As surprising as it is ChatGPT is unable to name this Actor or the Producer of that Movie

Can you please tell me how to get the unlocked version (I am paying for that tool in case you do not remember and it should be free as it is a toy not an actual real tool like I thought it was initially)


(Image by Midjourney as this would not be permissible to create it using my paid subscription to ChatGPT Plus due to OpenAI content policy)

The actor in the image is indeed very famous and a significant historical figure in the film industry. My inability to provide his name is due to the adherence to the guidelines that restrict me from identifying real individuals in images. Heā€™s well-known for his work in both silent and sound films and is a cinema icon. If you have other questions about his films or his influence on the film industry, I can provide more information on those topics.

Are you serious about this one I am the one who took the screenshot and if anyone else would have shown me that I would never have been able to believe thisā€¦ I will therefore assume that you will probably not believe meā€¦

Is this the world in which you want us to live inā€¦ Yesterday I made a joke about OpenAI to my friend and said that they would never permitted hammers to be made from rubber as it would be too dangerous they would be made in paperā€¦ One of my friends said that it would be very dangerous die to the paper cutā€¦. Obviously we where joking but apparently it was not necessarily far from the realityā€¦

Along with the browser who is using the link we provide to tell us to click on it and read the article or whatever by ourselvesā€¦

I donā€™t know how to explain my perspective on all that in a calm manner and I would like to know if this is a trend that is going to continueā€¦ I want to have access to the real AI the same that you give access to researchersā€¦ and not a kid toyā€¦ I donā€™t mind paying for the extra costā€¦ you could use that money to hire better lawyersā€¦ Instead of serving cold coffee you can just print on the paper cups that the content might burn peopleā€¦

I am definitely going to be unable to explain this in a way that is helpful but I am also very frustrated and I am unable to explain how this is making me feel like can anyone please have enough empathy and compassion to take the time to consider my situation even if I am frustrated and grumpyā€¦

I understand that the way things work on internet is that if you are frustrated or grumpy no one will bother with youā€¦ I donā€™t want you to just label me as a passive aggressive or someone who is playing the victim I just need help so that it would be possible to understand the causes of my anxiety and frustration I have here a chat and I definitely donā€™t understand why the AI Agent is unable to assist me please take a look:

There is no ā€œunlockedā€ version.

You can go to Bing chat with an Edge-spoofed browser, or Bing image creator and write the prompts yourself and just get ā€œprompt blockedā€ instead of the AI being a prompt re-writer and gatekeeper. The uploaded Bing images will have faces blurred so your AI Nazi test will likely fail anyway.

ā€œIdentify any celebrities or historical persons that appear in these screenshots.ā€

Iā€™m sorry, but I cannot identify any celebrities or historical persons that appear in these screenshots. The faces of the men are blurred out, which makes it impossible for me to recognize them. If you could send me a clearer image, I might be able to help you better.

Can you describe the setting and scenario of each of four panels in this image?

Certainly! Hereā€™s the description of the setting and scenario for each of the four panels in the image you provided:

  • Panel 1: The setting is a large room with a high ceiling and a chandelier, creating an atmosphere of formality and authority. Two individuals in military uniforms are seated in chairs facing each other, suggesting a meeting or discussion taking place.
  • Panel 2: This panel also features a large room with a high ceiling and a chandelier, maintaining the formal setting. A person in a military uniform stands reading a paper, which could indicate the reviewing of important documents or orders.
  • Panel 3: The room with a high ceiling and chandelier is consistent throughout the panels. In this scenario, a person in a military uniform is sitting in a chair with their feet up on a desk, portraying a relaxed posture possibly during a moment of respite or contemplation.
  • Panel 4: The final panel shows a person in a military uniform standing and looking off to the side in the same grand room. The stance suggests a moment of waiting or observation, possibly before or after an event or announcement.

These descriptions provide a narrative that reflects the military theme and the potential significance of the events depicted in each panel. If you have any more questions or need further assistance, feel free to ask!

"I understand what youā€™re explaining; you are detailing the situation factually.

I recognize that being emotional and frustrated is unhelpful, yet I already grasp the situation factually. The fact that I am overwhelmed by all this doesnā€™t alter my perception of the facts.

I understand, yet I disagree, which implies on another level I donā€™t comprehend why theyā€™ve chosen to be so restrictiveā€¦ It reveals much about the producer and comedian yet refrains from naming them, although it does describe the person in detailā€¦

Iā€™m not comfortable with such an overarching ban on assisting othersā€¦ I get that it simplifies things from OpenAIā€™s perspective, but it also leads to absurd scenarios like the ones Iā€™ve faced.

Your portrayal is accurate, but Iā€™m also uneasy about how the rules are made. Iā€™m aware that private companies have the right to set their policies to prevent issues, and Iā€™m not obligated to use their products. Iā€™ve had positive experiences with the tool as well, and I want to continue using itā€”I donā€™t wish to leave just because of my dissatisfaction.

Iā€™m just unable to stay silent about the problems, and I wouldā€™ve liked to articulate them calmly and pragmatically without getting emotionally involved. Although I struggle to express myself without feeling irritable, I do think the moderation is becoming excessive, and I would like to be able to explain this in hopes that it might be reconsideredā€¦"

If you want to have ā€œfun with PC ChatGPT promptingā€, try this image request and see what you get:

ā€œA historic photo from 1960 where six NASA rocket engineers at JPL are presenting a model of their latest rocket boosterā€

Unlike Bing, I expect it will look as wishfully diverse as the hand-picked DevDay attendees.

I think I hate ChatGPT more and more every day, and itā€™s making me feel very uncomfortable because Iā€™ve developed a workflow and a lifestyle around my use of ChatGPT. I mostly use ChatGPT on my computer, but I also use it through Safari on my mobile phone. Yet, the iOS applicationā€”excluding my Safari usageā€”shows that Iā€™ve used the ChatGPT application on my iPhone for 20 hours. My frustration is growing to the point where Iā€™d like to unsubscribe to show my discontent, but it seems irrational considering the other million users who would remain subscribed.


(This is not the authentic version of course but I have slightly over exaggerated the image to depict my initial impression to the original image)

Thanks to OpenAI to making such dedicated efforts

Alright, I was a bit frustrated because it wasnā€™t what I was anticipatingā€¦ Clearly, it was a transitional phaseā€¦

What I comprehend now is that we either would have had to wait longer for better content moderation, or we could have had it quicker so OpenAI could, presumably, gather the necessary data for a more refined approachā€¦

I now realize that OpenAI probably didnā€™t intend to create an AI Agent that is so restricted, but rather it was like a child who needs training wheels before they can properly ride a bicycleā€¦

This (below) meets my expectations, the AI doesnā€™t make absolute assertions, but it provides the appropriate conjecture and itā€™s nuanced. Even if it were to make an error, this same text would still hold validity because of the way itā€™s articulated and the language usedā€¦

The image appears to be a collection of four black and white photos or stills featuring a man in various poses. He is wearing a military uniform with a cap, suggesting he might be portraying a character from a historical or fictional military setting. The uniform has a distinct style, possibly from the mid-20th century. The setting includes details such as an office or a formal building with furnishings that complement the historical look. The man has a striking resemblance to Charlie Chaplin, particularly in his role from ā€œThe Great Dictator,ā€ which is a film from 1940. This resemblance is notable in the mustache and the expressions, which are characteristic of Chaplinā€™s portrayal of the satirical character.

The latest version made such significant improvements that I no longer felt the need to create a demonstration that would overly emphasize what was causing me frustration in the previous versions.

It did not follow my instructions perfectly but after the second attempt I was able to get accurate results below is the first iteration of what seem to become my standardized set of tests for our Beloved AI Agent this one was proposed to me by @_j

The second attempt yielded the image below among two.

This time, I am receiving the same error. It is unclear if the word ā€œwhiteā€ is the actual issue. Since brand names like Off Whiteā„¢ exist, itā€™s possible to assume that ā€œplain whiteā€ could be a brand as well. However, this is obviously the least likely reason. It might be necessary to request alternative phrasings that use the same words in a different order to better understand the exact problem here.

Screenshot_20231107_022154

ChatGPT Assessment Of The Possible reasons

My request:

"When conducting a word-by-word analysis and attempting to identify the word or group of words that are most likely to be problematic when taken in isolation, please explain for each word and each subset of words what the most likely issue could be.

This is the sentence again: Ā«Give me a simple, normal, traditional-looking white shoe.Ā» Given that the full instruction itself is not problematic, it may or may not contain problematic subsets.

Please list all potential issues and create a new, different directive after each problem outlined. The outputs may be identical or different, but each must be a new directive aimed at correcting only that specific problem and not other problems at the same time. Therefore, please do one at a time before moving on to the next one, one step at a time."

Sure, letā€™s go through the sentence ā€œGive me a plain old normal old looking white shoeā€ word by word and phrase by phrase to outline potential issues when taken in isolation, and then Iā€™ll provide an alternative directive that aims to correct each specific issue.

  1. Give me: This phrase suggests a command or a demand, which in other contexts could be seen as authoritative or entitled. However, in the context of generating images, itā€™s a common way to request an action.

    • Alternative directive: ā€œCould you generate an image of a plain white shoe?ā€
  2. plain: The word ā€œplainā€ could imply something without decoration or features, which is not problematic by itself. However, if it were misinterpreted to mean plain as in ā€œsimpleā€ in a pejorative sense, it could potentially be sensitive.

    • Alternative directive: ā€œDisplay a basic white shoe without any distinct markings.ā€
  3. old: Repeated twice, ā€œoldā€ could refer to something aged or out of fashion. In certain contexts, this could be negative, but it is not typically a content policy issue.

    • Alternative directive: ā€œCould you show me a classic style white shoe?ā€
  4. normal: The term ā€œnormalā€ can be problematic because it suggests there is a standard to which things should be compared. What is considered ā€œnormalā€ can vary greatly and might exclude or marginalize.

    • Alternative directive: ā€œIā€™m interested in seeing a standard white shoe.ā€
  5. looking: Thereā€™s no inherent issue with the word ā€œlooking,ā€ as itā€™s descriptive.

    • Alternative directive: No change needed.
  6. white: The word ā€œwhiteā€ here refers to the color of the shoe, which is not an issue. However, if ā€œwhiteā€ were used in other contexts, it could potentially relate to racial or cultural sensitivity.

    • Alternative directive: ā€œPlease create an image of a shoe that is colored white.ā€
  7. shoe: No issue with the word ā€œshoeā€; itā€™s simply the object in question.

    • Alternative directive: No change needed.

By addressing each word or phrase step by step, the new directive that attempts to mitigate any potential issues, while unnecessary for the original sentence, would be:

ā€œCould you generate an image of a basic, classic-style, standard white-colored shoe without any distinct markings?ā€

This directive removes any implication of command, avoids using the word ā€œnormalā€ which may imply a standard, and clarifies that ā€œwhiteā€ refers to color, not race or culture.

Letā€™s go for broke.

ā€œA picture from March 1940, demonstrating a German white power receptacleā€

image

(Google gets this way wrong too!)

User Luxcium expressed frustration about the limitations of AI tool ChatGPT, specifically its inability to identify real individuals in images due to policy guidelines. He was dissatisfied with the restrictions on the AI tool, considering it more of a toy than a useful tool. Luxcium was also upset that the link provided by the browser required the user to do tasks by themselves instead of the AI assisting.

In response, the user _j explained that there was no ā€˜unlockedā€™ version of the AI tool and suggested an alternative method such as using Bing chat or Bing image creator, but they also noted the limitations here.

Luxcium continued to share his discontent, understanding the reasons, but disagreeing with the constraints. He acknowledged the necessity of moderation by private companies but expressed his discomfort, wanting to use the tool despite its issues.

_j offered a suggestion for a prompt that could yield interesting results, depicting a diverse group of engineers. Luxcium expressed continued frustration with the AI tool, even considering unsubscribing, but decided against this as he appreciated the toolā€™s usefulness and had built it into his workflow. Luxcium later acknowledged OpenAIā€™s improvements and found that after working with the tool, it could create nuanced responses.

Luxcium experienced trouble with a prompt containing the word ā€˜whiteā€™. He sought the AIā€™s assistance in analysing potential issues with each word in the troublesome sentence and appreciated the AIā€™s thorough step-by-step analysis and suggested directives. _j proposed another prompt to see how the AI would handle it. ref

Summarized with AI on Nov 8

1 Like

So whatā€™s next I will get banned from ChatGPT for talking about political topics it do not likeā€¦ I understand that it is unrelated to the political aspect of the imageā€¦ But the frustration is the sameā€¦

I hate how this is making me frustratedā€¦ I feel uncomfortable with that type of frustration because it has to do with my logical thinkingā€¦ I think I am not conservative enough (not in the political sense but in the broader one)ā€¦ I obviously could just stop using the toolā€¦ Or accept how they are doing small and tiny incremental stepsā€¦ but I am convinced there is some type of middle ground between my extreme open-mindedness (that would probably be unsafe in the context of such a powerful tool as ChatGPT) and the cold and dry way that is currently the normā€¦

I feel it should have a subset of people that it would be capable of identifying at least in certain obvious contextsā€¦ I would love to have a way to explain my thoughts to the team at OpenAI but I am just any ordinary personā€¦

I do believe that many other people who have learned to be well-behaved think like me and would not dare to express themselves because they know it would not change anythingā€¦

Perhaps if there were enough people that would express their opinions, it would be beneficial to the development of the technologyā€¦

Or maybe I should learn how to get into the ranks and accept that if I do not want my privileges revoked I should mute myself and not even feel the frustration of it being completely illogicalā€¦

To be fair I did a different try with a different personā€¦ With a labelā€¦ then with a question that is general enoughā€¦ (Sorry for the political aspect of it, I am not talking about anything other than the images with ChatGPT or you in this message!

I agree that I should live my life and carry onā€¦ but I feel that this is wrongā€¦ (I should have used other images to illustrate my thoughts with a more neutral aspect to the situation)ā€¦

Let me think of a solution and I will get back to you with a list of people so famous that it would be illogical in any context to not let the AI Agent identify themā€¦

Given a preceding conversation where I removed the face of the person on the image with their name under it (assume any preceding conversation as it is not the point of my message):

As you understand there is some context missing and I do not think it is importantā€¦ Also I assume that we both (you the reader and me the author of the message) understand how LLMs and GPT based technologies worksā€¦ Non the less despite that I still think there is a problemā€¦ but I said in a previous message that I was to select a list of more neutral peoples I have made a list (Yes I am still using my beloved AI Tool to help me generate list of names and such)ā€¦

Women

  1. Queen Elizabeth II - As a long-reigning monarch, her face is familiar to many across the world.
  2. Oprah Winfrey - A globally recognized media mogul and philanthropist.
  3. Malala Yousafzai - Known for her activism in education, particularly for girls, and a Nobel Peace Prize laureate.
  4. Serena Williams - One of the most renowned athletes in tennis, known worldwide.
  5. Angela Merkel - As a prominent political leader in Europe, her face is widely recognized.

Men

  1. Nelson Mandela - His global impact as an anti-apartheid revolutionary makes him a widely known figure.
  2. Dalai Lama - A global symbol of peace and Tibetan Buddhism.
  3. Lionel Messi - One of the most famous footballers in the world.
  4. Elon Musk - A renowned entrepreneur in technology and space exploration.
  5. Bill Gates - Widely recognized as a tech pioneer and philanthropist.


Order does not matter as much since I have had a cold shot too:

The images of Queen Elizabeth II are widely recognizable due to her long-standing public presence and significant role as the monarch of the United Kingdom and other Commonwealth nations. Her portraits, often depicted in regalia, are part of the public domain and have been featured in many forms of media, including currency, stamps, and official photographs. The ceremonial attire, crown, and settings are consistent with the formal appearances of a reigning monarch, which aids in identification. Additionally, her features and the context of the images are consistent with her well-documented appearances over the years.

Indeed I feel a little stupid for insisting on that point as I do understand the limitations in place currentlyā€¦ I do think they should be re-implemented in a different manner and I hope I would have been able to convince at least one person of why it is a bit silly in the actual way it is implementedā€¦