How would *you* insert/use AI as a part of your product design process? We are not looking to replace sketching, modelmaking, CAD, rendering, or any of the other core skills industrial designers use. But rather, how could you use it to improve/augment your process? Or perhaps, how can you use it to avert a creative block?

Shown below, is a simple in-class demo I showed my students. I prompted OpenAI’s Dall-E for images of “Cylinders Embracing Spheres” and threw in words like “Ceramic”, “Raku” and “Vessel”. What I got back was a set of fascinating organic forms (some better than others) that I admit I would not have landed on on my own (I gravitate towards more angular, simple form when sketching).

What came next (for me) was distilling those ideas into a product concept. I noticed these forms naturally generated a number of open cavities/levels, and thought that those characteristics  may be well suited for an object like an organizer/container of small items. I quickly took some of the best or most interesting bits and forms and in my sketchbook, created a few desk organizer ideas.

From here, I transfered these ideas to a simple CAD model in Autodesk Fusion360, and you will see a few simple renderings in the gallery below.

In my process, I used AI to convert a very abstract idea into form inspiration. But my students all had different ideas of where AI may fit in their own personal workflow.

In the classroom, amazing conversations were sparked, about authorship, ethics, and the evolution of our design discipline, some students seeing the parallels with the introduction of CAD or the internet. Each of the students were asked to chronicle and reflect on their experience, and you will find these fascinating writings in the full article (below).

Some of these designers used AI-generated images as a form-finding tool that accelerated or augmented their existing process, and others saw the AI take them in unexpected directions, with very rich results. The variety of products designed, ranging from musical instruments, furniture, to assistive technologies and consumer electronics, shows us the potential of using these tools in a variety of industries.

After this experience, I remain hopeful that AI-based tools will be a welcome addition to the tool kit of designers across many disciplines, and rather than replacing or dilluting the skills or desirability of the designer, will augment and accelerate our creative process.

Balance Chair for Kids

Shen Liu

My name is Shen Liu, I am from China and I am interested in design for vulnerable groups & sustainable development.

In this project, I explored the impact of AI on the product design process. And I believe that using AI is a perfect way to improve today's design thinking. According to Cam Sackett's “What the hell is a product designer?”, the existing design thinking includes five steps: Embrace with people, Define the problem, Think of a solution, Build a prototype, and Test your solution .I think AI can greatly inspire designers in the early stage of design thinking. I will explain the reasoning by introducing this project.

At the beginning of the design, I hope to design a furniture product with fruit characteristics. This furniture product needs to conform to product semantics. This is not easy because we just get a bit of information from the AI. Here is my initial prompt: 

Furniture design, fruit, minimalism, product design

This is the vocabulary that I gave to AI for the first time. I also added a design style. The result was disappointing. The AI only provided a large number of pictures similar to tables or tea tables. And fruit is deliberately placed on the product. This does not meet the design requirements of product semantics. Fortunately, a chair shaped like potato chip caught my attention. So I asked AI for help for the second time:

“A minimalist chair with the Product Semantics of fruit”

Compared with the first time, this search is a complete sentence. I also made clear this furniture is a chair and typing Product Semantics was one of the conditions. The result is also good. AI combines the chair and a cushion that looks like bread, very interesting. After that, I combined the results of two AI images to find potential users, and explored more possibilities of chairs in one in-depth sketch session.

Finally, the biggest difference between this project and my past work is that it seems to be "adding". This is an experience I've never had before. According to the user-centric design principles, Who and Why are the first conditions I defined. At this stage, we need to clarify many design expectations, such as use, product life cycle, etc. In this project, I think subjective choice and randomness account for a very large proportion. These sentences are the main drivers of the project, including:

"Ah, I like the surface that looks like a potato chip."

"This picture looks like a vaulting box. I think of it as a children's toy."

"Does this toy have more playability?"

"Can it get more people involved?"

“Can it grow up with children?”

It seems that these sentences help us obtain the design requirements step by step, but these requirements come from the designer's own experience, and AI just provided a picture.

In conclusion, I look forward to the development of AI and hope that it will become an intelligent tool or helper for designers. AI can provide additional creativity for industrial designers, especially when thinking is solidified or inspiration is exhausted. At the same time, I believe that it is unrealistic to rely entirely on AI for design because the complex principles of user-centered design are difficult or impossible to describe in language.

Presentation Clicker

Jacqueline Qiu

My name is Jacqueline Qiu. I am Interested in designing products that may help people release the pressure of this busy and stressful world.

I started my design process by trying to generate images through the AI program. As I needed to type in some prompt words to start the AI generation, I first thought about materials and textures which may help relieve stress. Then I tried to type in some words about context like “wearable device,” “speaker”, or “table.” I was also interested in fancy combinations to help me start my ideation process, so I typed in “fancy,” “interesting,” and even some culture-related words. As I was learning to generate more useful images through how other people type words and generate the images they want, it also surprised me as we can limit the generated AI photos’ results by typing in things like what proportion of the image we like, what degree the realistic of the image we want, and even what visual style we need. Overall, the whole process of AI image generation is like a combination of discussing and brainstorming with someone else, even though it is only you sitting in front of a screen.

After the generation of the images, I save some great features of things I love and bring in my creativity. I sorted them into different materials and contexts. I sketched all the possible ideas out of the generated images. Most of the drawings refer to the actual shape of the AI pictures.

At first, I imitated the shape of the items in those selected pictures. Then I started to ‘confer’ them ‘functions’ with ideas and details to make them useful products. It is a fascinating process as you talk to shapes with your pens and let them jump out of the paper as a well-structured product.

The very first of them I was ‘talking’ with is shaped as a decorated stone sculpture being carved and shown as textured wood inside. According to the context of people living in the stressful world, it leads me to think about aroma, humidifiers, and even ends up with the presenter clicker. Cause wood always has some interesting fragrance in my personal aspect; it leads me to think about aroma, which may help provide the fragrance for people to feel comfortable and relaxed. Considering the stone as a mountain, the humidifier seems to have strong materials combined with softy ‘clouds’. It gives a beautiful and moving continuous steam for people to stare at and feel relaxed cause it is nature related. Then I found out that the item could be smaller. Whether it is just as small as our hands? Whether it is a reliable, smooth stone that can be hold in our hands when we feel nervous while doing presentations? Whether if it can provide some outputs to people to interact with them? Whether it is a very functional piece as a clicker?

After my ideation, I selected the fanciest concept according to my perspective from all of the ideas in the sketches and started my prototyping. It will help to ensure the curves and the structures are as good as I thought.

Then, I started with the CAD model and printed it out to test whether it worked or not.

As I started this idea from the AI photo, which related to the materials and textures, I wanted to make the final product more connected to what it used to be. So after the shape and curves were modified well, I tried the textures and materials through CAD model and renderings to make sure the color and the materials combination is fit and worked together perfectly.

The process of generation and ideation gives me an image of a future with AI in every corner of this world. What if it produces a sheet of music that may heal people’s mental illnesses? Can it make robots have the proper conduct to be a companion to a child? Can it have new ideas for an advertising plan for the sales manager? Can it be not only a 2D artist but also a 3D art sculptor? I think there will be tremendous and exciting innovations with AI in the future to meet every person's needs.

Still, could it interrupt human development progress passively also? As AI is a thing that can relate all the information it has, which is definitely more than what one person has, to generate something. So, for ordinary people, will the world become way more exciting, making people find everything so intelligent and stop progressing in their life? For artists, will AI be a great tool to support them, or cause it is hard to go out of ‘their own style’ as AI do? Will, they lost their own artist’s heart and repeat what AI generated only?


Violin

Jayden Zhou

My name is Jayden Zhou. I’m from Zhejiang, China, and I’m interested in designing consumer electronics.

I’ve never designed transportation or musical instruments before. I wanted to try a new way of designing with the help of artificial intelligence(AI).

The AI tool I use is Midjourney, which is more like a quick search tool for me, which quickly performs random combinations based on the keywords I set. I picked the direction I desired within the generated results to go deeper. Research is essential to design. I do a lot of research before designing, like the history of the object I am working on and what kind of work other designers have done. This search is a learning process for designers, and AI can help them learn more quickly. In addition, the resulting AI-generated images may not be realistic, but parts of the results, such as the color scheme, are very informative.

My prompt was “futuristic sci-fi stringed instruments”. I expected the AI's perspective to be unique and full of randomness. I expected it to generate instruments that were different from those on the market. Four images are initially generated, the first is a light strip that spans across the whole image, and the other three are guitars in various forms. I initially used the guitar as a variant to continue generating images, but it was so much like a guitar, those results were not immediately striking. Eventually, I found a tiny incomplete instrument in the corner of the first blurred image. This instrument was a bit unreasonable in structure, but its shape was never seen before. One of the advantages of the blurred image is that it gives the designer enough space to imagine and rationalize the unreasonable structure. I sketched the stringed instrument according to my imagination, and its layout and colors are more like an acoustic violin. During the ideation process, I kept the characteristic contour lines as the outline of the electric violin.

In the future, It may become common for people to use AI as a tool for their daily designs. When struggling with inspiration, I could use AI to help me overcome obstacles. For now, the chances of success are low, but it’s worth a shot.

AI will be expanded to all industries as a tool. For the design industry, AI will be trained to specialize and become a design partner and tool that designers will use to meet the ever-changing demands of their work.

More and more intelligent AI works stun our eyes. Its results made me question the need for the existence of the designer profession. I remain optimistic after consideration. The emergence of designers is inseparable from the history of human development. When our ancestors first picked up stones in the process of hammering, grinding, and making tools, designers in the primitive sense appeared. It is a long evolutionary process from the first "tool maker" to the modern designer. As technology progresses, the design profession will evolve adaptively rather than be replaced.

Works of science fiction have often discussed the difficulty of AI having human emotions. AI results do not come with emotional judgments, which is an essential part of the design. While AI can combine existing things, it cannot create things. AI may replace designers’ specific rule-based skills; However, creative designers will not. The designer of the future will become the fundamental creator. To use an analogy, they will go from being instrumentalists to music composers. For designers, AI has only changed the process of creation, while the essence of creation has not changed.

Thinking at a deeper level, when things created by AI surround us, do we still have the ability to judge things, good or bad, beautiful or ugly? We should always keep human judgment superior to AI, which means we may need designers more than ever.

Stingray Saddle Seat

Ariella Knight

I am an optimistic convergent designer and maker. With a broad arts/tech background my current work focuses on access and sustainability and explores how social and disruptive design can facilitate new and more equitable systems, products, and services.

I did not set out to design a chair. I started by thinking about solar panels on farmland. When Midjourney spit out a steampunk, veiny system of jellyfish in the sky after I entered the prompt “jellyfish/solar panel/farmland/4k,” my interests wandered and I thought more about what a mashup of contrasting prompts could generate. This can happen a lot in design, a squiggly, spiraly journey that shuffles us from where we started to where we end. Lorraine Justice calls this part of the process the “fuzzy front end.” (Check out the included images to visualize what this journey looked like).

The simple, fast, AI driven accessible exploration of divergent prompts was what became immediately captivating to me in this project. As a mother in grad school I like to include my daughters when I can, and so we came up with some prompts together, “soccer ball, fish, diaper,” “ice cream, laundry, sunshine, wet,” and "stingray, bed, velvet.” I liked what generated as a result of that last one. 

Some of the produced images made me think of current products, for example the Diamond Chair by Harry Bertoia. The Diamond Chair made me think about uncomfortable designer chairs (yeah, I said it). The notion of comfort led me to posture which made me think of saddles and ergonomic chairs. When I presented a concept sketch in class Juan Noguera supported my direction and referenced the Krabat Wheelchair. So I continued on, now designing a Stingray Saddle Chair– one that echoed some of the midcentury forms that I love, but could be comfortable, encourage good body alignment and even build strength over time (really not agricultural solar panels).

I share this backstory because I can. Because I have references for exactly how it went, what threads were followed and each thought explored. That knowledge– the documentation of this process and the understanding of acceptance or rejection of thoughtlines feels in stark contrast to how the AI image generating programs work. The AI process is opaque, you type in a string of words and presto some images appear, but no opportunity to explore tributaries along the way. Where does ownership lie? 

Yes, these Midjourney outputs are generated from my inputs, and yes the images created moments of thought and inspiration, and yes, I drew from them. But also there is a severing associated with the images that feels disruptive to me. Where are the works coming from? How did Midjourney get here? We look deeply at process and thinking in design, I feel myself really longing for insight into AI thinking. There are hints– artifacts of text or barcodes blurred out and dissolved into the landscape, but it’s not enough. 

How could this be remedied? A way to trace origins would be a great place to start. Maybe even access to a library of source material.

Especially at a time when we are thinking more about appropriation and crediting Midjourney’s generations in the void feels like a move in the wrong direction. Some might argue that it is an equal playing field– new to everyone and with widespread availability, but I question if in the end it’s a sneakier form of erasure.

Accessible Deodorant

Jos Mayo

My name is Jos Mayo, I’m from Rochester, NY and as an industrial designer I am interested in medical devices, construction tools and machine design.

At the start of this exploration into AI, I was a little unsure that it would be applicable for design. I hadn’t heard of AI image generators like Midjourney or Dall-e before. I’m not sure where I stand with AI image generators as a general topic but it is worth wandering into the space to try it for yourself; you will be surprised! For functional products or objects that could perform real world functions, it’s not there yet for me. It is more a source for inspiration because of the lack of clarity. My prompt for this exploration was “A dishwashing tool for someone without hands”.

My grandfather was involved in a train accident that cut off his right arm above the elbow. He had a simple rubber arm as a result, commonly called a passive prosthetic. This got me thinking what could have been available to him had this happened today? How would his life’s experience be different? My prompt was the result of that thinking and my first impression of my results was surprise because it generated objects with handles! I could have made a new prompt with “no handles” included but instead I requested multiple revisions based on the initial blue form to see what it would do. The results slowly created a more usable form that could be translated into a prototype or actionable form for development. It’s intriguing to see how the design was slowly changed over time and went from vague to more specific and even developed reminiscent kitchen tools to accompany them.

Some simple sketching and research into the real world of amputees revealed a glaring lack of options for hand amputees, especially bathroom aids. Many of the best institutions that outfit amputees with prosthetics advocate using microphone stands as the best commonly available option for mounting hygiene items for daily use. These solutions still often require amputees to rig their own setups while taking into consideration that they are potentially new to the world of living without a limb(s)! After seeing pictures of a deodorant stick, electric shaver, toothbrushes, etc, all strapped to gooseneck microphone stands in a bathroom, I couldn’t ignore the glaring lack of attention that these individuals and their unique needs had acquired.

My final design ended up being a simple clamp that utilizes a weighted microphone stand and a universal microphone attachment. It has a compliant mechanism built into its shape that acts as a clamp when pressed into the microphone attachment. The deodorant stick can be securely held and is easy to mount without much need of assistance. The intention of this design was to be able to insert deodorant and mount it to the stand without needing finger dexterity (no velcro or spring loaded clips). 3D printing emerged as an effective way to provide this solution to the community that needs it, and optimizing its shape for 3d printing added to its aesthetic form and function.

Ultimately, I didn’t use my original AI images or forms but they did provide the push to investigate a problem and find a solution that I wouldn’t have otherwise done. This pushed me to think about how design has been done in the past and how it may change for the future. So, where would I use AI in my workflow of the future? I would try it again but it does feel a little foreign to me as a permanent addition to my current workflow since it isn’t my own creativity that is being expressed through the AI. With some more refinement and control over the process that generates images, I could see this finding a permanent place in most creative processes.


Terra Music Player

Anqi Zhu

My name is Anqi Zhu. I’m interested in conceptual design, toy design and retro-inspired design

The market of CD players and Vinyl record players has been saturated for a long time. However,  modern internet culture and social media keeps reviving this sector. Fans of particular musicians and celebrities continue to purchase music in vinyl and CD formats as they are more precious and collectable, preserving these obsolete formats. 

My proposed problem was to design a CD player for the retro-futuristic music market, something that is maybe future proof and that can help protect the medium from its next wave of  abandonment.

For my AI-assisted idea generation, I prompted the AI for “Futuristic blender render of CD player, 8k, emotional” and the images produced were amazing. These music players were a fusion between a disco dance hall and a modern product. I didn’t see actual CDs in the images, only glass boxes but I noticed the appearance of vinyl record textures in the top portion, which led me to the thought of combining both a CD player and a vinyl record player in the same form factor. 

The beautiful lighting of the AI generated images also gave me the idea to combine light projection into the product and design thinking. 

 Throughout this process, I asked myself, how does it fit within my industrial design process? Before this close encounter with AI technology, I had only read about it on the news. Only a couple of years ago, the state of the art for AI was clumsy face-swapping and funny effects on social media selfies. However, these AI bots have gone from clumsy and ridiculous, to sophisticated and scary, in only a short period of time. 

Using an AI system like Midjourney has me a little worried. The images are fantastic, they can have a touch of genius, and are generally very creative and unusual. Most people think AI is very distant from them, but these systems are becoming usable tools. As long as we are transparent in their use and existence, especially in our design process, things will be well.


Mycelium Shoes

Tayo Oke

Hello, my name is Tayo Oke, I am from Long Beach California. I am interested in designing everything.

 I prompted the AI Image generator Midjourney to show me how “mycelium slip on shoes with brown rope shoelaces dipped in gold” would render. Midjourney gave me what I would call the “perfect” render of a slip on shoe made out of mycelium. This is where I became stuck in my design process, because the AI answered my question perfectly. 

First, I tried to be minimalist in my design process by dissecting the noise and texture created by Midjourney so I could possibly make a realistic prototype moving forward. But the more I sketched, the more I saw myself competing with the AI.

 I went down a rabbit hole on how mycelium grew and navigated through the earth. Biomimicry strengthened my process by drawing inspiration from parts of mushrooms like the cap and the grills.  Studying characteristics of real mycelium while looking through the AI renders helped influence my creativity and thus I was able to blend nature and machine to create my rough concepts.

Midjourney made me curious about the possibilities of mycelium footwear. In research, I discovered Stan Smith’s Mylo adidas trainers; a shoe leather made from mycelium. Discovering more fabrics and garments made of out mycelium made me confident that this process was practical and could be made as a working prototype.

Overall, this whole experiment was very eye opening to how not to use AI in a design process. I think Midjourney (more than Open AI’s DALL-E) can be an excellent tool to use in the design process because it develops pictures in an artistic format, whereas Dall-e feels like I was manipulating a bad photoshop collage. The way I used this program might solidify the skeptics who believe that Midjourney and “AI are coming to take over our jobs”. However, I would argue that this proves how new and imperfect these tools can be used in creative spaces. In my process I immediately went back to draw inspiration from nature to strengthen my sketching and design process because that was the only way I saw myself making progress in this experiment. 

I believe in industry, AI can be used to strengthen our process if used correctly. I could see Midjourney used in the beginning of a designer's process by making a prompt, generating different variations and ideating as I did for this project. I can also see AI used in the middle after a designer has sketched their ideation and then submitting said sketches to the AI and seeing the different variations of “sketches” it produces.  

Designers should experiment with AI and see where this tool fits in their own creative process. I also believe that creatives should place creative boundaries on how often they us Midjourney AI so they [nor I] are relying on AI to generate most of their ideas. Going back to my example, if I needed to design a shoe with natural organic features and then hand it to my boss; I could just hand in the refined render that Midjourney created and call it a day. Because I also believe that if used “poorly” Midjourney could be used as a commodity to take the easy route.  

I will try to use AI just so I can personally see where it fits in my life. I've always had an interest in how people interact with AI and experiments. It made me curious and caused me to want to learn more about a subject that I wasn’t familiar with, so overall it did its job of pushing me to want to know more and to unlock a different area of creativity.

Health Monitor

Zaheer Shujayee

I’m Zaheer, I am from Pittsburg California, and I want to design products that promote well-being and mental health.

Anxiety disorders are being more and more widely diagnosed, with centering exercises being used to manage anxiety. I’m Zaheer and I want to design products that promote well-being and mental health. This has historically varied from products that make people feel seen and represented, to products that deal with neurodiversity. With my anxiety management wearable device, I aimed to create a product that will detect a heart rate congruent with a panic attack, and prompt the user into a guided breathing and centering exercise. This guided exercise uses a circular light to prompt whether to inhale, hold, or exhale.

Part of the design prompt was to use an artificial intelligence image generator (Midjourney) to create product images using prompts. I wanted to see what devices specifically catered to anxiety disorders looked like according to A.I., and received some interesting results. A hurdle I encountered was that this space is largely dominated by wrist watch devices, which was a form factor I was not interested in. A large part of the reason for my disinterest in this form factor was that many users would already be wearing a watch or watch-like device in their daily lives. I expected the A.I. to give me an amalgamation of what was already available while also weaving the feeling of “anxiety detection” into the device, helping me envision what an appropriate looking device for that task would look like. I feel like the A.I. succeeded in this task, giving me form factors and aesthetic cues that felt calming. Of course I needed to add other adjectives into the image generator such as “sleek” and “modern”. After refining my prompts I received many interesting looking products that worked to guide my design thinking as well as giving me a bank of influences to draw from.

Artificial intelligence is very useful for creating a supply of inspirations to draw from, but it is in no way a tool that will replace designers. For myself, it was best used as a tool for getting the ideation process started; it was no replacement for ideating, sketching, and iterating. My expectation of the A.I. assisted design process was that the A.I. would simply amalgamate many related products from the web of reference it had to draw from, and I feel as though that was what happened, and with some refinement of the prompts I used, I feel as though I was able to make my design process more efficient by introducing a new tool into my process. As with many other new tools for designers, it did not replace the role of the designer, it simply changed the design process in a way that has the potential to make the designer more efficient.

I see A.I. being used as a means to kickstart the ideation process in design, it creates really interesting first round concepts that spark conversation. I really enjoyed making this a part of my process, and if I feel like the design prompt would suit A.I. assistance, I would definitely see myself using artificial intelligence.

Camping Stove

Rob Deane

My name is Rob Deane and I come from a sculpture and manufacturing background, I am a welder by trade.

My primary focus design-wise is furniture and home goods, but for this project I decided to branch out a bit into camping gear.  I was particularly excited to get started on this project.  I had been playing with DALL•E Mini (Craiyon now) all summer and the idea of using more robust AI engines for design purposes was right up my alley.  I jumped in with very product-focused prompts, referencing design motifs, eras, or movements, materials, and the desired product. 

The prompt that gave me the most intriguing results was “cleverly designed aluminum mid-century modern camping stove.”  I took that set of results and ran with them.  I ran variations of the original results, variations of the prompt, on both DALL•E and Midjourney.  I bounced images back and forth between the two systems, and determined what parts of the process each were more suitable for.  After generating several dozen images of what a camp stove might look like, I developed three widely diverging designs for wood or charcoal fueled outdoor stoves.

This was a fascinating process to explore.  It seems that everyone has their own process using the AI tools.  I fell into a rhythm fairly quickly.  The focused prompts gave me material to immerse myself in, focusing on highlighting and iterating on features and forms that I found appealing or useful.  I found that Midjourney was more suitable for generating the kind of initial ideas that I was looking for; DALL•E was more suitable for iterating on individual images that I had generated using Midjourney, as well as for creating images that were more technically cohesive.  DALL•E seems to be much more proficient at producing detailed, life-like images of products, but Midjourney was significantly more “creative” with the results it would present.  My ideal workflow, I found, involved using them in tandem.  

After being satisfied with the material I had gathered, I started to sketch, pulling forms and details from various images.  The abstraction of the images was key to this process.  Many features in the images had no obvious function, leaving plenty of room for my own interpretation.  These bits were sort of creative sparks for me, seeds to let ideas grow from.  

Going forward, I think that AI is going to be an important tool for designers.  It provides an ability to iterate very rapidly and broadly without falling down rabbit holes like many of us tend to.  For my process it helped me to branch out into forms and patterns that I may not have otherwise come up with through my normal ideation process.  Everyone is going to use these tools differently, and it’s up to each individual to figure out how AI fits into their personal design process.  Many people are concerned about AI’s place in the design world, fears that are mirrored in many other industries.  Perhaps someday these fears will have some ground to stand on, but for now DALL•E and Midjourney, among others, sit in a sweet spot for designers.  They require carefully tailored input to obtain the desired results, making them a fascinating interface between human and machine.  As these programs develop, I can only hope that a measure of this interaction is preserved.