|
|
|
|
Dear all,
I was recently surprised by seeing that one of my photos, submitted under Landscape, got points from AI; maybe this was advertized in the magazine, I don't know. The funny thing is that this particular photo was rejected, with 2% from members and 1% from experts, with a comment that there was nothing in it and it needed improvement in almost every respect, in spite of that it got 95% from AI 😆
Hi José,
very funny indeed, I had a rejection with 4% from members, and 42% from experts, from AI it got 14%. Don't know what to think of it.
Regards,
Jan
Quite frankly I think curation sucks!
Dear all,
I was recently surprised by seeing that one of my photos, submitted under Landscape, got points from AI; maybe this was advertized in the magazine, I don't know. The funny thing is that this particular photo was rejected, with 2% from members and 1% from experts, with a comment that there was nothing in it and it needed improvement in almost every respect, in spite of that it got 95% from AI 😆
Hi José,
can you post a screenshot ? Never heard about AI in curation so far.
Cheers, Hans-Martin
Quite frankly I think curation sucks!
José,
Curation is the DNA of 1x.
When I joined 1x in 2013 I was fascinated by works that made it through curation. I soon realizized my work was not at the level to become awarded. Not only rejections told me, if I was honest to myself the photography I did was below average (for 1x) then the time.
It was only when I discovered the critique forum in 2017 when I changed my modus operandi. Writing critiques forced me to look deeper, analyse why photos fascinated me, what was needed to take them and how they can be improved in post to have that wow-effect making them stand out from mediocre work.
I once read a statistic that said we're exposed to roughly 3000 images every day, pulling and fighting for our attention. Of course we can't remember all images, our brain filters probably 99% of them. But there are these few that stick. Try to find out what makes them stick for you. That probably leads you to techniques helping you to create photos not "being filtered" for others. It can be colors, layout, perspectives, light, tones, the subjects, mood, bokeh, there are many things and tools to learn.
Complaining about rejection is a very common thought. I did, I admit that.But it's confusing root cause and effect to a certain extend. It took me several years to reach a level where my work is largely accepted by curation, and I did not lose interest in photography or felt forced to work in a way I did not like. I enjoy learning the most in photography, and as I can see from your posts in critique, you're willing to learn as well. Maybe you just need a little more time to fascinate other photographers(!) by your doings. 1x is a hard environment to prove yourself, there is a lot of competition.
People invest a lot of time, have excellent skills, brilliant ideas to get to the photos ultimately shown on the front page. And nobody is never rejected, despite all efforts. Like I said, that's the DNA of 1x, curation.
BR,
Mike
Dear all,
I was recently surprised by seeing that one of my photos, submitted under Landscape, got points from AI; maybe this was advertized in the magazine, I don't know. The funny thing is that this particular photo was rejected, with 2% from members and 1% from experts, with a comment that there was nothing in it and it needed improvement in almost every respect, in spite of that it got 95% from AI 😆
Hi José,
can you post a screenshot ? Never heard about AI in curation so far.
Cheers, Hans-Martin
Good morning all,
A recently published photo with 65%|85% generated the following AI rating.
Just wondering if and how this is weighing in for publication.
Have a nice Sunday!
Kind regards,
Gerda
Dear all,
I was recently surprised by seeing that one of my photos, submitted under Landscape, got points from AI; maybe this was advertized in the magazine, I don't know. The funny thing is that this particular photo was rejected, with 2% from members and 1% from experts, with a comment that there was nothing in it and it needed improvement in almost every respect, in spite of that it got 95% from AI 😆
Hi José,
can you post a screenshot ? Never heard about AI in curation so far.
Cheers, Hans-Martin
Good morning all,
A recently published photo with 65%|85% generated the following AI rating.
Just wondering if and how this is weighing in for publication.
Have a nice Sunday!
Kind regards,
Gerda
Thank you, Gerda. I'm not active in landscape photography, that's why I did not see it. Maybe in the future the other genres are evaluated by AI, too.
Kind Regards, Hans-Martin
I deleted the photo from 1x, because I can see no point in keeping a photo that no one else can see, but I've attached a low res version of it.
I deleted the photo from 1x, because I can see no point in keeping a photo that no one else can see, but I've attached a low res version of it.
Thank you, José.
I wasn't primarily interested in the specific photo, but in the presentation of the AI result (as Gerda posted, see above).
Good light, Hans-Martin
I did not even notice that "AI" score until now - when I was looking for it. It's more "artificial" than "intelligent", I'd say. In my case the mismatch of curation score and the how-ever-generated score is even more significant. I doubt it is good for anything at this stage.
Would be intersting to hear why it was introduced and what it is expected to do - when it became a bit smarter.
Regards,
Mike
I did not even notice that "AI" score until now - when I was looking for it. It's more "artificial" than "intelligent", I'd say. In my case the mismatch of curation score and the how-ever-generated score is even more significant. I doubt it is good for anything in this stage.
Would be intersting to hear why it was introduced and what it is expected to do - when it became a bit smarter.
Regards,
Mike
Another curation mystery added. 🤷♀️
Imho AI should never be the biggest piece of the pie for something so subjective (unless it looks for objective quality indicators, but even those can vary), hence my curiosity for the weight in curation and also how this technically works in terms of curation. I hope it's in a testing phase now, otherwise I have some concerns about it. It can do a fair bit, but certainly not replace good taste. 😉
Let's see if they'll maybe cover it in an article as well.
I have just been out, Spring is in the air, the finest time for floral macros! 🌸
Enjoy the remainder of your Sunday!
I guess not all members can see this AI score or its not rolled out to every member?
I guess not all members can see this AI score or its not rolled out to every member?
Only in the category Landscape thus far.
I guess not all members can see this AI score or its not rolled out to every member?
Only in the category Landscape thus far.
Thank you for the clarification Gerda 😊
Well, if our photographs are to be judged by algorithms and Artificial Intelligence, we'll need smarter cameras - cameras that can communicate in their language.
:-)
I did not even notice that "AI" score until now - when I was looking for it. It's more "artificial" than "intelligent", I'd say. In my case the mismatch of curation score and the how-ever-generated score is even more significant. I doubt it is good for anything in this stage.
Would be intersting to hear why it was introduced and what it is expected to do - when it became a bit smarter.
Regards,
Mike
Another curation mystery added. 🤷♀️
Imho AI should never be the biggest piece of the pie for something so subjective (unless it looks for objective quality indicators, but even those can vary), hence my curiosity for the weight in curation and also how this technically works in terms of curation. I hope it's in a testing phase now, otherwise I have some concerns about it. It can do a fair bit, but certainly not replace good taste. 😉
Let's see if they'll maybe cover it in an article as well.
I have just been out, Spring is in the air, the finest time for floral macros! 🌸
Enjoy the remainder of your Sunday!
I noticed it too.
I do not like this movement, that in a digital dictatorship where we live in, artists (that are particularly human) that try to connect with human around us by sending emotions, messages that computers do not understand (yet), are judged by this artificial 'intelligence'.
I think art counterbalances the digital environment where we live in by its human character, it's capability to shock, to go beyond the bounderies.
AND: If we are curated by robots, than my 'worse' photos will be on front page and my best photos go into the bin (like Mike's). How intelligent is that?
I also find it very confusing all these different stats, but maybe I am not intelligent enough ;-)
Sorry, I am emotional, not a robot :D
Cheers!
Elisabeth
some links from our fellowphotographers on the subject:
and a lot of others :D
I did not even notice that "AI" score until now - when I was looking for it. It's more "artificial" than "intelligent", I'd say. In my case the mismatch of curation score and the how-ever-generated score is even more significant. I doubt it is good for anything in this stage.
Would be intersting to hear why it was introduced and what it is expected to do - when it became a bit smarter.
Regards,
Mike
Another curation mystery added. 🤷♀️
Imho AI should never be the biggest piece of the pie for something so subjective (unless it looks for objective quality indicators, but even those can vary), hence my curiosity for the weight in curation and also how this technically works in terms of curation. I hope it's in a testing phase now, otherwise I have some concerns about it. It can do a fair bit, but certainly not replace good taste. 😉
Let's see if they'll maybe cover it in an article as well.
I have just been out, Spring is in the air, the finest time for floral macros! 🌸
Enjoy the remainder of your Sunday!
I noticed it too.
I do not like this movement, that in a digital dictatorship where we live in, artists (that are particularly human) that try to connect with human around us by sending emotions, messages that computers do not understand (yet), are judged by this artificial 'intelligence'.
I think art counterbalances the digital environment where we live in by its human character, it's capability to shock, to go beyond the bounderies.
AND: If we are curated by robots, than my 'worse' photos will be on front page and my best photos go into the bin (like Mike's). How intelligent is that?
I also find it very confusing all these different stats, but maybe I am not intelligent enough ;-)
Sorry, I am emotional, not a robot :D
Cheers!
Elisabeth
some links from our fellowphotographers on the subject:
and a lot of others :D
Yes, curation is the heart of 1X, and yes, the volume of pictures is a big problem. The curation system at 1x seems to have some severe technical issues with its architecture ever since the introduction of the new version. These difficulties, IMHO, should be solved before introducing another new feature like AI. However, if the strategy for solving the existing problems consists of replacing the curation system by AI, then the core USP of 1X, curation by humans, be they members, experts or head curators, will be no longer existent.
It would be helpful to have some information about the intentions of the founders in this regard.
Regards,
Jan
Hi Jan,
you can handle this problem in two (or maybe more) different ways. Either you limit the uploads of each member (e.g. ONE photo per month), then you possibly are able to curate the photos by humans, or, if you allow more uploads you have to employ more people or use AI. The 15 head curators all are not paid (except they save the fee for the PRO accounts). Even if a head curator starts working early in the morning and stops late at night, it would be impossible to handle the current upload volume.
Last year the CEWE company in Germany was proud to announce that in a photo contest led by CEWE more than 600,000 photos were uploaded.
Now I ask you: What do you think how these 600,000 photos were evaluated ? By whom ? By CEWE employees ? By AI ? I don't know. But I don't care either.
I personally don't take part on these ridiculous contests.
Since the beginning of my membership in 2013 I sent 276 photos to the curators, about 30 each year. I think that's enough for building a representative portfolio for a gallery showing a photographer's work.
Good light, Hans-Martin
Member since 2017 - About 50-60 photos were uploaded and some were uploaded for critique only. I would not like AI to decide go/no go... because I fear that all the images would evetually look the same or would look like the "current trend". I'd personnaly would rather see less uploads if that is what is becoming a problem. I take great pride in a photo that's published or even better "awarded" because I know that expert actually looked at it. I have a flickr account for the rest of the non-publishable 1X photos...some made it to "explore" along with the school bus images 😃 but that is an example of AI for you.
Pierre
Hi Hans Martin,
Strongly agree with you about the numbers. Personally, I did upload roughly the same number of photographs since 2013, and I sure don't like the contests you mentioned.
I think the introduction of the now existing curation system with member, expert and head curators was a good idea to lower the workload for the head curators. I appreciate the work of the head curators very much, their work is very demanding and is the main USP of 1X. Many of the head curators are engaged in this work since I joined in 2013, which is quite a commitment.
But this system should be implemented technical properly, which very obviously is not the case.
It is impossible to diagnose the problems without knowing the details, but as a former programmer and consultant to high-tech projects, I am aware of the tendency to solve architectural problems with new features. I sincerely hope this is not the case with the introduction of the AI algorithms here.
I enjoy being almost 10 years a member of 1X. I still think it is the best photo site because of the quality of the pictures presented here, the superb magazine, and several other aspects of the site. That's why I renewed my subscription this month. But I think, as paying members, we are entitled to a functioning and transparent curation process.
Regards
Jan
Hi Hans Martin,
Strongly agree with you about the numbers. Personally, I did upload roughly the same number of photographs since 2013, and I sure don't like the contests you mentioned.
I think the introduction of the now existing curation system with member, expert and head curators was a good idea to lower the workload for the head curators. I appreciate the work of the head curators very much, their work is very demanding and is the main USP of 1X. Many of the head curators are engaged in this work since I joined in 2013, which is quite a commitment.
But this system should be implemented technical properly, which very obviously is not the case.
It is impossible to diagnose the problems without knowing the details, but as a former programmer and consultant to high-tech projects, I am aware of the tendency to solve architectural problems with new features. I sincerely hope this is not the case with the introduction of the AI algorithms here.
I enjoy being almost 10 years a member of 1X. I still think it is the best photo site because of the quality of the pictures presented here, the superb magazine, and several other aspects of the site. That's why I renewed my subscription this month. But I think, as paying members, we are entitled to a functioning and transparent curation process.
Regards
Jan
Hi Jan,
I agree with your post, but would like to add something. You mentioned adding new features and AI algorithms. New features are put on websites and cameras to placate buyers when developers are trying to buy time and they are between a rock and a hard place. AI algorithms are fine as long as they are not too narrowly defined or too precise, which might cause photos to be rejected very quickly. This could be reflected in the big differences between the member curation % and expert curator %.
AL
Exactly my sentiments, Al.
Regards
Jan
Hi Al,
To me it appears as if it's not given a lot of weight at this moment. Going by Adam Dauria's rant his photo got awarded despite a very low AI rating and I had a photo with members rating 4%, experts rating 46% and AI rating 98% that was rejected (as it would have been without). I consider it a funny experiment for now. I think it can measure very objective things and needs help on the subjective side. At least it gave us something to talk about. 😉
Have a nice Sunday, all!
Kind regards,
Gerda
The AI rating is new to me; probably because I haven't posted a landscape shot in some time.
Given the mismatch between "live people" and AI why publish the results. It only serves to reduce the credibility of one or the other or possibly both. They should not have rolled this out publicly yet. A better approach would have been to gather data behind the scenes and work on getting better correlation between the "live people" and AI results. Once (if ever) there was a reasonable correlation than it could be rolled out. Going live with AI at this point is ludicrous.
Mike
The AI rating is new to me; probably because I haven't posted a landscape shot in some time.
Given the mismatch between "live people" and AI why publish the results. It only serves to reduce the credibility of one or the other or possibly both. They should not have rolled this out publicly yet. A better approach would have been to gather data behind the scenes and work on getting better correlation between the "live people" and AI results. Once (if ever) there was a reasonable correlation than it could be rolled out. Going live with AI at this point is ludicrous.
Mike
Hi Mike,
I agree with your post. I will add that I do not think that the correlation between the two will ever be very high. The members make subjective decisions. The algorithm makes objective ones. The algorithm can only measure things that are measurable.
In my opinion, what they should do is have a scale or assign weights to member curation and the algorithm indicating how important each one is. The curation result would then be made up of those two things and passed on to the expert curators.
AL
Hello everyone,
I think Gerda explained quite well what this is about. Actually the AI does not have any weight in the decision making at all, 0%, it's just extra feedback to our members to give you another way to evaluate and get more information about our photos. We are soon going to publish an article in the magazine which better explains how it works. It's being developed by 1x photographer and AI professor Yan Zhang and it's extremely sophisticated. Can an AI learn to evaluate feelings in a photo? Yes, most likely by analazying thousands or millions of other photos with strong or weak feelings. It's a good question though and a very interesting field of research. It's certainly one of the hardest applications for making an AI and therefore a very interesting challenge.
Anyway, 100% of all final decisions are still being made by the head curator team.
At the moment the AI can only handle landscape photos but we are right now training it for other categories as well. It's in a experimental early stage and will be improved a lot during the coming years. It's already likely to be one of the absolute most advanved curation AIs in the world though.
The AI rating is new to me; probably because I haven't posted a landscape shot in some time.
Given the mismatch between "live people" and AI why publish the results. It only serves to reduce the credibility of one or the other or possibly both. They should not have rolled this out publicly yet. A better approach would have been to gather data behind the scenes and work on getting better correlation between the "live people" and AI results. Once (if ever) there was a reasonable correlation than it could be rolled out. Going live with AI at this point is ludicrous.
Mike
Hi Mike,
I agree with your post. I will add that I do not think that the correlation between the two will ever be very high. The members make subjective decisions. The algorithm makes objective ones. The algorithm can only measure things that are measurable.
In my opinion, what they should do is have a scale or assign weights to member curation and the algorithm indicating how important each one is. The curation result would then be made up of those two things and passed on to the expert curators.
AL
Hi AL and all,
Which objective decision could AI possibly make with artistic work? I cannot figure out one criterion that would be applicable to all photos in one category.
Having to comply with AI in curation will never promote uniqueness or diversity and means moving one step further away from true creativity... at least for the ones that depend on curation.
For commercial purposes that might make sense. In Holland we call it 'eenheidsworst'... a unit sausage.
The fun-factor is lost on me now.
Wishing you a creative start of the week!
Edit: Having read Ralf's reply that was posted while I was writing; I'm very curious for the article and relieved that AI currently has no weight in curation.
Ralf, thanks for the explanation!
The AI rating is new to me; probably because I haven't posted a landscape shot in some time.
Given the mismatch between "live people" and AI why publish the results. It only serves to reduce the credibility of one or the other or possibly both. They should not have rolled this out publicly yet. A better approach would have been to gather data behind the scenes and work on getting better correlation between the "live people" and AI results. Once (if ever) there was a reasonable correlation than it could be rolled out. Going live with AI at this point is ludicrous.
Mike
Hi Mike,
I agree with your post. I will add that I do not think that the correlation between the two will ever be very high. The members make subjective decisions. The algorithm makes objective ones. The algorithm can only measure things that are measurable.
In my opinion, what they should do is have a scale or assign weights to member curation and the algorithm indicating how important each one is. The curation result would then be made up of those two things and passed on to the expert curators.
AL
Hi AL and all,
Which objective decision could AI possibly make with artistic work? I cannot figure out one criterion that would be applicable to all photos in one category.
Having to comply with AI in curation will never promote uniqueness or diversity and means moving one step further away from true creativity... at least for the ones that depend on curation.
For commercial purposes that might make sense. In Holland we call it 'eenheidsworst'... a unit sausage.
The fun-factor is lost on me now.
Wishing you a creative start of the week!
Edit: Having read Ralf's reply that was posted while I was writing; I'm very curious for the article and relieved that AI currently has no weight in curation.
Ralf, thanks for the explanation!
Hi Gerda,
AI (artificial intelligence) algorithm sounds mystical. It is just a mathematical formula or program that looks at data (pictures) and compares the information inside of one picture against many others.
When I said that it can measure objective things, I meant things like photo noise, composition, color, dynamic range, etc. it can measure these things absolutely or relative to other photos.
Ralf things AI can evaluate feelings in a photo, but feelings are not measurable by themselves. I am guessing that he means feelings about a photo can be inferred by looking at many similar photos and correlating the information with their score or result. But no one can be sure that what is being evaluated are feelings or trends.
AL
To me the AI additions are an interesting experiment in the world of photography...
I am quite curious what a trained AI program could do in comparison to humans, usually humans are overestimating their consistency and capabilities... it will take time but a lot of image traders are using these tools already...
I hadn't noticed it so far but look forward to what it can add...
before it gets an active role in the curation process I would assume there would be a public announcement about that change...
artificial memory, algorithms, mathematical calculations, etc. etc. Of course, innovative measures for evaluating the present and the future. We already have curators who "judge" without having published a photo and a profile, perhaps the best artificial memory? How would such a memory behave to “evaluate”, today, a masterpiece by Giotto, Picasso, or a photograph by Ansel Adams? All interesting, innovation is welcome. At this point, if an artificial memory is more reliable than an emotion, a moment, a feeling in looking at a photo, a street, a landscape, why do we focus on the soul of the shot? On the description? If artificial memory replaces emotion then that's fine, maybe we'd all be better off. Important, one day, not to complain that the machine created by man, annihilates man.
It's just my thought, not artificial!
The AI rating is new to me; probably because I haven't posted a landscape shot in some time.
Given the mismatch between "live people" and AI why publish the results. It only serves to reduce the credibility of one or the other or possibly both. They should not have rolled this out publicly yet. A better approach would have been to gather data behind the scenes and work on getting better correlation between the "live people" and AI results. Once (if ever) there was a reasonable correlation than it could be rolled out. Going live with AI at this point is ludicrous.
Mike
Hi Mike,
I agree with your post. I will add that I do not think that the correlation between the two will ever be very high. The members make subjective decisions. The algorithm makes objective ones. The algorithm can only measure things that are measurable.
In my opinion, what they should do is have a scale or assign weights to member curation and the algorithm indicating how important each one is. The curation result would then be made up of those two things and passed on to the expert curators.
AL
Hi AL and all,
Which objective decision could AI possibly make with artistic work? I cannot figure out one criterion that would be applicable to all photos in one category.
Having to comply with AI in curation will never promote uniqueness or diversity and means moving one step further away from true creativity... at least for the ones that depend on curation.
For commercial purposes that might make sense. In Holland we call it 'eenheidsworst'... a unit sausage.
The fun-factor is lost on me now.
Wishing you a creative start of the week!
Edit: Having read Ralf's reply that was posted while I was writing; I'm very curious for the article and relieved that AI currently has no weight in curation.
Ralf, thanks for the explanation!
Hi Gerda,
AI (artificial intelligence) algorithm sounds mystical. It is just a mathematical formula or program that looks at data (pictures) and compares the information inside of one picture against many others.
When I said that it can measure objective things, I meant things like photo noise, composition, color, dynamic range, etc. it can measure these things absolutely or relative to other photos.
Ralf things AI can evaluate feelings in a photo, but feelings are not measurable by themselves. I am guessing that he means feelings about a photo can be inferred by looking at many similar photos and correlating the information with their score or result. But no one can be sure that what is being evaluated are feelings or trends.
AL
AL, many thanks for your hints and thoughts, they help me to have a better understanding of what AI can do in this respect. The only thing I could come up with in my lack of imagination was for a photo to be in/out of focus, but that's not necessarily an objective parameter throughout all styles.
I tend to agree with Franco and his doomsday scenario if it comes to AI, but let's not condemn the idea before it has come to full bloom. 😊
Ciao!
artificial memory, algorithms, mathematical calculations, etc. etc. Of course, innovative measures for evaluating the present and the future. We already have curators who "judge" without having published a photo and a profile, perhaps the best artificial memory? How would such a memory behave to “evaluate”, today, a masterpiece by Giotto, Picasso, or a photograph by Ansel Adams? All interesting, innovation is welcome. At this point, if an artificial memory is more reliable than an emotion, a moment, a feeling in looking at a photo, a street, a landscape, why do we focus on the soul of the shot? On the description? If artificial memory replaces emotion then that's fine, maybe we'd all be better off. Important, one day, not to complain that the machine created by man, annihilates man.
It's just my thought, not artificial!
We will try our best not to let our AI take over the world and eradicate mankind!
Jokes aside, according to scientists an evil AI is actually the most likely armageddon scenario, Read more here.
This is no JOKE Ralf, AI is already aligined to smoe specific names for publishing their works, the others are all stepped away.
There is too unfair publishing judgments in publishing photos in the main galleries now. my last photo that is rejected in both galleries have been favorited by four high scoring members of 1x, and I see photos not even favorited being published. AI is stupid if not constsntly watched and checked.
This is no JOKE Ralf, AI is already aligined to smoe specific names for publishing their works, the others are all stepped away.
There is too unfair publishing judgments in publishing photos in the main galleries now. my last photo that is rejected in both galleries have been favorited by four high scoring members of 1x, and I see photos not even favorited being published. AI is stupid if not constsntly watched and checked.
I'm not sure if Saad is right about AI beeing biased towards some members, but it appears to me that it is; I have nothing concreete, just hints.
José Almeida
Hi all,
I don't know if you remember some months ago, when there has been proposals from the system for tags when uploading an image. Interestingly my upload was analyzed (I suppose by AI) and the tags FEELING,DARK,LOW KEY,MODEL,EMOTION,PERSON,LIGHT came as proposals. Among many others.
Please note: feeling, emotion
Apparently AI is not as stupid as many human beings think.
Interestingly the members (humans) criticized the technical quality of the image. By the way, the inferior technical quality is an often criticized point in my (sometimes blurry) images of models behind fashion store windows. My goal is definitely not to present crisp images.
I suspect that artificial intelligence can already deal with this topic far better than humans.
Maybe next week I will upload another (blurry) image. I'm curious what the members will say.
