|
MicrostockGroup Sponsors
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - cascoly
Pages: 1 2 3 4 5 [6] 7 8 9 10 11 ... 165
126
« on: January 03, 2024, 14:38 »
based on previous years it's too early for any announcement - last year's was late Feb(?) and credits in early March (?) - dates may be off by a week or so
i won a bet last year against those who thought the program would be canceled - anyone want to take the bet this year & buy me a few americanos?
127
« on: January 03, 2024, 14:34 »
Does anyone know how long it takes to get an answer from Shutterstock support?
Uploading on my account has been suspended because of AI generated content on December 16th. I started deleting and I have deleted all AI generated content from my portfolio in two days, but by the 18th my account has also been blocked and my portfolio just disappeared.
Since then I sent a few emails to support, no answer yet.
've been getting a lot of rejections as 'ai generated' when they're photos - support answered within 1-2 days & gave me a case # and almost were accepted on re-submit was that your situation or were you actually submittimng AI images?
128
« on: December 31, 2023, 19:09 »
...
Let's hope it's not the Daleks. Although in that case, I am sure the Doctor would help out yet again.
yet we have decades of video evidence of the dalek existence which make them much more likely than other ET to be out there. is there a Dr in the house/tardis?
129
« on: December 31, 2023, 14:13 »
Yes, that will be my workflow going forward. Upsize to the max to see flaws for editing, then downsize to 6-8 MP.
My minimum is 36 cent on Adobe and many sales that used to be 99 cents are now 1.11 or even 1.18.
So prices are going up and we are getting some relief for inflation which is very good.
Over time, when the quality of the ai engines improves, we will probably be able to offer larger files.
since inflation in many countries was about 10% for a short period, dropping to 3% in US this year & some others, for those countries the SS price would need to rise to $.11 ($, .40 and $1.09 on AS for your numbers) to keep up and then only sligtly more in the last year, so you're doing ok (if in the US)
130
« on: December 31, 2023, 14:05 »
...
What everyone else agrees about is, you will be selling all future rights to your images, that are in this sale, forever and you can't ever use or do anything with them.
...
depends on the terms of the contract - buyer could grant a license to use the images as you wish (except, of course, uploading to a stock agency since you no longer have copyright)
131
« on: December 30, 2023, 14:00 »
Having more files declined again that look quite fine to me....
perhaps that in combo with upsizing creates problems.
both can be improved with much smaller file sizes, so I will try to reprocess and make them smaller.
Hope I can find the problems and make it work. Having files declined is a waste of time for everyone.
ai Giga improves the image in addition to upsizing - i then downsize to 6mp (since these are selling for 10c i see no reason to provide larger) - no problem getting accepted by agencies
132
« on: December 29, 2023, 15:57 »
some further commentary on the case - turns out the copied text was not a random article but cherry-picked: AI #44: Copyright Confrontation
Zvi Mowshowitz newsletter The New York Times has thrown down the gauntlet, suing OpenAI and Microsoft for copyright infringement. Others are complaining about recreated images in the otherwise deeply awesome MidJourney v6.0. As is usually the case, the critics misunderstand the technology involved, complain about infringements that inflict no substantial damages, engineer many of the complaints being made and make cringeworthy accusations.
That does not, however, mean that The New York Times case is baseless. There are still very real copyright issues at the heart of Generative AI. This suit is a serious effort by top lawyers. It has strong legal merit. They are likely to win if the case is not settled.
In a handful of famous cases, there seems to be an exception. Exactly as in the MidJourney examples, why are we seeing NYT article text almost exactly (but not quite) copied anyway in some cases? Because it is iconic.
Kevin Bryan: NYT/OpenAI lawsuit completely misunderstands how LLMs work, and judges getting this wrong will do huge damage to AI. Basic point: LLMs DON'T "STORE" UNDERLYING TRAINING TEXT. It is impossible- the parameter size of GPT-3.5 or 4 is not enough to losslessly encode the training set.
Ok, now let's see NYT examples. Here GPT spits out almost perfectly the opening paragraphs of a "snow fall" article from 2012. But this text is all over the internet - super famous article! That's why GPT's posterior predictions given the previous article paragraph are so good.
Likewise, in the famous Guy Fieri Times Square review, GPT repeats almost perfectly whole paragraphs. But these paragraphs have also been repeated dozens of times across the internet! That's why the LLM posterior probability next word distribution picks them up.
In practice, one can think of this as ChatGPT committing copyright infringement if and only if everyone else is committing copyright infringement on that exact same passage, making it so often duplicated that it learned this is something people reproduce.
My take? OpenAI can't really defend this practice without some heavy changes to the instructions and a whole lot of litigating about how the tech works. It will be smarter to settle than fight
.
>>> bold text my emphasis -- all caps in original much more detail in the newsletter: For free subscription: Don't Worry About the Vase | Substack https://thezvi.substack.com/?utm_source=substack&utm_medium=emailA world made of gears. Doing both speed premium short term updates and long term world model building. Currently focused on weekly AI updates. Explorations include AI, policy, rationality, medicine and fertility, education and games.
By Zvi Mowshowitz
133
« on: December 27, 2023, 16:04 »
you beat me to, it - i was about to, post this, too -- looks like a major case that may lead to a legal conclusion though I distrust the ability of knowledge-deficient judges to really understand the issues involved (much like the copyright office's decision i've been reading Wolfram(of Mathematica fame) on how ChatGPT works (available a on Kindle Unlimited for free) https://www.amazon.com/What-ChatGPT-Doing-Does-Work-ebook/dp/B0BY59PT5Z it gets complicated quickly and he explains why there is no connection between training and generation since the process from source to dataset is not commutative. like MJ et al., the data used for training is massive (even NYTimes huge content is dwarfed by several orders of magnitude). it should come down whether the scraping amounts to fair use. however, there are some significant differences from AI-gen images as the Timmes alleges wholesale reproduction of significant quantities of text by chatGPT - something no one ah as been able to show re AI-image generation.
134
« on: December 26, 2023, 13:46 »
...
In fact we all really need a more detailed rejection reason for images, AI images, Videos.
"quality issues - Thanks for giving us the chance to consider your image. Unfortunately, this image doesn't meet our quality standards so we cant accept it into our collection.Common issues that can impact the technical quality of images include exposure issues, soft focus, excessive filtering or artifacts/noise."
Current "quality issues" explanation gives absolutely no chance to understand, fix it and improve. Is it so difficult to make multiple select checkbox list for reviewers to select the exact rejection reason? It will really help both Adobe and contributors.
Great ideas & pix - my take is AS may (unjustifiably) consider them too 'soft' even tho photorealistic. as far as quality reasons, we've been asking for YEARS to get real comments about quality, otherwise it's just a guessing game - even alamy with its ridiculous rejected a batch when they find 1 error, has changed its previous policy & now at, least identifies the culprit and gives a better reason (tho w/o changing it's all or none policy) and even much maligned SS has always given detailed reasons for rejection (and responded when rejection reasons are challenged. eg, i get a lot of 'ai generated content' for photographs & get a case # to resubmit -- a pain, but most are accepted on 2nd try.
135
« on: December 26, 2023, 13:33 »
... I never entered metadata for each image, and now I have no way to find a specific image. Does anyone know a way to do a reverse image search, but instead of searching the internet, I want to search my hard drive? Thx.
i keep an excel sheet with my filenames, captions & tags - so i can find any image very quickly. [ my archives contain over 250k images including original s, intermediate processing, etc but my spreadsheets only need to record about 10% of that] details at: https://cascoly-images.com/building-a-microstock-tracking-system-part-1curious how you were able to sbmit without metadata?
136
« on: December 26, 2023, 13:22 »
...
Short sighted if they expect AI training to go on forever. Once used, it's not needed. Oh wait, that's our side, one use or two and done.
New training, maybe, and that's why everything rejected, they tried to slot over into Data Catalog so they could sell it for AI, without putting it up on the standard site. WE feed the data beast that's eating our income?
they don't necessarily sell training, they sell already trained datasets others can use - i'd expect these would be subsets of the main one - eg, medical, architectural, etc
as far as pumping the data library, i've found most 'eligible' are n not my best images (those are approved), but 'not quite' similars that would otherwise rejected. it's not in their best interest to inflate the training set as they'll lose sales from those items, versus the small income per individual image among millions.
Ah, and then we get our .03 when they license that specialized set? Hey when is the next exciting and much anticipated, Contributor Fund report coming? If they are concentrating on data sales and sets, where's our slice of the pie?
I opted out of the data library, as all I saw what everything rejected, went into that. I wasn't going to count on the pennies adding up from those rejects. I mean, not good enough to sell, but good enough to license a data?
well, you couild look at it as the farmer poisoning their own well
137
« on: December 25, 2023, 19:00 »
At the beginning of October, Adobe Stock was adding around 900k genAI images per week.
From October 16th for the next three weeks, it was just over 1 million each week.
From Nov 6-13, it was almost 2 million!! 1,930,975.
...
for comparison, what were the numbers for non-AI? that would help to see how much AI is affecting the total
138
« on: December 25, 2023, 18:57 »
...
At least no one will be fighting over the drumstick on Thanksgiving 
but they should give credit to John Madden who invented the multi-drumstick turkey for their NFL Thanksgiving broadcasts
139
« on: December 25, 2023, 18:28 »
... their upload limit.
...
Agencies needs to take action to control these stuffs to promote more authentic work and real artists.
Sometimes I feel there is a huge requirement for a Microstock Organization and all RF and RM companies should come under them. Through this they can control pricing and authenticity both. Moreover can also organize a annual microstock meet to reward and also motivate people to work for such field.
luckily most of us don't live in a country that would be able to dictate what private companies must do. -- who would control pricing? -- how would you force membership? -- how would this work when agencies are global? -- who'd decide what is acceptable or authentic?
140
« on: December 25, 2023, 18:19 »
Producing 100 images with AI takes just two or three hours?
do you mean to generate unfinished images,just prompts?yes,you can do even more in 2-3 hours,but the work is just started.
I just sent 10 AI images to Adobe and it took me 2 days!
I generated them with AI yes,but then I completely renovated,improved,added particular effects...it takes forever!
photographers and videographers will always be necessary,AI changes the game but cannot replace real content.
...
totally agree - those who keep citing mass quantities of similar images (as if it's new challenge to 'real' artists) seem to have forgotten the thousands of silly images that have been approved over the years ( tomato slices, )
141
« on: December 25, 2023, 18:13 »
...
Short sighted if they expect AI training to go on forever. Once used, it's not needed. Oh wait, that's our side, one use or two and done.
New training, maybe, and that's why everything rejected, they tried to slot over into Data Catalog so they could sell it for AI, without putting it up on the standard site. WE feed the data beast that's eating our income?
they don't necessarily sell training, they sell already trained datasets others can use - i'd expect these would be subsets of the main one - eg, medical, architectural, etc as far as pumping the data library, i've found most 'eligible' are n not my best images (those are approved), but 'not quite' similars that would otherwise rejected. it's not in their best interest to inflate the training set as they'll lose sales from those items, versus the small income per individual image among millions.
142
« on: December 23, 2023, 20:25 »
i had a bump from a special payout in april - otherwise Pond5 sales were about 30% over previous year - but hardly worth further contribution ass - i've left my portfolio sit & not uploaded since early 2022
143
« on: December 22, 2023, 16:07 »
Sounds like even more good content from experienced producers will be going near exclusively to Adobe.
They change nothing and keep winning.
no evidence that would happen - wishful thinking by anti-SS folk. SS wasn't harmed by those who stopped uploading and many of us who stayed report RPD for AS & SS hasn't changed much, while AS sales continue to be significantly less than SS (in my case 50-70% less) some of this is likely due to AS rejecting many images that sell on SS combined with AS rejecting most editorial images. (and batch rejects by AS haven't helped)
144
« on: December 19, 2023, 13:41 »
To me this topic is useless in the forum if there is no discussion on the particular problem with the problematic examples, it looks like personal problem between contributor and Adobe, but not professional. I can't believe in this, by my experience Adobe do it's job professional. But maybe there is some private case and better contact Mat or Adobe directly, if you don't want to share your images, something I understand...
the point is, individual images aren't relevant when entire batches from multiple established artists get rejected. and by posting here we discover it's not an isolated instance (or reviewer) but an ongoing major problem that AS refuses to recognize
145
« on: December 18, 2023, 20:05 »
Must of been created by AI. Hence the 10 extra fingers steps.
AI would have done a better job -- just ask chatgpt to do your re-design unfortunately, this is a modern curse - inexperienced designers who don't bother to do a needs analysis first, interviewing users to see what's actually needed; they just add what they think is an improvement, then dont bother to conduct an alpha test much less a beta test - they just ejaculate whaledreck into the wild without ever consulting the users we see this with ALL the agencies spewing new versions. [rant] it's not just here - it' s everywhere and much more annoying in more consequential apps -eg, not realizing that a backspace is a character, so when someone makes a typo, it invalidates the entry. there are simple routines to strip unwanted characters, including 'special characters and otherwise cleanup & validate the users entries but these newbie coders are too lazy to add an extra line of code and seem to have even fewer supervisors who understand software development in the real world. These are boring, non-sexy elements of software, unlikely to even be covered in academic courses [/rant]
146
« on: December 18, 2023, 13:40 »
next response from canva 'support':
We'll forward your feedback to our team about the review process. We improved our tools which caused efficient review turnaround time. This is because of the backlog we encountered in the last few months.
rejecting everything will certainly reduce the backlog!
147
« on: December 16, 2023, 13:14 »
I will try to send the images as editorial, they don't have any visible logo anyway. It makes no sense that the same images have been accepted as both editorial and commercial on Adobestock.
different reviewers, changing standards over the years
148
« on: December 16, 2023, 13:13 »
...
I see overall people thinking on quality only on technical aspect, but to me in one stock photo quality means also what value it brings to the collection and from there the quality of the Adobe Stock collection as a whole. Every image is like a small pixel who create the whole collection image....
perhaps, but we can't tell what the reason is when the rejection just says 'quality' - rejection because it doesn't fit is the old very subjective "low commercial value". While annoying (since no one really kows what will sell), a separate rejection for LCV at least tells us there's nothing we can do. otherwise its's just a big guessing game
149
« on: December 15, 2023, 14:19 »
mass rejects have been reported for some time now in previous threads
150
« on: December 12, 2023, 14:19 »
I'd asked how images could be rejected seconds after being uploaded - here's 'Ivy's response
The Canva Content team aims to build a diverse and expansive library where users would be able to find the most amazing and relevant images they can use for their every designing need. While we appreciate the artistic mind and tireless efforts our contributors put in every image they submit to us, we need to ensure they pass our quality guidelines.
We've reviewed your most recent image submission and found that most of them do not adhere to the Canva image quality standards. Based on overall artistic appeal, usability and technical quality, we are unable to accept them into the library.
You may refer to our Contributor Quality Guidelines for Photos and Graphics for your reference.
I'm now closing this for now, but feel free to reach out should you have another question or concern. We'd be more than happy to help!
We appreciate your understanding.
of course they didnt answer my basic question
in addition, i pointed out those guidelines didnt contain anything relevant to image contributors
Pages: 1 2 3 4 5 [6] 7 8 9 10 11 ... 165
|
Sponsors
Microstock Poll Results
Sponsors
|
Show Posts
