As promised here are the full answers Michael Howard from foto was kind enough to answer.
"Q. Do you see the AI contributors as a viable part of the paying supporting community?
A. I’m not sure yet. We are watching to see how Foto is used by everyone from all walks of life and seeing who supports Foto. Building a social media network centered around photography primarily, but images of other types are allowed starts by collecting data, observing, and then making adjustments. I don’t think there is a huge AI community on Foto. There are a few that I know of, but I do not think 50% of the Foto users are AI contributors. From what I can tell, it’s a tiny %.
Q. Is it possible that the AI image makers might eventually outnumber traditional photographers?
A. With how fast Adobe, Google, Apple, and Canvas are incorporating AI, it’s possible, but we don’t really know. No one knows. It’s all speculation. AI images could be just a fad like the NFT craze. To a certain degree, the market will sort itself out, but it will take time. I do think companies will start using AI more, and there isn’t much anyone can do about it. The genie is out of the bottle at this point, and billions are being poured into making AI better every year. It sucks, but that’s what is going on.
Q. Do you plan to allow the photographs posted in the feed to be used to train AI models?
A. No. We are not allowing photos posted to Foto to train AI. If we ever did something like that, it would be something every user would have to opt-in to and we’d need to pay contributors for allowing their photos to be used as training data. Perhaps this is a way photographers can make money and sustain themselves in the future, but again it’s too early to tell.
Can you protect the existing contributing photographers from having their content used in AI mashups?
Q. How might you clearly define original work from Images generated on the databases of other people’s work?
A. This is again something we are discussing, and we’ll talk about it in depth on the Foto podcast episodes about AI. There are some options available now, but we need to build a firm plan around them. We can do something with metadata and third-party software systems that are learning how to identify AI images.
Q. How do you plan to maintain the authenticity and artistic integrity of the community?
A. This ties into the question above. It will be done through metadata, third-party software systems, labeling images, and allowing Foto users to choose what they see. Current social media doesn’t give users control. We want to do the opposite and give users the ability to craft their own experience, but to do that, we need accurate labeling of images and accurate search. This will take years to perfect, if not decades.
Q. If ever you enable photographers to sell their work do you think original photography will be worth as much alongside fantastical AI images?
A. I don’t know. You are asking me to be a futurist and predict the consumer/art buyer market in the future. I can’t predict the future."
Michael also says he's going to record several podcasts with a few different people to discuss AI and the challenges it presents. That will most likely be available on https://fotoapp.substack.com
Enjoyed reading this one, I understand AI is here to stay in one for or another… I see it as a tool that helps facilitate how you might edit better in regards to photography. My hand is nowhere near steady enough to mask out to the enth degree when separating her from the background, but an AI tool could and it would get better overtime. I do not think AI should be used as a tool for generating images to make a story because it takes work away from the people who have trained long and hard to do this work.
I think if you are having to separate someone from a background after the shot there's an issue with the DOF, lens, or lighting. ;-) Doesn't removing or taking the background away in an image make it more digital art than photograph? You took a photo but turned it into digital art.
Thanks for going into the Foto AI quagmire. As you know I'm an early adopter there too, and I share your justified dismay, but I sympathise with them too. When is AI OK (and even when is AI not AI?) is rapidly becoming the biggest question of our times, and there is no 'good' answer to what should and shouldn't be allowed to be shown... I think the guys at Foto are trying with the best kind of intentions, but they're Canute in front of the waves aren't they? I have no answers to this but with AI in general, I suspect we're going to have to make some binary choices when the answers aren't binary at all.
Thanks for the comment Tim. I was up till 1am attempting to catch up with the comments on the foto app as since my conversation with him Michael has dropped a post asking the community their thoughts on AI. You could imagine how busy that particular thread is. I can’t find my reply to a reply now, but I was mentioning that foto was on the lookout for defined binary solution (which is understandable as he’s a coder) but there was a lot of grey area in this dilemma. I like the idea of an authentic tag and also that the community can flag questionable content. It could be the content with a certain number of flags is filtered by those questioning authenticity. Obviously, this can be abused so a ranking system for moderators would probably be needed. I’m sure this problem has been solved elsewhere.
As promised here are the full answers Michael Howard from foto was kind enough to answer.
"Q. Do you see the AI contributors as a viable part of the paying supporting community?
A. I’m not sure yet. We are watching to see how Foto is used by everyone from all walks of life and seeing who supports Foto. Building a social media network centered around photography primarily, but images of other types are allowed starts by collecting data, observing, and then making adjustments. I don’t think there is a huge AI community on Foto. There are a few that I know of, but I do not think 50% of the Foto users are AI contributors. From what I can tell, it’s a tiny %.
Q. Is it possible that the AI image makers might eventually outnumber traditional photographers?
A. With how fast Adobe, Google, Apple, and Canvas are incorporating AI, it’s possible, but we don’t really know. No one knows. It’s all speculation. AI images could be just a fad like the NFT craze. To a certain degree, the market will sort itself out, but it will take time. I do think companies will start using AI more, and there isn’t much anyone can do about it. The genie is out of the bottle at this point, and billions are being poured into making AI better every year. It sucks, but that’s what is going on.
Q. Do you plan to allow the photographs posted in the feed to be used to train AI models?
A. No. We are not allowing photos posted to Foto to train AI. If we ever did something like that, it would be something every user would have to opt-in to and we’d need to pay contributors for allowing their photos to be used as training data. Perhaps this is a way photographers can make money and sustain themselves in the future, but again it’s too early to tell.
Can you protect the existing contributing photographers from having their content used in AI mashups?
Q. How might you clearly define original work from Images generated on the databases of other people’s work?
A. This is again something we are discussing, and we’ll talk about it in depth on the Foto podcast episodes about AI. There are some options available now, but we need to build a firm plan around them. We can do something with metadata and third-party software systems that are learning how to identify AI images.
Q. How do you plan to maintain the authenticity and artistic integrity of the community?
A. This ties into the question above. It will be done through metadata, third-party software systems, labeling images, and allowing Foto users to choose what they see. Current social media doesn’t give users control. We want to do the opposite and give users the ability to craft their own experience, but to do that, we need accurate labeling of images and accurate search. This will take years to perfect, if not decades.
Q. If ever you enable photographers to sell their work do you think original photography will be worth as much alongside fantastical AI images?
A. I don’t know. You are asking me to be a futurist and predict the consumer/art buyer market in the future. I can’t predict the future."
Michael also says he's going to record several podcasts with a few different people to discuss AI and the challenges it presents. That will most likely be available on https://fotoapp.substack.com
Enjoyed reading this one, I understand AI is here to stay in one for or another… I see it as a tool that helps facilitate how you might edit better in regards to photography. My hand is nowhere near steady enough to mask out to the enth degree when separating her from the background, but an AI tool could and it would get better overtime. I do not think AI should be used as a tool for generating images to make a story because it takes work away from the people who have trained long and hard to do this work.
I think if you are having to separate someone from a background after the shot there's an issue with the DOF, lens, or lighting. ;-) Doesn't removing or taking the background away in an image make it more digital art than photograph? You took a photo but turned it into digital art.
If I’m doing hair, it’s more if I’m just cleaning up stray hair and things. Or just lifting the shadow slightly to reveal more detail
Thanks for going into the Foto AI quagmire. As you know I'm an early adopter there too, and I share your justified dismay, but I sympathise with them too. When is AI OK (and even when is AI not AI?) is rapidly becoming the biggest question of our times, and there is no 'good' answer to what should and shouldn't be allowed to be shown... I think the guys at Foto are trying with the best kind of intentions, but they're Canute in front of the waves aren't they? I have no answers to this but with AI in general, I suspect we're going to have to make some binary choices when the answers aren't binary at all.
Thanks for the comment Tim. I was up till 1am attempting to catch up with the comments on the foto app as since my conversation with him Michael has dropped a post asking the community their thoughts on AI. You could imagine how busy that particular thread is. I can’t find my reply to a reply now, but I was mentioning that foto was on the lookout for defined binary solution (which is understandable as he’s a coder) but there was a lot of grey area in this dilemma. I like the idea of an authentic tag and also that the community can flag questionable content. It could be the content with a certain number of flags is filtered by those questioning authenticity. Obviously, this can be abused so a ranking system for moderators would probably be needed. I’m sure this problem has been solved elsewhere.
Meanwhile over on Threads...
https://www.threads.net/@meta/post/DA_mM0Ey1qC?xmt=AQGzG91FnsHN3rFaY1Dfsk0ULDG51-jNR6-BAOZ-W7cPBQ