This may be the first profession to be subverted by AI | Ai Faner

The model before you, wearing a white T-shirt and denim midi dress, is wearing a standard professional smile. If the picture appears in a certain treasure, there is a high probability that it will be submerged in colorful colors and will be mechanically drawn by our fingers.

Stop and take a look, and you may feel that something is not right, her limbs are slightly stiff, the shadow under her neck is a bit strange, and the boots are of the same style as a low-quality Barbie doll. Yes, this is an imperfect AI model, but it will become more and more real in the future.

I am a model and I know that AI will eventually take my job.

Canadian model Sinead Bovell wrote this article for Vogue three years ago. Now, her prophecy appears to be coming true.

AI models are crazy

This AI model comes from the denim brand Levi’s.

In March, it reached a cooperation with AI model company Lalaland, and plans to use AI-generated models this year.

Authenticity is not in the same dimension for the time being, and AI models compete with humans in speed and variety.

According to Lalaland, the AI ​​models are inspired by real people but generated by AI.

Every aspect of them can be customized, including hairstyle, body shape, skin tone, posture, mood, etc., and the whole process even takes less than 5 minutes.

More AI model companies are growing like mushrooms after rain.

The start-up company Botika has launched an AI model service, which plans to charge $15 a month, mainly for small businesses. It is still in the startup stage, and the waiting list has reached 1,000 people.

In the video on the official website, a photo covering the face is placed on the head, and the foolish operation focuses on a silky smooth:

Gone are the days when one model fits all. Change models based on facial expression, skin tone, hairstyle, background, and more with the click of a button.

The AI ​​content generation website ZMO, the schematic diagram makes the folds and touch feel good, but the background is too fake without shadows. At present, only the waiting list is open, and the promotional copy really wants people to queue up overnight:

Put your products on models in minutes, not weeks!

The Dutch start-up company Deep Agency even calls itself an “AI photo studio and model agency” in a high-profile manner.

Their service sounds cheap: pay $29 a month, enter prompts, adjust backgrounds, lenses, poses, etc., and you can use AI to generate models.

In all fairness, the generated pictures are stiff and plasticky, and the price is more expensive than ChatGPT Plus.

Looking back at China, many e-commerce practitioners are also exploring independently with AI tools such as Stable Diffusion, and even re-developed the AI ​​model dressing system based on these tools. As early as March, someone made a highly completed AI model map.

There was such a hot search in Xiaohongshu before: “Is this the first profession to be subverted by AI?” It was about Taobao models.

But in the view of Simon Chambers, executive of the modeling agency, the usefulness of AI models is still very limited, neither emotional nor creative:

AI mannequins are best suited for basic imagery that is used for reference, not marketing or promotion that builds relationships with customers.

At present, most AI models can only take the low-end basic model, but the important thing is the possibility of the future. Even if they are shoddy, they have already begun to be commercialized.

Original artist and Hollywood screenwriter’s successor

The reason for the rise of AI models and startups is simple: where there is demand, there is supply.

Hiring professional models is a big expense for SMEs. Tracy Porter is a self-employed clothing brand. In order to save money and do modeling herself, she has previously hired short-term photographers, models, hair stylists and makeup artists, but it costs $5,500 for a shoot.

The domestic e-commerce practitioner “Langzhu Huitou”, who explored AI models earlier, also said: “This year we don’t need to spend 40,000 to 50,000 a day to hire photographers and models… It’s enough for us.”

The e-commerce business that AI is aggressively invading is precisely the most reliable source of income for models. Not to mention complete replacement, the emergence of AI has given the market more choices, which may reduce the treatment of human beings. Hollywood screenwriters and original artists are examples from the past.

Hollywood screenwriters are protesting against the participation of AI in creation. They are not afraid that AI will replace them, but that editors will write better by themselves, but they will be forced to rewrite the garbage generated by AI. In the process, their wages will depreciate in disguise.

Similarly, many original painters have become AI repainters. Party A feels that the image has already been generated, and the original painter only needs to simply repair it, and the price is kept very low, resulting in the hourly salary of the retouching being far lower than the original one.

Some people now think that instead of using immature AI models, it is better to directly hire models who have just entered the industry, and the average time and cost are not necessarily higher.

AI models not only affect human models, but the cloud of unemployment hangs over the entire industry chain. Makeup artists, hair stylists, photographers and even creative personnel may be affected.

Fashion magazine Vogue has used tools such as MidJourney and Dall-E to put AI models on the cover in its spring issues in Brazil and Singapore.

AI is proving itself, whether it can work in e-commerce photography or dabble in difficult fields such as fashion magazine covers.

The first to bear the brunt are those unknown models who are regarded as the “replacement” of AI models. In many places, including the United States, models are employed as independent contractors. So, when the wave of AI models hits, they may be hard to protect.

This is what Os Keyes, an AI ethicist at the University of Washington, puts it, that ordinary models, photographers, and artists are particularly vulnerable to generative AI because they lack the power to structure:

Apps such as Deep Agency, if they work, will further displace already precarious creative workers and divert money to such startups, making profits more concentrated.

On the contrary, the supermodels who dressed up to attend the Met Gala and were active in the red carpet and big shows can temporarily rest assured that their personal images are still valuable in the face of technology.

The CGI mannequins from a few years ago are a reference. Wearing Prada, Dior and Gucci, they are personalized by stories made up by humans, creating rich commercial profits.

A successful model is already planning for a rainy day. At the end of April, 50-year-old supermodel Eva Herzigová claimed to have participated in body scanning, motion capture and other shootings, and plans to create a digital avatar “MetaHuman” and use it in future advertising campaigns.

“We can completely change her appearance, from hair, clothes, makeup, poses to virtual environments, with the click of a button.”

Convenience is one thing, but whether fans accept it is another. People may no longer believe that “there are pictures and the truth”, just like facing the “arrested” Trump and the “wearing down jacket” Pope. In the comments section of MetaHuman, a fan left a message:

Miss your soul and the twinkle in your eyes.

AI models are not synonymous with diversity

There’s no shortage of clamor for AI models, as they’re a quick way to show more faces, skin tones, and body shapes.

Part of the reason Lalaland CEO Michael Musandu started his business was that he had difficulty finding suitable models himself:

As a boy growing up in Zimbabwe, I rarely saw anyone who looked like me in photos, commercials or runways.

But the diversity of AI models has also been questioned by people in the fashion industry.

Levi’s, which announced the introduction of AI models, was brought to the forefront because of this matter. People are outraged that this giant company with sales of $6.2 billion in 2022, which talks about being diverse, is unwilling to directly hire diverse models.

Levi’s, which smells the crisis, will not reduce the on-site shooting of human models. The role of AI models is to supplement rather than replace. When each product is shot by one or two human models, they are responsible for adding icing on the cake and producing more variety faster Photo.

These words are not convincing to the models who have been involved in the vortex.

In fact, it is precisely marginalized models such as people of color and plus-size models who are most likely to be affected by the wave of automation, because they are more likely to engage in lower-level jobs and have fewer opportunities to choose from.

According to Vogue’s Size Inclusivity in Business report, only 0.6 percent of the looks shown in last year’s autumn/winter seasons, which included New York, London, Paris and Milan, were plus-size. In March, the Melbourne Fashion Festival was criticized for a lack of disabled models.

Yet ironically, brands ignore real minorities and use AI models to represent minorities. According to Richard Lachman, a professor at Toronto Metropolitan University, the trend is disturbing:

Brands want to create models that appear to fit society’s profile, with all skin tones, body shapes, and ethnicities, but it’s creating an illusion of idealization. AIs aren’t real people, and they don’t actually enrich diversity and increase the number of jobs.

At the same time, the AI ​​model itself will also be affected by the bias of data and algorithms, and then some biases will be aggravated.

Vice reporter Chloe Xiang found that Deep Agency generates blond white female models by default. If you want to change the appearance of the model, you must add additional words such as race and age. Then there is the possibility that among the thousands of extra AI pictures, women with other hairstyles and races will be even rarer.

So, some experts suggest that AI models should be staffed by a diverse behind-the-scenes team, including marginalized human models who are paid to provide shoots to make up for the lack of demographic data in the fashion industry. But a gray area ensues: who actually owns the data.

When we discuss the various impacts of AI models on the fashion industry, Canadian model Sinead Bovell always asks:

It’s good to see diversity, but who exactly gets the benefits and who gets the automation?

What Joseph Schumpeter called “creative destruction” may hold the answer: the arrival of a new invention, like electricity and appliances from the 1880s to the 1920s, is like a gust of wind blowing through industries and entire economies .

On the one hand, new technologies eliminate certain commerce and industries that exist widely in the economy; on the other hand, new technologies and new industries replace those technologies or industries that have collapsed.

Cheaper, faster, and more diverse AI models can certainly speak for minority groups, but they often do not benefit these groups themselves, at least they cannot directly benefit, and the benefits are more likely to be attributed to the technical elite.

In the long run of history, technology will benefit mankind, and all people can benefit from it. But in the not-too-short transition period, someone will act as the silent price, just like tears disappearing in the rain.