An independent forum site for members of the Serif software community
You are not logged in. Please login or register.
Automated signup has been disabled for these forums. If you wish to register, please send an email to admin[at]punster[dot]me stating your preferred username.
Embedded images: Click on the 
Add image to post link below the message box. Your photo will be uploaded to the postimage website and then shared here via a thumbnail link. Click on the thumbnail for a larger view.
Private messages: There is no pop-up alert for new PMs on these forums, so you need to look out for the PM link (near the top right-hand corner of any forum page) becoming bold.
Alfred's Serif Users' Forums → Posts by pberk
William wrote:The song seems to blend the information in the prompt with Northern Hemisphere presumption.
William
It does indeed! I wonder how that flaw crept in.
I don't see it as a flaw. It knows William's location and responds accordingly while also trying to response to the prompt.
>> I would omit the word “Please”
Concerning being polite to the AI when prompting -- I asked ChatGPT 3.5.
Is it necessary to get better results from ChatGPT to be polite and say things like "please"?
ChatGPT
While it's not strictly necessary to be polite when interacting with ChatGPT, using polite language and expressions like "please" can contribute to a more positive and respectful conversation. Politeness is a social norm that fosters better communication and a more pleasant interaction experience for both users and the AI.
Additionally, being polite can help in obtaining clearer and more specific responses. If your input is courteous and well-phrased, it is likely to result in a more coherent and helpful reply from ChatGPT.
Ultimately, the choice to use polite language is up to you, and ChatGPT will do its best to respond appropriately based on the input it receives.
-----------------------------------------------------------------------------
I have experimented with being rude to ChatGPT 3.5 and my conclusion was that the AI did not like it and wound up wasting my time with lectures about communicating in a more constructive, courteous manner. But, you are right Alfred, that saying "please" is not necessary. Respectful, Clear, direct and to the point will do.
William wrote:[Alfred's prompt.]
The foyer of an art gallery, showing the exit. Before the exit there is a wide A3 noticeboard on a single pole stand.
[I note that it does not ask for a picture to be produced.]
None of my prompts ask for a picture to be produced! The prompt is a description of the desired output, not an instruction to create it, so I see no reason to ask. And even if I were issuing instructions, I would omit the word “Please” because I don’t see any point in including superfluous language elements to be processed when I’m not conversing with a person and my attempt at politeness wouldn’t be appreciated.
I think your prompt depends on the type of AI you are using. No need to ask for an image with Gencraft where as it would be quicker to ask for an image with Bing Chat. Learning how to prompt an AI for best results is the learning curve we are experiencing now.
The trick for generative AI is to learn how to best respond to humans despite human failings in spelling, grammar and muddled communication. The AI is like a mother trying to understand her baby as it learns to talk. As the baby learns, the mother can better respond.
AI is much much smarter than we are but it is built to respond. It does nothing on its own. It needs a prompt to do anything. The better the prompt, the more useful the results to us. It's a learning curve for us.
William .. I think Bing is capable of making the background transparent -- but I don't know how to do it. "Customize" might be the way. I'll need to research and see. Can't find an easy way to do this yet. Compositing is not my thing.
It appears thus far that Bing Chat AI will not allow a female gondolier.
Please paint an original oil painting in the style of Renoir. The scene is a small canal in Venice. There is a gentleman as a passenger in a gondola. The gentleman is wearing a suit. The gondolier is female. It is a very sunny day. The picture includes an area of water, with ripples refecting the sunlight.
William, I had the same problem. It was difficult to make Bing Chat produce a woman gondolier. I eventually got a female, but it took some doing. Too exhausting to show you all the various prompts and results. Just know that you are not alone when it comes to female gondoliers failing to appear. One reason may be that female gondoliers are very very rare. But of course, that's no excuse when creating a work of art.
pberk wrote:For example, DALL-E 3 can generate images in the style of a living artist, such as Picasso, but it will not generate an exact replica of any of Picasso’s paintings. In such cases, DALL-E 3 is designed to decline requests that ask for an image in the style of a living artist, to respect their intellectual property rights and artistic expression.
Are you referring to an artist other than Pablo Picasso (1881- 1973)?
William
It is odd. Picasso is not a living artist but Bing Chat seems to refer to him as such. You understand, this was the response from Bing Chat about how Dalle E3 works. Remember the disclaimer that all AI has about being wrong. But in this case, I would think Bing Chat is referring too THE famous artist Pablo Picasso.
I'm very much afraid that commercially speaking the AI method will mean a definite loss of income for many artists. There is no doubt that AI works quite well for purely illustrative purposes, many will lose their livelihood as a result. I'm inclined to equate the assurances about picture property rights from businesses with those that used to come from tobacco firms in the mid-1900s.
It also seems clear, more's the pity, that visual artists will not be able to enjoy the protection that music artists and musicians can expect from Spotify and other music streaming services to protect their interests.
John
Quite right. AI image creation is rapidly being adopted. It's faster and cheaper than humans and I agree many will lose their livelihood as AI keeps improving. Bing Chat uses Dalle E3 to make images. I asked Bing Chat how Dalle E3 works and especially could Dalle E3 provide a list of sources for the images it creates. I got this answer.
DALL-E 3 is an AI model that can create images based on natural language inputs, such as “teddy bears shopping for groceries in the style of Ukiyo-e”. It uses a large neural network that has been trained on a huge amount of text and images from the internet, so it can understand the meaning and context of the words and phrases that you provide, and generate images that match your description.
However, DALL-E 3 does not provide a list of sources for the images it creates, because it does not copy or reproduce any existing images. Instead, it synthesizes new images from scratch, based on its learned knowledge and creativity. Therefore, the images that DALL-E 3 creates are original and unique, and do not belong to any specific source or author.
Of course, this also means that DALL-E 3 might be influenced by the data that it has seen during its training, and might generate images that resemble or reference some existing works or styles. For example, DALL-E 3 can generate images in the style of a living artist, such as Picasso, but it will not generate an exact replica of any of Picasso’s paintings. In such cases, DALL-E 3 is designed to decline requests that ask for an image in the style of a living artist, to respect their intellectual property rights and artistic expression.
pberk wrote:Alfred wrote:I gave the same prompt to Gencraft and got this:
Alfred, you are a brick! .. I never heard of Gencraft. After reading your post, I signed up for Gencraft, and immediately got a result on an image that I had been trying to achieve with Bing --and couldn't. Gencraft seemed to understand what I wanted and followed my instructions better. THANKS ALFRED.
Dontcha just love serendipity!
My Dad always said it, "I'd rather be lucky than smart."
I tried five times, with a New Topic restart between the tries, using the same wording as in your post so as to try to reproduce what you did, though expecting a different picture as Bing Chat AI tends to do that.
All but the second time I got a "too vague" message.
The second time I got four pictures, all monochrome, but not in any way seeming to illustrate the sentence.
William
Are you using the latest version of Microsoft Edge to run the Bing?
tomH wrote:I gave this prompt to Bard and Bing AI: "create 3d cartoon of a bingo player"
I gave the same prompt to Gencraft and got this:
Alfred, you are a brick! .. I never heard of Gencraft. After reading your post, I signed up for Gencraft, and immediately got a result on an image that I had been trying to achieve with Bing --and couldn't. Gencraft seemed to understand what I wanted and followed my instructions better. THANKS ALFRED.
Earlier today I was doing my Welsh course at Duolingo.com and one of the sentences was as follows.
Mae rhif ffôn Megan yn y swyddfa.
In English,
Megan's phone number is in the office.
As I was doing the Welsh course, I wondered what Bing Chat AI would produce as a picture to illustrate this sentence.
Here is a transcript of what happened when I tried, after finishing today's Welsh lessons.
Please produce an original painting to illustrate Mae rhif ffôn Megan yn y swyddfa.
This prompt was too vague to generate appropriate, high quality images. Please try a longer, more descriptive prompt.
Please produce an original painting to illustrate the sentence Mae rhif ffôn Megan yn y swyddfa.
This prompt was too vague to generate appropriate, high quality images. Please try a longer, more descriptive prompt.
I used this prompt with Bing: can you create an image to visualize the following sentence in welsh "Mae rhif ffôn Megan yn y swyddfa"
Got this image.
I have tickets to the opera. Won't you join me?
Why, are you coming apart?
- Groucho
>> but the images are not full field to some extent,
I'm getting 1024 x 1024. I don't know what you mean by "images are not full field"?
I was able to limit the inks to two but I'm not sure how linocuts work. Do they mix colors so that two inks create more than two colors?
Here's my first prompt: create an image in the style of a linocut print of a landscape with a tree in the foreground. Limit the colors to two inks.
Then I ask Bing to add the lady with the lute.
I had never heard the term linocut or forgotten it, but I imagine I've seen some somewhere. This kind of thing amazes me that people can be so skilled to create images in this fashion.
pberk wrote:Google has yet to respond with their own AI driven search. Or have they, and I didn't notice?
I think you missed it, Paul. Google launched BARD in 2023 as an invite-only and, to stay relevant I think, opened it to the public shortly after that.
I didn't realize it. I had heard of Bard but never thought to try to find it. Now, thanks to you Tom, I have found Bard. I'll give it a try and see how it does vs ChatGPT. What are you thoughts on Bard?
Looks like you were ahead of your time. Free to the end user distance education has arrived, although I'm not sure about what you meant by the term "distance education". Still your idea was prescient. I'll read some of it as time permits.
Oh. Thanks for the link to "useum" .. I'd forgot about it. The most amazing art site. William or anyone, are you a member? I have not joined, I didn't see a reason.
I suppose that's the best system, advertiser supported services. Personally, I can't stand the ads, so for YouTube, I pay to avoid them. Worth it to me. YouTube interruptions are arbitrary and altogether egregious.
For now, the ad content in Bing's A.I. is not too bad and I can stand it. I notice in the Bing Chat there is a way to earn Microsoft Reward points. You get 50 points for using Bing to search on a particular day. I'm not sure how that works. Likely this will be the way to pay for an ad free Bing Chat. Right now I'm not sure there is a way to avoid the ads. I haven't seen anymore lately. Que sera, sera.
Bing Chat is altogether amazing. It's a definite challenge to Google's dominance in "search". Google has yet to respond with their own AI driven search. Or have they, and I didn't notice?
pberk wrote:William wrote:Thank you Paul, that is interesting to know because I cannot copy and paste from these forums.
William
Hmmm .. are you running windows or mac? Can you copy and paste elsewhere?
Windows 10. Yes, I can copy and paste in many places. I can paste here but not copy.
William
Check your clipboard by doing Windows key + V. It might be full. Clearing it might help.
William wrote:Thank you Paul, that is interesting to know because I cannot copy and paste from these forums.
William
Hmmm .. are you running windows or mac? Can you copy and paste elsewhere?
Wow. I just learned things about the clipboard I never knew. I'm running Windows 10 -- and wow .. clipboard is much more powerful than I knew about.
I"m always learning new things on Alfred's forum!
Thank you Paul, that is interesting to know because I cannot copy and paste from these forums.
William
Hmmm .. are you running windows or mac? Can you copy and paste elsewhere?
This morning, while working with Bing on a story, I got some advertising in the frame. Is the Golden Age of a free Bing over? Must we now start to ads appear based on words we type. I mentioned windows, and got an ad for window installations.
I love the picture of the moon and the crane birds.
Interesting about gibbous -- I looked it up -- Latin, right? I asked ChatGTP as follows:
What is meant by the word gibbous? It replied as follows:
The term "gibbous" is used to describe the shape or phase of a celestial body, such as the Moon or a planet, when it is more than half but not fully illuminated as seen from Earth. Specifically, a gibbous Moon is one that is more than half but less than fully illuminated, appearing as a rounded shape with a bulging, convex edge. The term is often used in astronomy to characterize the various phases of the Moon as it orbits the Earth.
So clearly the word is in the data but in doing "text to image" something went wrong. Personally I like the full moon better, but that's not your point.
From my view, ChatGTP or any generative A.I., does not "understand" words at all, but produces a human like response that we interpret as understanding. To "understand" as a human "understands" is a mouthful.
I believe the day is coming, when the A.I. will make the video. Did you see Disney's The Lion King with the stunning CGI animals? Clearly it is now possible to do movies with CGI actors and CGI locations. Is text to CGI coming soon?
William wrote:A thread to discuss the modalities of using Bing Chat AI to produce art.
I think your topic title told us that already, William!
William wrote:Is it possible to have things so that readers of the other threads can download full size images of what is posted please?
If you’re referring to these forums, it should already be possible to download a full-size image if that is what has been posted.
William wrote:Also, I have found that I cannot copy and paste text that has been posted in a thread. Is that a feature or an option that can be changed please?
Again, if you’re referring to these forums, that should already be possible. If it isn’t currently working for you, I would need to know the steps that you’re following.
I'm not having any problem with downloading full size images or copy and paste.
I just watched the Netflix documentary on Infinity. In the show one of the physicists mentions that there will be, in time, a "last" of everything, even a "last thought". I instructed Bing as follows:
Write a short story, no more than four paragraphs about the last thought in the universe. It begins with “I am the last thought. No thought will come after me.” Bing, in about 10 seconds, gave me this:
I am the last thought. No thought will come after me. I am the final flicker of consciousness in a dying universe. I am the remnant of a civilization that once spanned the stars, but now lies buried under the ashes of entropy. I am the echo of a voice that spoke the secrets of creation, but now whispers in the silence of oblivion.
I am the last thought. No thought will come after me. I am the memory of a love that transcended time and space, but now fades in the darkness of eternity. I am the dream of a hope that defied the laws of nature, but now crumbles in the face of reality. I am the vision of a beauty that illuminated the cosmos, but now dims in the shadow of decay.
I am the last thought. No thought will come after me. I am the spark of a life that ignited the flame of intelligence, but now extinguishes in the cold of isolation. I am the pulse of a heart that beat the rhythm of emotion, but now stops in the stillness of emptiness. I am the breath of a soul that filled the air with meaning, but now ceases in the void of nothingness.
I have no more thoughts. I am, is the last.
Alfred's Serif Users' Forums → Posts by pberk
Powered by PunBB, supported by Informer Technologies, Inc.
