Photographers Should Be Worried... and Excited by Generative AI | Midjourney and DALL-E2
HTML-код
- Опубликовано: 8 июл 2024
- In this video, we explore the impact of generative AI on photography and its possibilities for photographers. We compare the results of different AI generators such as DALL-E2 and Midjourney, highlighting their strengths and weaknesses in generating realistic human faces and stock photography. We also discuss how photographers can use generative AI to develop and share photography ideas and enhance their photography journey. Watch this video to learn more about the potential benefits and drawbacks of using generative AI in photography.
I encourage you to view the incredible work of the following artists mentioned at the top of the video:
👉 Noah Asanias: / aibynoah
👉 Jos Avery: / averyseasonart
👉 Nick St. Pierre: / nickfloats
If you're interested in learning about, or trying out Midjourney and DALL-E2:
🚀 Midjourney: www.midjourney.com/
🚀 DALL-E2: openai.com/product/dall-e-2
In the video, I mention several prompts. Prompts are human-readable instructions to the AI, telling it what to create. Here are some I used in Midjourney to create the images in the video:
----------------------
"studio style photo of model with brown medium curly hair, 1979 disco makeup, lying on stomach up on elbows facing camera wearing sparkling red spandex pants and a black mesh top, background half yellow half blue, shot on Fujifilm Superia 400, lighting studio light 2 feet off floor, shot type full length horizontal --seed 1 --v 5.1"
"studio style photo of female model in fancy red dress holding black balloons, beige seamless background, shot with Sony A7iii, 85mm lens f/9, bright studio lighting, full length --v 5.1"
"1940s, medium-closeup, ektachrome, 25-year-old woman, long blonde hair, big band, white party dress plunging neckline, fancy jewelry, hollywood nightclub scene --v 5.1"
"1980s, medium-closeup, kodachrome, teenagers, bright clothing, mtv style, shopping mall --v 5.1"
"[image URL] as female action hero, 2005, jungle, soldiers --v 5.1"
----------------------
👉 LEARN PHOTOGRAPHY HERE:
edverosky.com
Thank you, Ed. This was really enlightening for me, in terms of both A.I. as a useful enhancement tool and in context of art. I continue to think of photography as art and a record of the real world. Enhancement to emphasize the photographer's view of the subject or elicit an emotional experience is part of the art. It may be argued that a completely A.I. (fictionalized) image is no less valid that the painter expressing his/her interpretation of what their mind imagines. Yet, when a photographer captures a real human face and the life lived it expresses, or a spring blossom, a bird in flight, a newborn child or a graduation, or a family event, that buck standing in the silhouette of a sunrise in the morning mist, etc., to me these are genuine, true and a record of life. Looking at an early photograph of people who have lived and died often draw us in, offering a glance into a time before us, be they historical figures or an immigrant farmer or the face of Geronimo. A fictionalized image cannot - ever - be this.
Totally agree. Thanks for that. 🌟
Great information!!😮😮😮
I like photography,I like the process,I like getting out with my camera,I like seeing what I can find…..AI will never be able to produce pictures of my family and friends creating memories…..my nephews wedding,my daughter on her horse in our own arena…..just not interested in AI…..For me photography will never die as I’m sure it won’t for millions of others….specific genres may well be at risk eventually but lots won’t….this AI conversation on RUclips is getting pretty boring now.
Portable cameras didn’t kill the studio, instant didn’t kill film labs, auto didn’t take over our jobs, digital didn’t kill film, iPhones didn’t kill cameras, and generative AI is not actual photography. But many jobs photographers were traditionally needed for are going to go to digital creators.
We’re absolutely fooked
Hey Ed, great info. I do have a question about Midjourney and using it to superimpose a photo/portrait onto a different background:
>>Wouldn't your portrait image then become part of Midjourney's database of images it can reference from?
>>If, so, is that something you'd want (or the person who is in the photo)?
>>Also, aren't all the images publicly visible, or can you have a walled-off area where images you generate aren't seen by anyone?
I've only played with Midjourney once, and that was several months ago. They no longer have 'free' preview accounts, and I'm too cheap to join. :)
I’m only guessing, but I imagine any image posted publicly anywhere on the internet is potentially available to these systems. And I think I read that certain paid plans can operate in “stealth” mode so your images aren’t visible on the Midjourney site. Not sure exactly what that means but hopefully someone who’s more familiar can comment on this.
Generative AI is a challenge and an inspiration for photographers. I want to see how Midjourney will influence photography's style and meaning. But no AI can ever replace a photographer's perspective and creativity.
Old school photographer here. Glad I'm retired.
I think it’s all moving so fast now, nobody would blame you.
No, this is art. Is like you hire someone to create a picture based on your imagination. For all people without talent is something wonderful!
So far largely overrated. Using the service as a sketch board - yes, it does its job. Other than that - as a wedding and portrait / e-commercial / report photographer, I see no threat coming from ai. It might find its own niche.
A few years ago, I wrote about how computer generated images were being used in place of actual photography for products. But so far it’s actually more cost-effective for most brands to have actual photos created than hire out 3D art experts. I still see the most affected area as being commercial stock photos.