How will AI affect "making" versus "taking" photo's?

[Steven Lee]

In other words, I am stating that we have a promising explanation for ALL human behavior, which will reshape (started already) all branches of science.

 

[warden]

Potatoes can be creative and artful too

AI is a tool. Artists often employ tools.

Sometimes artists employ potatoes as tools. Ruth Asawa did so brilliantly.

Ruth Asawa | Untitled (SF.045c, Potato Print – Branches, Purple/Blue)

Ruth Asawa, Untitled (SF.045c, Potato Print – Branches, Purple/Blue), c. 1951–1952. Relief print, sheet: 13 5/8 × 9 in. (34.6 × 22.9 cm) Image: 13 5/8 × 9 in. (34.6 × 22.9 cm). Whitney Museum of American Art, New York; purchase, with funds from the Drawing Committee, the Director’s...

whitney.org

 

[Arthurwg]

Besides, if those models were so accurate, one would expect that fields like psychiatry and economics would have a much firmer grip on reality.

It's not the economists, it's the people who refuse to believe them.

 

[Alan Edward Klein]

Perhaps not yet, but eventually it very well might be. I’m not aware of any evidence the brain is more than a computational device (although there are plenty of conjectures). It’s exceedingly complex, but still a machine.

Also, Jackson Pollock could draw very well.

What distinguishes us is morality and knowing right from wrong. Would you want a soul-less machine deciding whether to go to war or execute another human based on some formula in its program? That's why machines can't be creative. They're formulaic. There is no room for heart.

 

[Milpool]

I understand where you’re coming from but there is no evidence to suggest the brain is anything more than an extremely complex, nevertheless soul-less machine. Neuroscience is still in relative infancy when it comes to the brain at a macro level, so it will probably be a while, but the things you’re calling distinguishing characteristics are in all likelihood just a matter of complexity.

What distinguishes us is morality and knowing right from wrong. Would you want a soul-less machine deciding whether to go to war or execute another human based on some formula in its program? That's why machines can't be creative. They're formulaic. There is no room for heart.

 

[Alan Edward Klein]

I understand where you’re coming from but there is no evidence to suggest the brain is anything more than an extremely complex, nevertheless soul-less machine. Neuroscience is still in relative infancy when it comes to the brain at a macro level, so it will probably be a while, but the things you’re calling distinguishing characteristics are in all likelihood just a matter of complexity.

Even if you're right, and I disagree, machines can't have that kind of complexity. They're really Mr Potato Head.

 

[Ron789]

I find it interesting to see that the discussion triggered by my initial post has evolved towards the role of AI in "art" (photographic or other).
I consider many (if not most) professional photographers to be craftsmen rather than artists: product photography, fashion, food, etc. images, all for advertising, marketing and sales purposes.... Though some may bring such photography to a level that one might call "art", most produce fairly traditional images, responding to the demand and preferences of their clients. They often do that with great craftsmanship but the artistic content is typically very thin, if any at all. I think those photographers will face hard times since AI can respond to the demand and preferences of those customers faster, more flexibly and cost effectively than a human photographer can.... "making" photos.
I do see a future for family, portraiture, journalism, sports and documentary etc. photography, since this requires the photographer to be present and interact with the subject.... "taking" photos.
So, my opinion regarding my own question: many of those who "make" photos may be facing hard times; this who "take" photos will be less affected by AI.

 

[warden]

I find it interesting to see that the discussion triggered by my initial post has evolved towards the role of AI in "art" (photographic or other).
I consider many (if not most) professional photographers to be craftsmen rather than artists: product photography, fashion, food, etc. images, all for advertising, marketing and sales purposes.... Though some may bring such photography to a level that one might call "art", most produce fairly traditional images, responding to the demand and preferences of their clients. They often do that with great craftsmanship but the artistic content is typically very thin, if any at all. I think those photographers will face hard times since AI can respond to the demand and preferences of those customers faster, more flexibly and cost effectively than a human photographer can.... "making" photos.
I do see a future for family, portraiture, journalism, sports and documentary etc. photography, since this requires the photographer to be present and interact with the subject.... "taking" photos.
So, my opinion regarding my own question: many of those who "make" photos may be facing hard times; this who "take" photos will be less affected by AI.

I think photographers that offer headshots are already feeling pressure from AI.

 

[MattKing]

I think photographers that offer headshots are already feeling pressure from AI.

The photographers that work in advertising are the most vulnerable.

 

[Sean]

AI has been helpful in my overcoming the learning curve phase of many apps, which definitely includes "Capture One" image editor. It can be a real slog getting through the learning curve on these vast image editing applications. Every time I get stuck I ask chatgpt something like, "In Capture One what are a few ways to get rid of a blue shift in my image", and it will quickly spit out a 'more than one way to skin a cat' style answer with step by step. I pick the one that makes the most sense to me and that I'm more likely to remember. I'm starting to ask it less and less. So in the context of AI helping me photographically, this is a slight yes.

 

[koraks]

and I will happily state that emotional needs and interactions are also just searches, and they are rational and not fuzzy at all.

If you've not yet read 'Brave New World' (which I find hard to imagine), I implore you to keep reading it until this 'happy' conception wears off.

 

[TomR55]

In other words, I am stating that we have a promising explanation for ALL human behavior, which will reshape (started already) all branches of science.

Now, by a promising explanation of “ALL human behavior” I assume you mean a “scientific explanation.” Assume that this is true. Then “reshaping science” follows, and is a good thing because it fits within the modus operandi of science. I don’t know that the scientific method, which is reductionist at its core, or its products satisfies the “human” need for a meaningful existence. Perhaps this doubt underlies a lot of the discussion on this thread.

Scientism has been in and out of vogue for some time (over a century). In the early twentieth century, some believed that a meta-mathematics was just around the corner. Godel proved this impossible within the definitions, axioms and theorems of formal logic, which is one of the foundations of mathematics and, by extension, science. A core group of technocrats today believe that the application of science to social and political problems is the only way out of our current situation.

I am unconvinced … and lately concerned.

Again, I am not saying you’re wrong about a possible future for science. Rather I am suggesting that scientific explanations of all phenomena might not satisfy the human need for “something more.” Maybe a “purpose or mystery" (the teleological) pursuit of these intangibles (even if proven to be products of some algorithm) are essential to human happiness—as fleeting as that might be.

 

[koraks]

I don’t know that the scientific method, which is reductionist at its core, or its products satisfies the “human” need for a meaningful existence.

Nor this supposed unraveling of the richness and mystery of the human experience through scientific methods something that I've seen many (or, come to think of it, any) person working within the scientific arena seriously purports or even believes. It seems to be mostly a phenomenon attribute to science by some, rather than an ambition that is noticeably existent within the scientific arena.

On a related note, I sometimes like to think that we may be capable of modeling fairly complex systems (biological, social, ecological etc.), but that the truly representative model will turn out to be of equal complexity to the reality we've tried to model, resulting in the net gain of our actual understanding being more or less zero. Replication without understanding - the intellectual parallel to the cycle of life that has been around or 3.5 billion years on this planet. We're gaining on you, God/Spaghetti Monster/whomever!

 

[Milpool]

Even if you're right, and I disagree, machines can't have that kind of complexity. They're really Mr Potato Head.

That's a view. Some/many/most (?) humans seem to have some desire to think they are beyond Mr. Potato Head at some "fundamental" level etc. Whatever works.

 

[Alan Edward Klein]

That's a view. Some/many/most (?) humans seem to have some desire to think they are beyond Mr. Potato Head at some "fundamental" level etc. Whatever works.

AI was developed by humans.

 

[MattKing]

AI was developed by humans.

So was Frankenstein.
And Barbie.

 

[Milpool]

AI was developed by humans.

Well I can't argue with that.

 

[Steven Lee]

Again, I am not saying you’re wrong about a possible future for science. Rather I am suggesting that scientific explanations of all phenomena might not satisfy the human need for “something more.” Maybe a “purpose or mystery" (the teleological) pursuit of these intangibles (even if proven to be products of some algorithm) are essential to human happiness—as fleeting as that might be.

TLDR: There are many unsolved mysteries in the universe for us to be excited about. But our brains are no longer on that list.

The AI technology merely models our brains, and they aren't the most interesting or even complex thing out there. I suppose the core disagreement that I have with koraks is the premium he places on the "richness and mystery of the human". I no longer see anything rich or mysterious. Looking at the mountains of evidence we're rapidly producing, I am becoming more and more convinced that we're just moist and unreliable pattern matching machines with fairly unsophisticated value functions, mostly related to imitation according to Gerard. I am not a believer in "art" either, -- it's just the oldest surviving form of product positioning. Mona Lisa is no different from a Macbook. And that's why so many "artists" feel threatened by AI, apparently they are not feeling any advanages offered by "richness and mystery" of their own brains.

Consider the following thought experiment. Suppose you have a baby born in the summer of 2025. Mount a stereo camera and a couple of microphones on his head, that record everything that's happening around him and store it forever. Have him wear them nonstop 24/7 for the next 15 yeras, capturing 100% of information that enters his brain. I think it's very likely that a model trained on his data will be a full digital replica of that person. To validate its completeness, you present the model and the 15 year old teenager with a situation/question/dilemma and they'll have identical response. Our understanding of how a models operate should evolve by then as well. The field of mechanistic interpretability is rapidly evolving.

 

[koraks]

I no longer see anything rich or mysterious.

Uhm. Okay. Sorry you feel that way.

I understand what you say. 15 years ago I might have said the same. We all learn.

Good luck with the mechanistic interpretability. Please consider my reading suggestion, though. Ideas like yours are kind of creepy. Not just because they're wrong (they are), but because there's probably quite a few people who are as misguided by them as you appear to be - for now (I've not yet given up hope on your case).

PS: in your thought experiment about the baby and the model - don't forget to ask both the 15-year old human and the model how they feel about themselves. Then ask yourself how you feel about their answers.

 

[Alan Edward Klein]

...

Consider the following thought experiment. Suppose you have a baby born in the summer of 2025. Mount a stereo camera and a couple of microphones on his head, that record everything that's happening around him and store it forever. Have him wear them nonstop 24/7 for the next 15 yeras, capturing 100% of information that enters his brain. I think it's very likely that a model trained on his data will be a full digital replica of that person. To validate its completeness, you present the model and the 15 year old teenager with a situation/question/dilemma and they'll have identical response. Our understanding of how a models operate should evolve by then as well. The field of mechanistic interpretability is rapidly evolving.

You mean the 15-year-old teenage computer model will be just as incomprehensible as my daughter was when she was a 15-year-old teenager? That isn't much progress.

 

[BMbikerider]

Ai is not nor will ever be, in true sense of the word, a category of art, but eventually it will be called that, and will sell as such, and there will be opportunists who will take advantage of it, faking it up even further.

But there is a very simple solution - stay away from it, make your own images as you know how, and don't bother thinking of Ai as being a competitor. If you bring yourself down to that level, you don't respect your own work.

Your second paragraph outlines exactly what I intend to do.

 

[Steven Lee]

Uhm. Okay. Sorry you feel that way.
I understand what you say. 15 years ago I might have said the same. We all learn.

See? Now you too have learned what Internet forums are for, and why moderation is harmful.
You can't enjoy subtly insulting strangers on the Internet when moderators are present! ;-)

 

[TomR55]

TLDR: There are many unsolved mysteries in the universe for us to be excited about. But our brains are no longer on that list.

The AI technology merely models our brains, and they aren't the most interesting or even complex thing out there. I suppose the core disagreement that I have with koraks is the premium he places on the "richness and mystery of the human". I no longer see anything rich or mysterious. Looking at the mountains of evidence we're rapidly producing, I am becoming more and more convinced that we're just moist and unreliable pattern matching machines with fairly unsophisticated value functions, mostly related to imitation according to Gerard. I am not a believer in "art" either, -- it's just the oldest surviving form of product positioning. Mona Lisa is no different from a Macbook. And that's why so many "artists" feel threatened by AI, apparently they are not feeling any advanages offered by "richness and mystery" of their own brains.

Consider the following thought experiment. Suppose you have a baby born in the summer of 2025. Mount a stereo camera and a couple of microphones on his head, that record everything that's happening around him and store it forever. Have him wear them nonstop 24/7 for the next 15 yeras, capturing 100% of information that enters his brain. I think it's very likely that a model trained on his data will be a full digital replica of that person. To validate its completeness, you present the model and the 15 year old teenager with a situation/question/dilemma and they'll have identical response. Our understanding of how a models operate should evolve by then as well. The field of mechanistic interpretability is rapidly evolving.

Don’t see any of this approaching “proof.” But it demonstrates that a little knowledge is a dangerous thing.