Written by – Wang Yabin Creator – Blindart
In the past two years, AI has been the hottest topic in our industry. But how can AI specifically help us? What benefits are there to using AI technology in our daily work? What advantages does working with AI have compared to traditional methods? I believe many of my peers are still unclear about these questions. AI technology has developed to possess various powerful functions, yet for ordinary technical workers, it often remains a vague concept. Many of AI’s powerful features have not been widely utilized. Although the number of practitioners truly benefiting from AI technology is still relatively small, especially in our industry, everyone has gained some basic understanding and recognition of its powerful capabilities over the past two years. This was particularly true after the new version of Photoshop launched its AI features last year, providing everyone with an opportunity to personally experience this technology. Some friends have become completely reliant on AI assistance in their daily work; for instance, when I’m retouching photos of a lawn, I find it very convenient to use the AI features in Photoshop for intelligent repairs. The powerful capabilities of AI have been witnessed by countless peers. AI technology is likely to gradually be integrated into our daily work as people become more familiar with it, just like curve and liquify tools that we frequently use.
Photography: Zhang Daxi (Chunxi)
Using AI features in Photoshop to generate images requires English input for keywords, which forces many people to use translation software alongside, significantly limiting the use of AI and stifling creativity. Even simple translations of a few Chinese keywords into English can easily result in errors or inaccuracies, let alone lengthy Chinese texts being translated into English, which can deviate significantly from the original meaning. Fully mastering AI functionality presents a considerable difficulty barrier, ultimately limiting our own ideas and requiring simpler demands from AI, which greatly undermines AI’s powerful capabilities. If using AI significantly increases our work time, and we eventually find that using it is less time-efficient than doing it ourselves, this outcome is unacceptable for many and contradicts the original purpose of AI. Additionally, we face the issue of high costs associated with foreign software. The AI features in Photoshop are not free, which is the biggest reason why AI technology has not been widely adopted. However, despite the numerous obstacles and issues summarized above, as long as AI is sufficiently excellent and powerful enough to genuinely assist us, these obstacles and problems are not insurmountable. The popular domestic AI tool “Pixel Cake” has also utilized AI algorithms for its work, and the highly popular “Monet Garden” effect on Douyin in May also employed AI technology to transform images into oil painting effects. Currently, domestic AI technology primarily relies on automated computer generation rather than being based on textual descriptions, which results in the effects produced by such AI being less professional and lacking in detail, thus not yet suitable for professional artists. Nevertheless, AI technology is increasingly accessible to ordinary people. What we need to do now is to promote the power and benefits of AI.
This time, I have specially invited a highly popular AI author on the Xiaohongshu platform and her team (Blindart) to collaborate with me. The team consists of two members, Ivy and Langlang, who operate four different accounts on Xiaohongshu, each with varying content, including AI works in portrait photography as well as works related to animals and clothing products. The works presented here come from the account Dreamart, focusing primarily on Chinese models, while other accounts feature works based on foreign models. Daily updates of this type of work are largely generated by AI and can be completed quickly, allowing for updates twice a day. However, for business collaboration advertisements, real-life shooting or product photography methods are employed, and then AI technology is used to complete the works, which naturally takes longer to produce.
The team is not large, and one person takes on many tasks, with even just one person completing AI creations. Currently, most AI creators are still primarily hobbyists with secondary income from their work. However, the Blindart team has already turned a profit, transforming AI into a legitimate job, meaning that using AI for creation is no longer just a hobby.
In a white studio setting, AI technology was used to generate a dark sky, black beach, and withered flowers, creating a doomsday atmosphere.
In a white studio setting, AI generated a boat full of roses and a sunset sky effect, creating an elegant and romantic atmosphere. AI can realize environments from fantasies and dreams, which is its most enchanting aspect.
AI works in the fields of beauty, clothing, and luxury goods.
Q: When did you first come into contact with AI? What initially drew you to this technology? And why did you decide to form a professional AI team to specialize in this field?
Blindart A:We started learning about AI technology in the spring of 2023 when it began to gain popularity in Europe and America. Initially, we used AI simply for our own amusement, creating some works during our spare time at work. As we became more familiar with AI software and as AI technology itself became more powerful and mature, the quality of our works improved, leading us to create an account on Xiaohongshu to share our works, thus starting our routine of regularly updating our creations.
AI National Style Masterpiece
Using AI, various types of national style works can be produced, such as the two works above, one featuring a cool and handsome female image, and the other exuding feminine elegance.
Since none of us are professional art students, we named our account Blindart, meaning “blind art.” This expresses our attitude that, although we are not formally trained, we can still create artistic works. Just like a blind person who cannot see the road, as long as they have a cane, they can walk! AI is our cane; even if we haven’t received much artistic training, or if I wasn’t previously in this industry, it does not prevent our team from using AI’s help to realize our imagination and create works that belong to us. With this mindset, we continuously update and share various themed works, slowly beginning to receive more and more likes and comments, and while receiving positive feedback, some brands have also approached us for collaboration. This has made us feel the immense potential of the AI industry.
AI Artwork Appreciation
These are also national style works, but this set is more flamboyant and sexy. The image of the black phoenix has a strong visual impact, and AI brings many imaginative ideas.
Q: Although you both lacked experience in design or retouching work, it did not hinder your creation of works. Even I, who have worked in retouching for over a decade, was surprised when I first saw your works. They do not show any signs of amateurism, and I even heard that some of these works were entirely generated by AI, yet the realism of the images has reached a level that is indistinguishable from real.
It is precisely because of this that I broke my previous impression that AI works are certainly fake and can easily be distinguished from those created by humans. I have come to recognize the current stage of AI and deeply feel its powerful capabilities, and I am amazed by its rapid progress. Additionally, one thing that surprised me is that using AI software does not require us to be particularly skilled in professional design software, and the effects of the generated works can compare favorably with those produced by designers with many years of experience, even potentially surpassing them. So how do you usually use it for creation? Or, to put it simply, what specific roles can AI technology play for you? In the entire process of creating your works, what key role does AI play?
Blindart A:Rather than saying AI can do things for us, it is more about what we want it to achieve. AI is actually an extension of our inspiration, turning what we want into concrete images. It can generate not only things we have seen in reality but also things that do not exist in reality, only in dreams, which is the most powerful aspect of AI. If its ability was only to generate a flower or a flame, we could accomplish these tasks ourselves, and its existence would lose much significance. Indeed, it can help us save a lot of work time, but this is merely the most basic and easiest capability it can satisfy among our many requests. Because AI does not have many bottlenecks and frameworks like humans, the more ideas the user has, the more it can generate. As long as the generated images are morally permissible, AI can produce and use them! So we often joke that the biggest limitation of AI is the person using it. Because of human limitations, there are still many things AI can do that we cannot showcase to everyone, and humans have become the biggest factor limiting AI’s maximum functionality.
AI Artwork Appreciation
AI can generate everything for us, even creatures we have never seen in reality. For example, neon-colored butterflies create a magical, surreal effect through AI generation.
In our work, there are also two different situations when using AI. For instance, if the creation is for a business collaboration, we must create based on the brand, referring to the brand’s past tastes and the requirements and product concepts they propose. In this case, AI plays a role in helping the product find a suitable artistic environment. For example, if the product is a perfume with a pristine white bottle, we can generate a flower in the background using AI, which is what we mean by an artistic environment. In the process of creating such works, the product is the main character, and everything revolves around this theme. However, if the creation is a non-profit work, like the daily updates we maintain on Xiaohongshu, we can unleash our creativity without restrictions. In this case, imaginative ideas and inspiration take center stage, and the fewer limitations there are on using AI, the more it can assist us, greatly enhancing its role. In summary, it is not about fearing AI’s inability to meet our demands, but rather about fearing that we cannot come up with more outstanding ideas.
AI can also create effects that can be photographed in reality but are relatively dangerous. For example, this series of works was inspired by Pink Floyd’s album cover “Wish You Were Here.” The behind-the-scenes footage of their album cover can be found online, and those interested can take a look. Some of these effects are dangerous and must be operated by professionals. In the past, without AI technology, we could only rely on fire to achieve this, but now we can entrust AI to accomplish these beautiful yet dangerous images.
Q:Since the birth of AI technology, there has been constant praise and skepticism. Some believe AI is a good thing that will inevitably lead to innovation and breakthroughs in our industry as technology matures. Others have a strong prejudice against it, believing that the widespread use of AI technology will become a tool for laziness, and allowing AI technology to be used will inevitably lower the quality and taste of works in the entire industry, resulting in future artistic works lacking human warmth, only cold images generated by machines. This group believes that the impact of AI on the industry has more drawbacks than benefits. I think it is normal for differing opinions to arise, and I view it as a good thing because through this process, truly good things will remain through continuous improvement, while bad things will be forgotten due to their inability to address people’s doubts. Although the current situation sees half of people actively embracing AI technology while the other half, comprising traditional retouchers and designers, are relatively resistant, it does not affect the increasing number of people benefiting from AI. You two are among those beneficiaries, as AI technology has enabled even those who were not professional designers or retouchers to create hundreds of artistic works. You also have a deeper understanding of the powerful functions of AI technology. As beneficiaries of AI technology, can you share what changes it has brought to you? What tangible help has it provided? How do you view those who are currently resistant to AI technology? How should we correctly perceive the role of AI in assisting our work?
The effects generated by AI are entirely determined by the creator. Often, to achieve a stunning final effect, an AI render is generated in advance, and then a real person is photographed based on that render. In this process, AI is just one of the tools; the initial concept and preparation of props and models have not been reduced due to AI. So how can using AI be considered lazy? It should be a more mentally demanding task, as it allows for more refined creation of works and greatly increases the accuracy of the final product.
Blindart A:I entered the design field through AI. When I have better ideas, I can express them myself through AI without relying on others. Since I started using AI, I have found myself more willing to imagine because AI has lowered many barriers to design, but I don’t see this as a bad thing. Although AI is not as complex to operate as professional design software, it still has its thresholds. The only threshold comes from your creativity and aesthetics. I believe that without my previous love for art and accumulation of experiences, I could not have created so many works through AI. Every piece generated is the result of keywords I input, as well as the outcome of my selection from the various generated images. Moreover, if we think carefully, many times we are attracted to a painting, a movie, or a song not necessarily because of its highly sophisticated techniques. Of course, having advanced techniques is great, but it is not the key factor that attracts us; the most crucial aspect is whether the creator has put their heart into their work during the creative process. I am not saying that technique is unimportant; I am just emphasizing that sincerity is more precious. Many people are resistant to some AI works because behind the shoddy production lies not only the immaturity of the technology but also the creator’s lack of care. Therefore, even if AI technology is powerful and has incredible functions, if the creator does not inject genuine feelings into its usage, it will only yield poor reviews.
Designers and AI need to achieve a harmonious relationship. Based on thousands of interactions between humans and machines, we have truly understood AI’s logic, allowing AI to better understand our needs, and it is through these thousands of interactions that we have formed our current style.
In conclusion, humans will always be the masters of AI. What your work ultimately looks like depends on your creativity, your aesthetics, and the patient process of conversing and revising with AI.
Q: In my spare time, I also play around with AI, but the results generated by AI are often unpredictable, frequently producing strange or unsatisfactory images, wasting time and often causing headaches. However, your team updates works twice a day; how do you achieve such high stability? What specific AI software do you use to assist in your creations?
Blindart A:The maturity of our works and the formation of our style have been achieved through thousands of dialogues and interactions with AI. When we create, we always have a preliminary creative draft and continuously provide it with new prompts, refining it step by step. AI is not as simple as some believe; it requires more than just a simple phrase or a few sentences while one lounges with a coffee. To achieve stable and accurate results, one must continually explore and understand it. For example, if you want to generate a flower, you may need to input a lot of text, but most of that content may be unnecessary descriptions. Once you become familiar with AI, you will understand which content is crucial for generating images, allowing you to generate a flower with just a few words, significantly improving efficiency. Additionally, AI is an evolving tool; although it is not alive, every choice and instruction you give will accumulate over time and ultimately change its character. Even if different people use the same AI software with the same text inputs, the final images generated will still differ, which is quite magical. AI is not a cold machine; the longer you use it, the more it understands you, leading to the illusion that it is just a personal tool serving only you. Those who do not use AI for long cannot experience this wonderful psychological feeling; perhaps this is the bond that forms. We usually use the AI software Midjourney for our creations, which is also a common tool in the AI community. Langlang, who was a photographer before, uses Photoshop because the works generated by Midjourney may not have high clarity and size, so she enhances the quality of the works. If the generated images have flaws or oddities, they also need to be fixed using Photoshop. Business-type works require higher standards, so we will conduct secondary creation based on the generated images. In simple terms, we combine product materials with images generated by Midjourney, as previously mentioned, business works will always involve real product or model photography, so we cannot solely rely on Midjourney; we also need to create with Photoshop. These are the two tools we commonly use. I must emphasize again that tools are not the key to creation; the creator is.
When generating complex images with AI, multiple iterations are often required to refine the details. This process can reach many iterations. Because of this, the works do not exhibit the roughness previously associated with AI; behind this is the significant time investment and our sincere emotions towards this body of work. Therefore, using AI can be a very simple yet also very complex task, depending on how you choose to use it.
Q: In recent years, AI technology has not only become popular in our professional circles but can also frequently be seen in many popular entertainment and leisure software. For example, the AI function that generates Monet-style oil paintings on Douyin gained immense popularity shortly after its launch, attracting many people to create similar effect videos. It has a strong appeal for everyone. Those willing to try it are not limited to us; AI technology is quietly entering thousands of households. My final question is, from a long-term perspective, what significant impacts and benefits do you think AI will have on our entire industry?
This set of works would have incurred huge time and financial costs if completed entirely by makeup artists and stylists, as the current Chinese style is very popular among young people, and related industries are raising prices! If we can generate clothing, accessories, backgrounds, and props through AI, we can avoid such massive initial investments. So, in the future, if customers want to wear a luxurious outfit and jewelry for a photoshoot, even if it’s a gown worn by Princess Diana, we can meet all their requirements; how could there be a dissatisfied customer? Every order would be a good review..
Blindart A:I believe the biggest impact will be that more newcomers will be introduced to our industry through AI, stimulating industry development. We must recognize that for an industry to develop sustainably, it must attract a continuous influx of young people. The works produced by young designers differ entirely from those presented by experienced designers; this difference is not merely due to experience. The imaginative and novel ideas of young individuals are incredibly precious as they are fleeting. For an industry to break through existing bottlenecks, it requires not only the research and development of new technologies and the experience of older employees but also the rich ideas and bold creativity of young people. AI is currently helping young people eliminate many barriers. Who has determined that one must be proficient in various design software to possess boundless imagination? Perhaps that was the case in the past, but with the emergence of AI, it is no longer so. The emergence of AI will undoubtedly transform many industries, and many previous rules and requirements will no longer be applicable. For instance, our team consists of two girls. This represents a significant opportunity to break existing industry bottlenecks because the perspectives of women differ completely from those of men, leading to different feelings about things, and thus, the creations will also be entirely different. In a special type of video, male and female photographers were invited to shoot the same model, and we ultimately discovered that although it was the same model in the same scene, the model presented entirely different qualities through their lenses, resulting in completely different final products, each with its own strengths. This addresses a phenomenon currently present in the industry, which is that the themes and styles of works are too similar and lack freshness. If the birth of AI can keep the industry continually fresh, it can curb the frequent occurrence of plagiarism. AI has lowered the barriers to design, so it is no longer necessary to rely on complex design software to realize one’s creativity, which is beneficial to the industry, but it is also bad for some. Why are some people so hostile towards AI technology? Because AI technology has made them feel a sense of crisis; AI can achieve effects that took them years or even decades to learn to complete with complex operations in an instant. For example, complex cutout techniques can now be easily accomplished by AI! However, before this, cutouts were a complex and tedious task, which is why they could showcase an individual’s skills. When someone only knows how to use complex software and shortcuts but has an empty mind, they will find it challenging to have any appreciation for AI. Because when faced with AI, they don’t know what text to input since there are no standard answers for AI generation. Therefore, we have always emphasized that AI is not a lazy tool; rather, it is a tool that keeps the mind more active and busy. If we spend a lot of our work time on creativity and ideas instead of on the previously complex and repetitive basic tasks, the mundane work becomes much more interesting for those who truly love creation. Isn’t that a good thing?
Young people have the advantage of resonating with their peers, so the themes they create can resonate with the young customer base. For instance, today’s young people are keen on collecting figurines, but it is unlikely to have these items during a photoshoot. Although they may seem like mere toys, if they were to be the same size as the images generated by AI, they could cost thousands of dollars each; having a room full of figurines would be prohibitively expensive. As mentioned earlier, AI can reduce costs, but the key is whether you can think of these youthful elements while generating works. Even if a customer requests a figurine, if you are not familiar with it, the generated effect will reflect elements from your era, which the customer will not appreciate. However, if a young designer is tasked with the design, for example, if a customer says they want to be photographed holding a LABUBU like Lisa, the young designer would know exactly what they are talking about! Thus, after the emergence of AI, young people are actually the group that benefits the most, as they not only know more but also love to learn, which is why AI is said to keep the industry continually youthful.
2. Experience the Power of AI Technology
AI can do everything for us; any materials we desire can be generated as long as we dare to imagine, even things we have never seen in reality. Isn’t that cool? Let’s take a look at the world created by AI.
3. Summary and Reflections
I am very interested in AI technology because I have clearly felt the fatigue in our wedding dress industry over the years. Engaging with a fresh AI team has continuously surprised me; their works are not only particularly professional but also feature themes that are especially novel and interesting. Just the word ‘interesting’ is worth recommending. Perhaps they are newcomers just entering the industry, or perhaps AI creation is not subject to many restrictions, resulting in their works being different from what I usually see. Rather than saying their AI works are interesting, it is their interesting personalities that make their works enjoyable and engaging. If our industry continuously welcomes imaginative young people, even those of us who have been in the industry for a long time will be reactivated and inspired, experiencing a second youth. AI has provided many opportunities; it allows photographers who do not know Photoshop to communicate with retouchers not just verbally but by generating a rough effect image to show the retoucher, which is much better than relying solely on verbal descriptions. It also creates a new mode of interaction between us and our customers; for demanding customers, we can innovate through AI or use AI to modify unsatisfactory areas. AI can also remedy some regrets; for example, this series of works depicts the Forbidden City in Beijing. If some newlyweds wish to photograph there but have a tight schedule and live far from Beijing, they can find a similar ancient building in their city for the shoot, and then use AI to modify the details, transforming the local ancient architecture into the appearance of the Forbidden City. Such outdoor shooting cases are quite common; for instance, if a photoshoot in Sanya is scheduled but the weather is terrible, with overcast skies, AI can help remedy those regrettable moments.
While writing this article, I came across a piece of content on Xiaohongshu that opened up a new layer of thought for me. This content involved a customer generating an AI image for a dessert cake maker and then asking the cake maker to replicate it exactly. In the end, the cake maker indeed created an identical cake based on the AI image. Previously, customers often complained that we did not understand them, while we blamed the customers for making abstract and impractical requests. Could there be a chance in the future for us to ask customers to generate AI images for us to reference, or for us to generate AI images before the shoot for customer reference? While we have always had reference images, they were based on others’ pictures, so it is not surprising that the final results were similar. Every girl wants to look particularly beautiful! It is not unreasonable to ask you to recreate a piece featuring Dilireba. Of course, AI-generated images may not be 100% replicable, but compared to abstract textual concepts and images in our minds, actual images hold more value. If the customer generates an image of nine dragons coiling around a cake, they are well aware of the cost; even if it could be made, they would likely not order such a complex cake, showcasing the key role AI plays in eliminating impractical requests and helping us better understand customers’ requirements.
A few months ago, Douyin began featuring such AI videos, where images of deceased loved ones are combined with personal photos to generate video content of the two embracing. This truly elevates AI’s role and the significance behind it!
4. 6 Steps to Understand the Behind-the-Scenes Process of AI Creation
The most important aspect of the entire photoshoot is the initial creative planning. AI is just a tool; it cannot replace thought and creativity. If there is no rough image in your mind during the creative process, you will end up in a state of generating ineffective images with AI. Currently, the AI tools available do not yet achieve the state of “wanting to get what you want,” but with initial planning, you can limit the creative range of AI with language, bringing it as close as possible to the “image in your mind.”
2. Scene Creation with Midjourney
With the initial creative planning in place, the next step is to use Midjourney for scene creation. Transform your ideas into short phrases and input them into the Midjourney software, continuously adjusting the prompts to achieve the ideal image. In the AI photography segment, the focus should be on the composition and lighting of the scene. Generally, AI’s lighting aligns well with real light paths, so the focus is whether the light and shadow produced by Midjourney can be simulated under real studio conditions. Additionally, the scenes created by Midjourney, especially complex ones, often contain many distorted details, such as unclear backgrounds or distorted objects, but these detail issues do not need to be considered in this early stage.
3. Image Repair with StableDiffusion
After obtaining images nearing the ideal state, Midjourney enlarges them once, resulting in images around 1-2 MB in size. For creative purposes, this image size and precision are inadequate, and there will definitely be many imperfect details or obvious “AI feels” present. StableDiffusion is needed to reconstruct the details of the image while also enhancing its size.
4. Simulating Lighting in the Studio
After StableDiffusion repairs the image, the details and texture of the image are improved. The photographer needs to assess the current state of the AI image to determine the lighting and color tone of the subject. During the studio shoot, efforts should be made to replicate such lighting and tones.
Sometimes, AI models occupy space in the AI-generated images. If real models are to replace AI images, the AI model’s position will correspond to the real model’s placement, and the lighting on their faces and clothing must be restored or improved during the shoot. When discussing shooting, it is essential to mention the background color used in the studio. Many photographers might immediately think of using a green screen for shooting. However, in our operations, green screens are neither the most used nor the preferred choice. More often, we opt for a white background.
6. Post-Production Fusion
After the shoot, the real model or product is cut out and placed in the AI image. At this point, StableDiffusion can be used as a fusion method, or Photoshop can be utilized. Comparatively, using StableDiffusion to merge lighting is somewhat easier. However, since we have already simulated the lighting and basic tones in the earlier shooting steps, using Photoshop becomes much simpler as well.
Selected Portrait Photography
If you like it, please share it on your Moments
For more original works, please visit
Click the image to purchase immediately
