What is happening to photography? Hi my friends - I've been absent from the site for a few months trying to get my head around all of this AI stuff. Maybe I'm just getting old (I am)...

twosummers Loc: Melbourne Australia or Lincolnshire England

Hi my friends - I've been absent from the site for a few months trying to get my head around all of this AI stuff. Maybe I'm just getting old (I am) but it is truly frightening. From a photography interest perspective I have been looking at 2 applications - one is called DALL-E and the other is Midjourney. Here's the scary part - I spent some time writing a prompt (you have to tell these AI machines what to do) - let's say you want a photograph - it can be just a regular one or an exiting one (say for a poster or an advertisement). The prompts can be what you need the photo to depict, the aspect ratio, camera model, lens, aperture, filter, shutter speed, weather, location - you can be as creative as you like. So from a prompt of about 50 words I got this image in a few seconds! If you are a graphic designer (for example) you've just saved yourself a bunch of time and brainpower. I read today that images of people created by AI cannot be distinguished from real photographs.......

(Download)

Bill_de Loc: US

twosummers wrote:

Hi my friends - I've been absent from the site for... (show quote)

There a section here for AI.

Here's the link:

https://www.uglyhedgehog.com/s-143-1.html

----

Have fun

twosummers Loc: Melbourne Australia or Lincolnshire England

Oh no! you beat me to it! Thanks, I'll repost over there.

CHG_CANON Loc: the Windy City

Lower your expectations to match your AI results? You can't tell if it's real? The focus, perspective, the pixel-level details, and you're even hesitant? What ??

Toment Loc: FL, IL

If you like the result, what’s the difference how it was made. Your 50 words created it…

User ID

twosummers wrote:

Hi my friends - I've been absent from the site for... (show quote)

Welcome to the 21st century. Enjoy !

User ID

CHG_CANON wrote:

Lower your expectations to match your AI results? You can't tell if it's real? The focus, perspective, the pixel-level details, and you're even hesitant? What ??

Did you say something ? Anything ??

CHG_CANON Loc: the Windy City

User ID wrote:

Did you say something ? Anything ??

I haven't yet outsourced my wit to AI ....

Longshadow Loc: Audubon, PA, United States

CHG_CANON wrote:

I haven't yet outsourced my wit to AI ....

Bill_de Loc: US

CHG_CANON wrote:

I haven't yet outsourced my wit to AI ....

If/when you do, you will become our resident Nit-Wit!

---

StanMac Loc: Tennessee

twosummers wrote:

Hi my friends - I've been absent from the site for... (show quote)

Two summers, that is an amazingly detailed action image (notice I didn’t say photograph). I would be very interested in seeing your 50 words that were used to create this image. I have not used any AI application (not sure if I ever will) but I am curious about the inputs that drive the process that creates the imagery. Could you share those words, please?

Stan

Wallen Loc: Middle Earth

twosummers wrote:

Hi my friends - I've been absent from the site for... (show quote)

AI have improved by leaps & bounds, but I have yet to see an AI generated image that is indistinguishable from a real photo. To my eyes, there is always something off about their images.

twosummers Loc: Melbourne Australia or Lincolnshire England

StanMac wrote:

Two summers, that is an amazingly detailed action image (notice I didn’t say photograph). I would be very interested in seeing your 50 words that were used to create this image. I have not used any AI application (not sure if I ever will) but I am curious about the inputs that drive the process that creates the imagery. Could you share those words, please?

Stan

Here you go Stan (probably more than 50 words!) - Whilst this prompt uses Midjourney variables the image I posted above was produced by ChatGPT - bear in mind that if you use the same prompt you'll get a different result as everything (documents, images etc) are generated each time! - it's very clever. I'm buying into this tech and compiling thousands of prompts by category (I need to get out more)

Here is the prompt -

Capture a thrilling car chase scene on a busy city street. The cars should be moving at high speed, weaving through traffic with skyscrapers in the background. The lighting should be clear and bright, highlighting the action and creating a sense of urgency. Use a Canon EOS 5D Mark IV DSLR camera with an EF 70-200mm f/2.8L IS III USM lens to capture the car chase. --ar 16:9 --v 5.1 --style raw --q 2 --s 750

Enjoy

You can of course use this exact prompt a dozen or more times and get a dozen or more unique images. If you modify the prompt you'll get something else of course. Rinse and repeat

James

User ID

CHG_CANON wrote:

I haven't yet outsourced my wit to AI ....

Ahh ! A work in progress !

User ID

StanMac wrote:

Two summers, that is an amazingly detailed action image (notice I didn’t say photograph). I would be very interested in seeing your 50 words that were used to create this image. I have not used any AI application (not sure if I ever will) but I am curious about the inputs that drive the process that creates the imagery. Could you share those words, please?

Stan

Having that 50 words, you could submit them multiple times to see how the result changes.