|
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 - HalfFull
Pages: [1] 2 3 4 5 6 ... 23
1
« on: Yesterday at 08:37 »
1.337 / 5.000
How will misleading content in the Adobe library be removed? According to the examples given, everyone is knowledgeable about their own region.
I searched for "Istanbul" as artificial intelligence content in the Adobe library. I know very well because I live there. 85% of the images on the first page are misleading. How can an employee working at Adobe and living in another country know this?
Or how can I know whether the image about "Hamburg", which I have never been to, is misleading?
Actually, I wonder where exactly artificial intelligence images are used. Artificial intelligence images are definitely not used in magazines, newspapers, press and publishing websites in Turkey. It was used in the news of major newspapers a few times in the beginning. However, since the use of models that did not look like Turks in a news about Turkish Retirees was found quite ridiculous, the content of the news was changed immediately.
I am sure that the customer knows what kind of image they want. It is very good to separate Real Photo and AI in searches. But the selection tool could be in an easier place. It could be in the selection in the upper left corner.
(By the way, we can participate from Turkey, but Adobe is closed to Turkish customers. We can check our portfolio by changing the "tr" extension in the internet address bar to "fr". That's why the first page I see and the first pages you see may be different.)
The interesting question over AI is its potential impact on the stability and mental health of the general population. It's already been responsible for triggering civil disobedience via fake events generated accidentally and intentionally by local & foreign actors. The new battlegrounds of the future. We're only at the beginning of AI really so without some sort of safeguarding, control or regulation to protect jobs and the truth, the internet is at risk of ending up as a mass of information that no one believe anymore. A lot of the fact checkers we see today are driven by companies with political agendas that use it as a way to push their own version of the truth by bending it here and there. In other words, who is going to control and provide you with the truth in the future? For instance, if we look at the amount of AI imagery available at various agencies compared to real / photographic imagery, and, that they use the AI images as well as real ones when creating new datasets, the % of real imagery (correct data) will become less and less and potentially provide even less accurate results if things are not managed correctly. If the same happens with written facts online, people will start to doubt everything they see even if it was actually the truth. AI has so much potential to do good things, to make a positive difference in life but if we're not careful, it could also do a lot of damage. Imagery, while important to us, is only a small element of the whole AI thing and I worry that our current crop of politicians etc are not really up to the job. They see $$$ signs first and fail to grasp the long term damage it could cause if not handled correctly, probably because they know they won't be around when the preverbal hits the fan and that'll be for someone else to sort it out. For us older bods that only have a few years left to work it's probably not a major issue but for the younger generations that still have 20+ years left to work, they'll see jobs being wiped out with minimal replacements in return as businesses look to use the tech to slash more and more employees to improve their bottom as politicians try to tax them and the remaining employees more and more to make up for the shortfall in cash in the treasury coffers. Of course, we don't really need to worry about it... our current crop of global leaders are bound to have it in hand 🙃
2
« on: January 13, 2025, 09:26 »
Maybe try some ai just to see what it is all about. It is a tool like camera or photoshop and not better or worse.
The reputation of junk/ai comes from the fact that most ai on social media comes from amateur users, playing with free ai. it is just as bad as the junk mobile phone shots.
As for photos being "real"...
High quality stock images are anything but.
The people images are fake as hell, using specific lighting, make up, styling and then post processing to create fake fantasy people.
"Professional" food images are created with motor oil, styropor ice cream, all kinds paints and a ton of post processing in photoshop.
Nature? Created with fake lenseflare, the sky exchanged for another image (now done with gen ai in photoshop), all kinds ditsracting elements removed or other elements added in collage style content, then overfiltered and overprocessed.
Stock photo images are not real. They are often just as fake as ai. But not clearly labeled as photoshop fantasy creations.
Only editorial is real.
With my buyer hat I would just use what looks good.
If the people are fake ai or fake overfiltered, makeup, processed monster creations - who cares?
If it carries the concept I want to promote, that is fine with me.
For travel editorial I would use real images on location, but if it includes people, the location might be real, but the models might be fake photoshop creations.
Customers should have a very easy way to sort content for their use. And it should be sticky.
But it is not possible to say the fake photoshop make up overfiltered "traditional camera stock" is more real than what producers create with ai.
I've tried it... it was worse for me. I was wasting a stack of time muck ing around with stuff I could do by hand properly in less time. AI is not necessarily high quality, as you said, it's fast and fast + high quality rarely go together. Lighting is smooth but not necessarily real looking. It ends p too perfect and that's one of the many things that makes it stand out as AI.
3
« on: January 13, 2025, 08:16 »
Recent top sellers list has only 2 AI creators out of top 10. Previous week, 7 out of top 10 were AI creators. It was a bad week for AI creators since Adobe Stock made this change.
https://contributor.stock.adobe.com/en/insights/best/contributors?start_date=2025-01-06
But.the bestseller list supposed to be based on sales of files from the last 3 months?
So how could a search change implemented in the last few days affect the bestseller list in this way?
Unlessthere is something intentionally strange going on.
I always thought the bestseller list should be an editors choice.
This now looks like one.
Most recent top sellers list is based on the previous week's number of sales as far as I know.
I think the port size and % of sales to new files added recently may also be a factor. A lot of AI people are new contributors and as a result may well feature higher up in this. But...it's not really that important to be honest. I always pray to avoid those charts... and thankfully do and still rank from 75 - 800 through the year. However, Adobe have made a conscious effort to change away from AI being the default. They won't have done it just for the heck of it and it'll either be driven by customer behaviour data (how many are switching it off for each search etc) or, they're receiving complaints from clients that they can't find what they want for the massive volume of AI imagery that is clogging up the search results. Either way, that's all I need to know from a business point of view for future planning. As with most fads / trends, there comes a point when the client (and their customers) become tired of a certain look and AI certainly has a "Look" that is easily spotted. As was mentioned earlier, I believe AI imagery will stay around but will be used mostly by people who don't have a big budget or, by the masses who want an image to post in a forum, send to a friend or for a personal project. AI imagery in the broadsheets etc is now gaining a name as being junk, cheap and flooding the internet. It is the "Clip Art" of the future. As a designer or client, do you want to be seen using images that are viewed as cheap or fake by a large proportion of the general public who could also be your customers? Personally, I still haven't used AI because it doesn't help me. It's another complication that requires a lot of processing, amending and fiddling around with to get it to do what I want. I just find it quicker to do it myself and at least that way I get what I want. It's about getting the idea from my head to the screen / page. I have a feeling a lot of the AI contributors it's like pulling the arm of a Slot Machine and hoping something interesting might turn up. They'll get lucky once in a while but mostly they'll produce stuff that will rarely sell... if ever. Hence the masses of junk landing all over the internet.
4
« on: January 11, 2025, 15:15 »
Doesnt surprise me. I was reading an article in a broadsheet the other day about complaints that the internet was being buried by poor quality AI imagery. I could also imagine a lot of buyers will become fed up because, lets face it, while there are some good AI images there is also a lot of rubbish and it must be a PITA wading through all that just to find something good countless times a day (image buyers for advertising agencies etc).
5
« on: January 07, 2025, 11:58 »
"Should there be a requirement to label AI-generated content? Whats your take on this?"
My understanding is that there is a requirement. If you've used AI to generate elements used in the image it has to be declared as "AI was used to generate this image" tag.
There is an exception that relates to the use of filling expanded backgrounds in photoshop or, editing out parts that are distracting but, if you add a person, animal or element to the image it has to be declared as AI.
As you mentioned, there are a lot out there that don't. I now use "Content Credentials" from Raw File -> Photoshop -> finished Tiff / Jpg. This way I can prove the origins of my work and how it was created.
6
« on: January 07, 2025, 07:09 »
I pray to god that they use Shutterstocks submission process and not that PITA one iStock use. It's one reason I don't submit much too them, takes way too long.
7
« on: January 07, 2025, 06:36 »
Not great... I already have three Getty Accounts due to mergers of Corbis etc With any luck the regulators may say no but I won't hold my breath.
8
« on: December 13, 2024, 08:44 »
I've been busting a gut to get out a lot of my spring / easter work on the basis it would take at least 1 month or more to be reviewed and I wanted them up no later than mid January (all .png) .... two batches, one submitted about a week ago, maybe less. The next one two days ago. Imagine my surprise when I seen the notification confirming arrive this morning saying they'd all been approved 😳
Wow... it's like old times again, hope this continues 😀
9
« on: November 26, 2024, 03:50 »
Downloads up, $ up and running steady... ebb and flow as expected. Just less than 20k images... no AI. Total covering a 6 day period. Mon - Sat
Nice stats!
What's your return per download in $? <1 or >1?
I would guess ~ 0.75.
Thanks. I average between 0.87 - 0.94. As with most things, this will swing one way or the other but annually, that's what it is.
10
« on: November 25, 2024, 12:21 »
I like to check the top sellers list... not to see if I'm on it but to make sure I'm not. Thankfully, I manage to avoid that algorithm.
11
« on: November 25, 2024, 12:06 »
Downloads up, $ up and running steady... ebb and flow as expected. Just less than 20k images... no AI. Total covering a 6 day period. Mon - Sat
12
« on: September 26, 2024, 04:44 »
Thanks for all the support over the years and good luck in your new role!
13
« on: September 23, 2024, 05:40 »
Cheers  I mainly sell stills <18k images. I've been submitting to Adobe for about 10 years and elsewhere for 14yrs. Around 108k dls from 2020
14
« on: September 23, 2024, 04:29 »
I'm surprised this seems rather confusing for a lot of people? I believe I even read where it stated what the weekly sales rank is?
It's simply your rank, in terms of downloads - relatively to ALL the other contributors - for that week. Lifetime rank is simply for all contributors over the LIFETIME.
So, if you get more downloads than more people - your rank goes up. If other people get more downloads than you, your rank goes down.
So you 'could' be ranked #1, if for example you had '3' downloads, and 'everyone else' had 1. Or, you could be ranked #10000 if you had say 1000 downloads, but 'everyone else' had 1001+...
Really quite simple. Your weekly ranking is just simply how many downloads you've had relative to everyone else in the system.
of course,but why is there anyone who thinks the opposite? 
I think we all bounce between 2 values in the rankings.
if your ranking value is an average of 30,000 you can probably bounce between 40,000 and 20,000
if your average value is 20,000 you probably bounce between 15,000 and 25,000
if your average value is 1000 you probably bounce between 800 and 1200
the higher you go,the less variation you have in the rankings,simply because there are fewer contributors at the top.
I don't know whether the ranking follows sales or money,I have some doubts about it,it probably follows both the number of sales and money.
Weekly rankings will vary based on what seasonal content you have. Like the movement of the tide and over time, you see the patterns in the data. Lifetime rank for a change.
15
« on: September 23, 2024, 03:09 »
While Adobe have problems with their reviewing process in terms of speed of reviews and some errors in declining content incorrectly, I've always found they respond quickly when you contact them via the form about such problems.
The last two occasions I had to contact them was either on the Friday or Saturday and by Tuesday the following week at the latest they'd already replied and corrected the problem. While I wish these errors never occurred, mistakes do happen and hopefully they are working towards refining the system to prevent them and the lengthy queues... which at present aren't acceptable. But I've always found the Contributor Support team to be very helpful and responsive. They do a great job and no need for a petition.
16
« on: September 17, 2024, 11:17 »
Yeah, same here. Timing is a little off but no doubt one will be appearing before too long!
Edit... just arrived. Nice xxxx additional payment!!
17
« on: September 13, 2024, 04:30 »
Both client and contributor sites working here - UK
18
« on: September 12, 2024, 10:59 »
Dipped a little today from 395 to 419. Lifetime at 481.
Stats are for 3.5 days.
Nice!! I need to double my sales to get from 1,100 to 400. I wonder how many sales 100th rank has per week.
Cheers... Looking back at the stats, a rank of 75 = 1640 in a week and a rank of 150 or so is in the region of 1250
19
« on: September 12, 2024, 10:08 »
Dipped a little today from 395 to 419. Lifetime at 481.
Stats are for 3.5 days.
20
« on: September 03, 2024, 10:13 »
Starting to move back up from a low of 750... seems like the summer slow down is coming to an end!!
Is that the "beginning" of this week? (I.e., Monday/one days sales?) Just noticed when you posted it... If so, wow, good job!
I wish It was for six days. I grabbed it last week but didn't get time to post. Rank 548 as of today. Looking at previous years stats it takes until mid September to regain normal momentum again.
21
« on: September 03, 2024, 04:35 »
Cheers...I've been at it for quite some time now. Anything is possible but it is harder now than it has ever been due to the volume of new files being added. Makes it difficult for new work to be seen and to become established in terms of sales.
22
« on: September 03, 2024, 02:47 »
Starting to move back up from a low of 750... seems like the summer slow down is coming to an end!!
23
« on: August 26, 2024, 09:45 »
Hopefully this will get through. The server for this forum seems to be on its last legs. Taking minutes for pages to load etc and often returning server errors.
Anyway, I wanted to persist in trying to post this. As I mentioned above, I also had a file declined for because I'd failed to flag the image as "AI Generated." when the image wasn't AI generated at all.
I emailed Adobe on the 24th and they've already responded and confirmed the image is now live. Given the 24th was a Saturday, this is a very fast turnaround. Mistakes happen but it's the way they are dealt with that is important and I feel Adobe have been spot on here.
Anyone else experiencing this problem should just use the "Contact Us" webpage referral. You'll receive an email that you can reply to so that you can chase up or provide attachments etc.
24
« on: August 26, 2024, 06:52 »
don't panic!
just write them to explain & they'll usually let you re-submit - i've had this happen several times
The same thing has happened to me. I've emailed them with details and screenshots etc of the psb files. These things happen and provided the deal with it in a reasonable timescale then no real harm done. However, regarding, "Being allowed to resubmit". If the review process was only a few days... even a week, then I'd go with that but as the review process is now 1-2 months, being allowed to resubmit seems more like a punishment for something that was no fault of your own. Its hard enough to plan submissions for time critical events without having to factor in a possible "incorrect refusual appeal".... you would need to submit Christmas images in April / May just in case you need to appeal, resubmit the file and still have the image available for buyers to buy, use and print jobs etc for clients I think the sensible thing is to review the error.... and if it's deemed an error, approve the file so it can go on sale. It would also avoid the possibility of the resubmission being accidentally declined for the same reason!
25
« on: August 24, 2024, 05:21 »
really slow week here. dropped from 3900 to 4700 pos
hope you guys doing better than i xD
And next week, you go up, someone else goes down.
A waste of time to watch or care about position. Bragging rights? That and the numbers are unreliable. People with the same downloads, have different position. People with the same position have different numbers of downloads.
What determines position, if it's not just downloads as we've been told is how the position is created?
Only if you don't want to improve. I have a chart in my office that records best / worst weekly position for each month of current and prev year along with best ever for that month. This allows me to monitor performance swing when I submit new imagery. Bragging rights don't come into it as this data can only really be used when comparing your own performance from year to year. Tag it alongside monthly earnings and it becomes a powerful tool for your business.
Pages: [1] 2 3 4 5 6 ... 23
|
Sponsors
Microstock Poll Results
Sponsors
|
Show Posts
