pancakes

MicrostockGroup Sponsors


Author Topic: Dataset earnings - opt out  (Read 19069 times)

0 Members and 1 Guest are viewing this topic.

« Reply #50 on: October 03, 2024, 08:23 »
+1
Screenshot-2024-10-03-at-16-22-22" border="0


« Reply #51 on: October 03, 2024, 09:16 »
+1
1458 here. Not bad

« Reply #52 on: October 03, 2024, 10:34 »
+2
a) Re: hiding the initial "opt-out" - I suspect it was so the big "google/blackrock/etc backed" companies could quickly steal/etc as much without many people knowing. The reason the companies offered the "opt-out" AFTERWARDS... is to stifle potential competition. because say if google has access to say 100% of the "clips", and a competitor only has access to 30% (because of people opting out after the trickery was done) - then the "google/blackrock/vanguard" companies can make models that look more realistic (& hence more 'desirable' to use, aka more sales).

b) Odd. I opted out - and yet - apparently my stuff was still used. "Possible" within the small window of having opted out from pond5 - and when they 'announced' it "maybe" they sold some data to someone else... unlikely, but possible.

c) Companies that "license" the dataset should be paying in perpetuity (i.e., monthly, regular income) for ANY "data training", because obviously they want to make "perpetual revenue"- and by stealing/using other people's data with the current "one-time peanuts payment" - essentially you are making extremely low cost competition for yourself for peanuts. You should be compensated EVERY single time someone makes an image/video/etc based off ANY kind of model derived from your works.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #53 on: October 03, 2024, 13:15 »
+1

c) Companies that "license" the dataset should be paying in perpetuity (i.e., monthly, regular income) for ANY "data training", because obviously they want to make "perpetual revenue"- and by stealing/using other people's data with the current "one-time peanuts payment" - essentially you are making extremely low cost competition for yourself for peanuts. You should be compensated EVERY single time someone makes an image/video/etc based off ANY kind of model derived from your works.

You go to a mechanic and he replaces your rotors and brake pads. Then you should pay him a fee, monthly, every time you use the brakes on your car?  :o

The images are used one time, to create a model and that machine learning model, creates the new images. Your images, and mine, are never USED again.

When you license a stock image, the contract often says in perpetuity now. Same for a training use for AI. It has been licensed for use, forever, not for a single use. If you prefer that to a single use, which is what the actual use was, then fine. We still get nothing more than what the training has paid for.

Another complication, if there are 10,000 images of a pickle, used to train the software, then every time the software makes a pickle, 10,000 people should be paid again, for having their image used in training the machine? How do you know which images were used for every new image? There is no direct connection back to any specific image, when the machine creates a new image, from what it has learned.

I'd really like someone to answer that last question.

« Reply #54 on: October 03, 2024, 17:48 »
+1
Thanks for the heads-up. I checked it and got $87.26.

« Reply #55 on: October 03, 2024, 19:30 »
0
Quote
You go to a mechanic and he replaces your rotors and brake pads. Then you should pay him a fee, monthly, every time you use the brakes on your car?  :o

The images are used one time, to create a model and that machine learning model, creates the new images. Your images, and mine, are never USED again.

When you license a stock image, the contract often says in perpetuity now. Same for a training use for AI. It has been licensed for use, forever, not for a single use. If you prefer that to a single use, which is what the actual use was, then fine. We still get nothing more than what the training has paid for.

Another complication, if there are 10,000 images of a pickle, used to train the software, then every time the software makes a pickle, 10,000 people should be paid again, for having their image used in training the machine? How do you know which images were used for every new image? There is no direct connection back to any specific image, when the machine creates a new image, from what it has learned.

I'd really like someone to answer that last question.

You are comparing apples to oranges, actually more like apples to palm trees on a beach. Pretty much no relation whatsoever.

a) The mechanic didn't steal the rotors, pads, etc to sell them to you.
b) You aren't in the business of making rotors, pads, etc that someone else steals from you to sell forever in direct competition, and doing whatever they can to simultaneously try to prevent you from future sales while they make a lot of sales.

The people behind the "AI" (which is not a thinking machine, and it is for the most part a jewish/"identify"asjewish consoratorium of people) are engaged in a combination of massive theft, and then trickery/deception/manipulation/etc to try to get you 'out' of business for a teensy tiny payment - because they want a "in-perpetuity" model where people use THEIR stuff, NOT yours.

In other words - they are trying to trick you into selling your "asset" 1x (to them) for peanuts (i.e., $0.10) - so they can then resell it 10,000+ times without ANY compensation  to you (i.e., to them $100,000+ for your single asset).

Single use is one thing, because it is used by a single entity. Not for derivative works to be resold (i.e., you don't sell your license 1x so then the company that 'bought' it can sell it 10000 times).

Re: 10,000 - since this is a SOFTWARE program - VERY VERY VERY EXTREMELY easy to do via micropayments. And yes - there IS a connection. You tag & tokenize the models - and then that is how you get paid. I actually DO know how to do that. It is VERY easy to do. I guess you don't have the technical background/expertise to understand that - which is fine - I do. And I am telling you - it is VERY easy to implement - it is simply a matter of doing it. Of course - the people running the "ai" companies want you to think you "can't" - but that is a load of crap - they want you to think that because then it makes it easier for them to try and steal your stuff from you.
« Last Edit: October 03, 2024, 19:38 by SuperPhoto »

« Reply #56 on: October 04, 2024, 11:56 »
+1
I have been opted out since they offered that option. Yet I still received a dataset payment. It was as big as the $38 dollars I made all year on P5 with about 6k assets.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #57 on: October 04, 2024, 12:45 »
0

The people behind the "AI" (which is not a thinking machine, and it is for the most part a jewish/"identify"asjewish consoratorium of people) are engaged in a combination of massive theft, and then trickery/deception/manipulation/etc to try to get you 'out' of business for a teensy tiny payment - because they want a "in-perpetuity" model where people use THEIR stuff, NOT yours.


Really? No discussion from you, with attacks like that, are worth answering.

I'll just quote you:

And the psychos LOVE playing "the victim"  to make them appear above reproach. They are prolific liars, and when/if confronted - tend to double down on their lies to try and confuse the person challenging them - because most (normal) people are good people, and find it really difficult to believe there could be such psychotic sociopathic liars. Sadly though, they do exist - and a good portion of them pushed this total con (surprisingly) on most of the world.



« Reply #58 on: October 04, 2024, 15:58 »
0
Really? No discussion from you, with attacks like that, are worth answering.

I'll just quote you:

Lol. Huh? Your statements make no sense... What "attack" are you talking about, and on whom? Are you somehow implying that your lack knowledge on the subject matter means that anyone educating you on something you don't know, and/or contrary to your belief system is an 'attack' on you? And the 2nd quote - if you are referring to that same consoratorium - yes - that is correct - they do like playing the victim, and do indeed double down on lies... but that also for convid...

You keep repeating (different threads) that you believe (albeit incorrectly) users can't be compensated, and don't seem to like hearing anything to the contrary. Just because you 'believe' it doesn't mean its a fact. The fact is - users can actually be compensated, in perpetuity, for the usage of their images & videos in various models. Of course - if you read bloomberg all day (the same people working together with the 'ai' companies, and presenting the 'ai' stuff as "impartial news" when in fact its promoting their partners products & services) - then perhaps you might believe that. But that belief is all it is, a belief, and it is incorrect. Users actually CAN be compensated fairly, and it is a matter of integrating into the products. Do you not want photographers and videographers to be compensated fairly for their work? Are you against that?

If you'd like to clarify what you are trying to say, because again your response makes no sense. You should read up though on the people who own the companies, look into their backgroudns - as well as review the actual algorithms (not the bloomberg articles, but actual tech articles) on how the 'ai' actually works...
« Last Edit: October 04, 2024, 16:12 by SuperPhoto »

« Reply #59 on: October 04, 2024, 15:59 »
0
---


Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #60 on: October 09, 2024, 12:47 »
+1
---

Perfect quote, since I might as well be writing to an empty black mirror.

Speaking of ignoring and repeating? You didn't answer?

You
go to a mechanic and he replaces your rotors and brake pads. You pay for the parts and labor. Then you should pay him a fee, monthly, every time you use the brakes on your car?  :o

A parallel would be, the AI software PAYS to use your image, one time, and only one time, why do you expect to get paid, every time the software creates a new image, which is made without accessing or using your image again? They have paid for the parts and labor.

The images are used one time, to create a model and that machine learning model, creates the new images. Your images, and mine, are never USED again. Single Use.


 

Related Topics

  Subject / Started by Replies Last post
3 Replies
3489 Views
Last post March 25, 2021, 14:54
by Clair Voyant
1 Replies
4234 Views
Last post May 13, 2021, 09:51
by Stillsands
109 Replies
20054 Views
Last post October 11, 2023, 10:42
by pancaketom
1 Replies
5022 Views
Last post October 07, 2023, 11:17
by Uncle Pete
9 Replies
3534 Views
Last post October 10, 2023, 04:09
by Her Ugliness

Sponsors

Mega Bundle of 5,900+ Professional Lightroom Presets

Microstock Poll Results

Sponsors