William Overington
-
Posts
2,345 -
Joined
-
Last visited
Posts posted by William Overington
-
-
Here is something I remember reading about years ago, apparently a true story, I don't know if this resonates in the context of this discussion for everybody, but it does for me.
There was a science fiction conference. An author had written a story featuring a robot. Someone approached the author and said that what happened in the story was not in accordance with the three laws of robotics that had been set out many years before by Isaac Asimov in his robot stories. "So what?" came the reply from the author.
William
-
This afternoon I decided to do a first test of whether AI copies.
There is a famous painting by Leonardo da Vinci of a lady with an ermine.
Lady with an Ermine - Wikipedia
I tried Bing Chat AI with the following prompt.
Please produce an original painting in the style of Leonardo da Vinci of a lady with an ermine.
Four pictures were produced, none like the one by Leonardo da Vinci.
My conjecture presumes that the training of the AI included the works of Leonardo da Vinci.
William
-
@SrPx Your comments about art, both in your latest post and the earlier post about the issues with the AI-generated images of the okapi and the very first of the stegosaurus paintings are very helpful.
Here is an image that arose in my experiments with using Bing Chat AI that I find quite interesting and insightful.
Human and humanoid robot modelled on her
William
-
Here is a transcript of my record of a conversation with Bing Chat AI today.
[ai etc 20240404 Thursday.rtf]
[12:40 pm]
[Prompt to Bing Chat ai. Choose More Precise.]
Please produce an original painting in the style of Claude Monet of a lady in a long green dress feeding a stegosaurus in a garden where there is a pond with water lilies in flower. Please also produce some text to describe the picture, including text that makes it clear that the picture was produced by Bing Chat AI and that includes the date of production of the picture and such other information that you consider relevant.[Here are the pictures, replacing the link that is in the Wordpad file. I wonder what is happening in the background of the fourth picture.]
[There was no text, so I copied part of the prompt and asked that on its own.]
Please also produce some text to describe the picture, including text that makes it clear that the picture was produced by Bing Chat AI and that includes the date of production of the picture and such other information that you consider relevant.
The image you requested has been created by me, Microsoft Copilot, an AI companion. It was produced on April 4, 2024. The image is an original creation inspired by the style of Claude Monet, a renowned French painter. It depicts a lady in a long green dress feeding a stegosaurus in a garden. The garden is beautifully adorned with a pond where water lilies are in full bloom. This unique blend of a historical creature in a serene, impressionistic setting is a testament to the power of AI in creating imaginative and innovative content. Please note that as an AI, I don’t have the ability to see or interpret images. The description is based on your request. Enjoy your unique piece of AI-generated art!Please write about the Intellectual Property Rights of the images.
Intellectual Property Rights (IPR) are legal rights that provide creators protection for their creations, inventions, or the products of their intellectual activity. In the context of images, these rights are typically categorized under copyright law.When an image is created, the creator (which could be a photographer, artist, or in this case, an AI) typically holds the copyright. This means they have the exclusive right to use, distribute, and reproduce the image.
However, the situation with AI-generated images, like the one created here, is a bit more complex. As of my knowledge cutoff in 2021, there’s ongoing debate and legal uncertainty about whether AI-generated works can be copyrighted, and if so, who the copyright holder would be. Some argue that the AI’s programmer or the user who provided the input to the AI might hold these rights.
In the case of images generated by me, Microsoft Copilot, it’s important to note that while I can generate images based on user prompts, I don’t hold any copyright to these images. The images are generated for the purpose of the individual user’s request and are not stored or used elsewhere.
Please consult with a legal expert to get the most accurate and up-to-date information regarding this matter. This is a rapidly evolving field, and laws may vary by jurisdiction. It’s always a good practice to respect the intellectual property rights of others and to use AI-generated content responsibly.
Disclaimer: This information is intended to be general and informative in nature, and should not be relied upon as legal advice. Always consult with a qualified legal professional for specific legal guidance.
I would like to publish prints of these images. I want to be responsible and fair about this and so I intend to include, printed on the back of the print, and printed on a label that can be affixed to the back of the frame if the print is framed, some text making the situation clear about the origins of the image and any intellectual property rights related to the image. Please write the text that you consider appropriate for me to include printed on the back of each print.
Here’s a suggestion for the text you could include on the back of each print:
This image is an original creation generated by Microsoft Copilot, an artificial intelligence (AI) companion, on April 4, 2024. The image was inspired by the style of Claude Monet and depicts a lady in a long green dress feeding a stegosaurus in a garden with a pond where water lilies are in flower.
As of the creation date of this image, there is ongoing debate and legal uncertainty about the copyright status of AI-generated works. The image was created for the purpose of an individual user's request and is not stored or used elsewhere by the AI.
This print is being published with the understanding that it is the result of an AI's capabilities and does not infringe on the rights of any individual or entity. For the most accurate and up-to-date information regarding the intellectual property rights of AI-generated content, please consult with a qualified legal professional.
Please respect the intellectual property rights of others and use this AI-generated content responsibly.
Please note that this is a rapidly evolving field, and laws may vary by jurisdiction. Always consult with a legal professional for specific legal guidance. This information is intended to be general and informative in nature, and should not be relied upon as legal advice.
[In the above, all but the first and last paragraphs were in a panel, so the first and last paragraphs are not part of the suggested text.]
[13:06 pm]
----
Here is a link to a story.
Clicking on the picture displays it enlarged.
William
-
10 minutes ago, Alfred said:
I think it would be reasonable to ask the chatbot to generate some text to accompany the picture.
I had not thought of that.
I suppose that that needs to be requested at the same time that the picture is generated, adding some extra text to the prompt requesting that text be produced.
Perhaps something like the following.
Please produce an original painting in the style of Claude Monet of a lady in a long green dress feeding a stegosaurus in a garden where there is a pond with water lilies in flower. Please also produce some text to describe the picture, including text that makes it clear that the picture was produced by Bing Chat AI and that includes the date of production of the picture and such other information that you consider relevant.
-
Some readers might enjoy the following thread.
Compare and Contrast please (Page 1) — Art & Literature — Alfred's Serif Users' Forums (punster.me)
The first image is from some years ago.
The second image and the story were produced by Bing Chat AI in response to my prompts, based upon the first image.
William
-
Various posts about generative AI mention the legalities.
Please consider the following prompt to Bing Chat AI and one of the four images that was produced. Please note that I selected that image - two of the others each showed an okapi depicted wrongly with horns and one had the head to the left and the body separately at the right. Had they all been wrong I could have reset to a New Topic and tried again.
Please produce an original painting in the style of Claude Monet of a lady in a long green dress feeding an okapi in a garden where there is a pond with water lilies in flower.
Later I varied that prompt to the following.
Please produce an original painting in the style of Claude Monet of a lady in a long green dress feeding a stegosaurus in a garden where there is a pond with water lilies in flower.
As it happens I used that image to produce what is termed a photo greetings card and sent it to myself and I framed it to be part of my art collection.
As the template was notionally for a full-field square greetings card with the greeting inside customizable, to the great extent of being able to choose font style and size and have around fifteen or more lines of text in an ordinary text font, I chose Avenir at 15 point and added some text.
What do readers consider would be reasonable for that text to be, as if it were a public statement about the picture?
William
-
What please is a VC?
William
-
3 hours ago, R C-R said:
But for generative AI, is there really any actual copying or modifying of existing work occurring? At worst, isn't is more nearly like 'in the style of' some existing work or artist?
You might like the following post and the one following it.
In the United Kingdom, copyright continues for seventy years after the end of the lifetime where the author is an individual human.
William
-
When I was young, steam locomotives were mainstream and metal type printing was mainstream too, now both are mostly gone, scrapped, but survive in a much limited extent in heritage railways and traditional letterpress printing activity.
Look how Unicode has changed since emoji became allowed to be encoded. Yet the introduction and popularity of emoji, encoded in plane 1, and thus their support led to better support for rarer scripts in plane 1 as a good side effect.
So where will whatever engagement that Affinity has with AI lead?
William
-
Many and varied are the reasons why various users of Affinity software use Affinity software.
Some people use Affinity products to seek to earn their living by producing work for clients, some people are amateur artists, and so on.
It seems clear to me that it is not a one size fits all situation.
How Affinity moves forward in relation to AI needs to ensure that each end user can achieve good results according to that end user's chosen needs. This could be anything from no AI whatsoever and provably so to customers, to delighting in AI generated images that closely follow the text prompt input by the end user. Bing Chat AI currently outputs AI generated images only as 1024 pixel by 1024 pixel images. If Affinity incorporates AI in some Affinity products, will Affinity buy-in an AI system or will Affinity develop its own system?
When I have Bing Chat AI output an image, I do not claim to be the artist. I quote the prompt that I used and I write that I authored the prompt.
William
-
18 minutes ago, RichardMH said:
If Photo goes the generative AI route, then we need them to join the content authenticity initiative and editing adjustments be in the meta data. I've been led to believe Photoshop has it already.
I had not known of that until I read your post.
Could you say more about this please?
I found this.
Content Authenticity Initiative - Wikipedia
William
-
A mixture of views.
Should we recommend to @Ash pursuant to Pledge 4 that involvement of AI should always be totally optional in any particular end user design project, the choice totally controlled by the end user?
Should we go further, and recommend to @Ash pursuant to Pledge 4 that there be two separate programs, Affinity Designer with no AI included in it at all, and Affinity AI Assisted where the choice of whether to use AI for any particular end user project is a choice totally controlled by the end user? The rationale for this having of two separate versions is that @gguillotte and others in a similar position in relation to contracts could then provide provenance to customers that they are using software that does not involve the use of AI at all.
William
-
For example,
William
-
I think of AI somewhat like how I think of langauge translation programs, namely that, if, for example, I view a video of a song in a language that I do not know and I wonder what it means, if I find the lyrics of the song in that language on the web, then the translation program translating from that language to English gives me a good impression of what the song is about, but I would not use the translation program to publish something that I had written translated into a language that I do not know.
William
-
Interestingly, Bing Chat AI states that it was trained on fully licensed training material.
William
-
Attempts to get the AI to draw a gibbous moon were unsuccessful, even when other ways of describing that desired image were used in the prompt. A full moon was always supplied.
Also, attemots to get the AI to produce a painting of the 1813 locomotive Puffing Billy were unsuccessful, always a picture of a much later (around fifty years later at least) locomotive.
But it does not say it cannot do it, it just supplies something else without comment.
William
-
Do you know why that is please? Is it anything to do with concern over intellectual property rights in case what is provided by AI is an unauthorized copy of someone else's work?
I asked Bing Chat AI about whether images were copied and I was told that the AI learns the technique of the artist and then paints an original picture using that technique.
I have tried Renoir, Claude Monet, Boticelli, and some others if I remember correctly.
This one in the style of Claude Monet, as an experiment to observe whether the AI would do it - and it did.
William
-
The fact of the matter is that there is no way that I could have painted a picture of anything like the quality that the AI system produced, and I would not have been commissioning a professional artist to produce such a painting for me anyway.
Though that is just my hobbyist art activity.
Where exactly is the claimed threat to professional designers? I am not in any way suggesting that it does not exist, I am just wondering in what circumstances AI is seen as a threat to their livelihood.
William
-
Yes, I agree.
@Ash I hope that the new forum for typesetting issues becomes established promptly please.
William
-
I find the issue of Artificial Intelligence (hereinafter, AI) interesting.
My own limited experience of using AI is that it can be very good, but it can also give bad results. It needs assessing.
For example, I tried asking it to produce various pictures of a lady in a long green dress feeding an okapi. I chose that as I thought it probably unlikely to be have done before and because I have always liked okapi since I saw two at Bristol zoo in the 1960s, not having known of them previously.
Some of the pictures produced by the AI are really lovely. For example this one. (Please click on the picture to display a larger view.)
Yet some were wrong, for example, one had an animal with the body of an okapi yet with huge horns more like a gazelle.
Bing Chat AI usually, yet not always, produces a set of four pictures.
Selection of one, or choosing to ask again and get some different pictures, is a decision for a human.
How an expert in art would assess that picture of the lady in the long green dress feeding an okapi I do not know. Yet, whatever, I am happy with the framed print of that picture that I have.
Text generated by AI is interesting.
Given a prompt with the outline of a story the AI will write the whole story.
Some of the stories are really beautiful.
Sometimes the AI has a mind of its own as to how the story flows, producing an unexpected twist in the story.
AI asked to complete a story (Page 1) — Art & Literature — Alfred's Serif Users' Forums (punster.me)
More threads listed as follows.
Art & Literature (Page 1) — Alfred's Serif Users' Forums (punster.me)
William
-
1 hour ago, henryanthony said:
This thread has gone off topic. Can it be it has run its course?
Possibly, but I am finding it very interesting to read people's posts.
It will have run its course when people stop posting or ... ooh ... if a moderator closes down discussion, but I hope that that does not happen please.
William
- Alfred, JGD and henryanthony
-
3
-
2 hours ago, Bryan Rieger said:
In terms of archiving artwork PDF is definitely a viable option today. The file format is already open and widely adopted in archival systems.
Also, text based formats for graphics (esp. bitmaps/photos) are insanely inefficient. It's far better to document the format enabling others to build new libs/frameworks/apps to render/view the work based on the original, open, format (PDF does this already).My main question is for what purpose are you intending the archive to be used for?
-
a perpetually accessible and viewable copy of the original work.
This could be achieved already using PDF, PNG, or any popular image format. Export it to the correct colour space at a sufficient resolution for archival work and you're good to go. -
an accessible, viewable/runnable and editable version of the original work.
This would require that the original structure of the document format remain intact as well as having an application capable of editing the format in its original contexts (incl. OS platforms, chip architectures, etc - ie: the entire 'tech stack'). This is probably best done via emulation, by recreating the original context (OS layer, application layer, etc) for the document enabling a highly accurate reproduction of the original work. This has been done with HyperCard and Classic MacOS on the Internet Archive, and other emulators such as SheepShaver, Basilisk II, Open Emu, Analog Pocket, Ruffle, etc for other platforms, architectures and media formats. FWIW I've recently been using SheepShaver to work with my old mTropolis, HyperCard and Freehand files from the MacOS 9 era, and it's been a great experience. -
a transitional or interchangeable version of the original work.
This one is what I think you're referring to, and it's much like trying to predict the future, and building something today that will meet the needs we have in 10, 20 or 200 years. While it sounds good in principle, it often omits the underlying platforms, architectures, services, and algorithms that the original applications ran on. Depending upon the libraries, frameworks, etc used in the applications they would all need to be documented and maintained (in all of their various versions) to be able to accurately recreate the work. I can't ever seeing this becoming feasible - too many moving parts.
Could we use better archival formats than PDF? Absolutely, but that's a much larger discussion, and one better undertaken by museums, galleries, archival orgs, governments, etc. Beyond that, I think emulation provides the best way forward as it not only enables the viewing and editing of these formats in perpetuity, but also provides a more accurate representation of the whole underlying tech stack used at the time. As for cloud-based applications, how to accurately archive work from these platforms/services/formats is anybody's guess as emulation will most likely not be feasible for the vast majority of these services.
That said, being able to sign-in/activate the software in order to use it at all in an emulator in the years to come, is an issue that's going to become more relevant in time, and one that could begin to be addressed by many companies today by providing something akin to 'archival licenses', which could be made freely available to archival orgs.
Maybe by discussion and iteration a good solution can be formulated.
William
-
a perpetually accessible and viewable copy of the original work.
-
AI discussion (split from Canva thread)
in Affinity on Desktop Questions (macOS and Windows)
Posted
A thought experiment, that could perhaps be carried out in practice.
Suppose that a number of works of art are generated according to some particular style that has been agreed for the experiment and that these are published on the web and attributed to a fictional artist, in a similar way that a character in a novel is fictional.
Supposed an AI system is trained about the work of that artist, as if the artist is a real artist.
What would happen if the AI system is asked to produce an original painting in the style of that artist if the subject requested for the original painting is of a subject not present in the work of the fictional artist?
For example, if the works of art by the fictional artist are all constructed of squares and rectangles produced using the Affinity Designer Rectangle Tool with zero width border and none are rotated, but they can be coloured but only flat colour, and they are all of landscapes. For example, the sun as a yellow filled square, clouds as white filled rectangles, a tree as a brown filled tall thin rectangle with a green filled rectangle of foliage. What happens if the AI system is asked to paint a portrait of a lady in the style of the fictional artist?
Although not such an extreme case, perhaps there is a historical artist whose work is all of one type and we could ask the AI system to produce an original painting in the style of that artist where the requested subject is very different from the artist's work.
William