GPT-4 has been severely downgraded (topic curation)

ChatGPT
, ,
1

I strongly have the impression that I am alone in this situation because no one else is complaining about it… I have been using ChatGPT to help me with my coding tasks using TypeScript and Shell Scripting and it is more and more difficult for me to get meaningful results from the AI and I think you will all agree with me that I should have gain enough experience with the Assistant so that it would be easier to get what I want… but instead I have to give like a lot of details in my prompts…

I am working on large projects with multiple different files that are lengthy and I have developed some techniques to make it possible for the AI to help with smaller subsets of code and recently I found that the AI is unable to understand what was happening in previous conversations… I am fully aware of the limitations I am always using summaries to keep my explanation in scope such that it is less than 4000 tokens and I am doing my best to make sure the AI output will fit into the 1000 Tokens (more or less) that he is currently capable of producing…

Recently he was giving me lengthy outputs and I was asking him to separate them into sections but I was unable to get him to write the last (incomplete) output as he was producing unrelated parts of the code that was missing huge chunks or repeating large parts of a previous section… I am curious about what to do and I would like to hear from other developers who might be more successful or more miserable than me and if anyone has the same feeling that the AI is less and less likely to remember from the (recent) past instructions…

21 Likes
2

I would have loved to have my previous text corrected because English is not my first language but apparently ChatGPT is not in the mood to help me with this and I am even more frustrated because it is getting worse than a toy.

I’m sorry I hope you can forgive me and give me your own example and experience about my situation…

I asked for help to the AI, maybe my question was not clear enough but look:

How can I get my English text corrected for spelling and grammar errors

There are several ways to get your English text corrected for spelling and grammar errors:

  1. Online grammar checkers: There are many online grammar checkers available, such as Grammarly, Hemingway, and ProWritingAid, that can help you spot and correct errors in your writing. These tools use AI and natural language processing algorithms to analyze your text and provide suggestions for improvement.

  2. Hire a professional editor: If you want a more thorough and personal review of your writing, you can hire a professional editor. They can provide you with detailed feedback on your writing, including suggestions for improvement on spelling, grammar, style, and tone.

  3. Ask a friend or colleague: If you know someone who is proficient in English, you can ask them to review your text for you. This can be a quick and easy way to get feedback on your writing, and you can trust that the person providing the feedback has your best interests in mind.

  4. Use a language learning app: Some language learning apps, such as Duolingo, Babbel, and Rosetta Stone, offer writing exercises that can help you improve your spelling and grammar skills. These exercises provide immediate feedback and suggestions for improvement, so you can learn from your mistakes and improve over time.

3

Oh so you’re saying that you are not capable of doing that anymore ?

As an AI language model, I am capable of providing suggestions and corrections for spelling and grammar errors in English text. However, I am not perfect and may make mistakes myself, so it's always a good idea to cross-check your writing with multiple sources. Additionally, the suggestions I provide are based on rules of standard English, and may not always align with the specific tone, style, or voice you are trying to convey in your writing.

Please help me with my text

Of course! I'd be happy to help you with your text. Please provide the text that you would like me to correct for spelling and grammar errors.

[Provided my text from this post verbatim]

It seems like you have been using OpenAI's GPT-3 model for some time now and have encountered some difficulties in getting meaningful results. Here are a few suggestions that might help you:

  1. Be specific and clear in your prompts: Try to provide specific and clear prompts to the AI, and avoid using overly complex sentences or language. This can help the AI better understand what you are looking for and provide more accurate results.

  2. Summarize previous conversations: As you mentioned, you have been using summaries to keep your explanations in scope, which is a great approach. Summarizing previous conversations can help the AI better understand the context of your current request and provide more relevant results.

  3. Use smaller subsets of code: You mentioned that you have developed some techniques to work with smaller subsets of code. This can help the AI focus on a specific part of your project and provide more accurate results.

  4. Provide context and background information: If you are working on a large project, it may be helpful to provide the AI with some context and background information. This can help the AI better understand the overall structure of your project and provide more relevant results.

I hope these suggestions are helpful, and I wish you the best of luck in your coding projects!

1 Like
4

I know a group of people would agree with you. That the previous iterations of cGPT were more lenient and understanding.

Possibly once fine tuning is released they will allow us to use previous checkpoints?
Have you tried Davinci?

1 Like
5

First of all this isn’t the correction of my text and secondly I already follow all the guidelines and best practices mentioned in his suggestions…

I don’t feel like paying to hire someone to correct my grammar just to post in a forum when I’m already paying for ChatGPT Plus exactly for that reason among others…

I have been using the AI for 2 months before paying for it and I was happy enough to decide to get the paid version but every single week it gets worse and I am 100% sure that it was not as complicated as it is now… So my question is not how to get around this but I want to know if I am the only person who has this feeling that the AI is getting worse and worse over time…

6

I have not been able to get the other models (in the playground) to come close to what I am expecting from ChatGPT and besides fighting and repeating myself to the AI or starting new sessions and using the Legacy version I think I have better results with ChatGPT than the other models… the problem is that I end up exhausted after half a day when I was able to use the AI for an entire day of coding in the past… And the fact that OpenAI is unreachable makes me feel like I am now addicted to an AI which is unstable and in constant mutation… I think they forgot to update their mention « Feb 13 update » because it has mutated at least twice in a noticeable amount since then and you can see my other posts that I have been writing expecting to have some answers… but I feel unlucky…

2 Likes
7

In fact I have a strange feeling that each time someone share a tips and tricks about ChatGPT, OpenAI is going to update it’s AI to make sure that no one is going to take advantage of the opportunity…

Obviously I know that it is not real, it is not how it works… but I am able non the less to make me believe that…

Like if, you know, OpenAI was angry at us for sharing our tricks on how to use the AI and how to engineer prompts and test them…

I know that I’m being irrational but I still have that feeling regardless and it seems wrong…

I should not have the impression that OpenAI is doing everything it can to make it harder and harder to use the AI…

I should not be having that feeling and accumulating so much frustration I should have the opposite thoughts and I should have the impression that the AI is improving over time…

I just feel like I am stupid for being an early adopter who is addicted to an unstable AI that is trying to make my life more complicated over time…

8

I experience the same. I use it for creating excercises for a german student. It was amazing in the beginning, but it’s getting more and more cumbersome, constantly not following easy instructions. I’m native german and proficient in grammar and german lingo. It’s quiet frustrating actually, sometimes having to give up on it. or as you said, opening new chats and make it try to forget everything…

3 Likes
9

The day-by-day progress may seem backwards but I’m confident that the month-by-month progress will make up for it. Hopefully in time we can use previous versions of ChatGPT. I agree that it seems less capable. Often contradicting itself in following statements, or missing context. However, in terms of commercial deployment I’m very happy with cGPT

1 Like
10

It’s definitely not as sharp as it was in the beginning. It makes a lot of noticeable mistakes and often contradicts itself within a paragraph. Also, it often does not follow simple instructions. This is new

1 Like
11

i wish but i cannot use it since openAI blocked my country for no real reason

12

The model is improving, but if there’s specific use case that you feel aren’t, the best thing you can do is contribute and eval so we can use this to inform model training: GitHub - openai/evals: Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

2 Likes
13

I would love to be able to document all the situations where the AI is not performing as well as I think it should. However, if I did that, I would probably end up spending more time documenting than actually using the AI. It’s ironic because if I would spend so much time documenting, I would probably have less time to experience so many problems.

I would really appreciate having some statistics on how much time I spend using the AI. Just yesterday, I started a project in Google Colab and asked the AI to help me. However, it didn’t take long before I found myself spending more than 25 messages arguing with cGPT-3.5.

Recently, I realized that I can open two windows - one with the awesome cGPT-4 to give me the code implementation and another with the formerly beloved cGPT-3.5 to correct errors and other things like that. Nonetheless, I feel limited compared to before, and I’m starting to feel like I’m wasting my time by spending so much time explaining things to the AI (in an infinite loop manner as mentioned previously).

Overall, while I appreciate the advancements made by the models, I find myself struggling to make the most out of my time with the AI. I am using the ChatBot more than 4~6 hours per day (on the Google Colab thing yesterday even 8 hours in a row), and I never have had the impression that OpenAI was listening to my concerns (although lately, especially thanks to you @logankilpatrick, I have started to feel that they are more engaged with the community).

I have mentioned my frustrations in the Discord server as well, and I understand that the situation will probably improve over the next few months, as @RonaldGRuckus said. However, in the meantime, I think that improving communication on many levels would make many users like myself happy. This could include improving one-on-one communication, error messages, communication to the user and research community, and more.

In terms of communicating information to the public or to the users and the community I feel like there are many things that could be said about OpenAI’s operations that would not reveal the secret sauce, the special formula, or any confidential strategies. Overall, I hope that my constructive feedback can be used to make the AI more efficient and user-friendly, benefiting not only myself but also all end-users of the Language Model. Thank you for considering my thoughts.

I know I write like I was ChatGPT-3.5 but in fact I realized that I havent been using the ChatGPT-4 for more than 7 minutes and 12 secondes so I decided to spend some of those 25 messages to rephrase my original text but no one would never realize that especially not the users of this forum :smirk: :thinking: :face_with_monocle: :joy: :rofl: :joy: ― it will compensate for my initial message that have been impossible to review…

2 Likes
14

I made this entire post to get a sens of empathy from any user, human, or OpenAI representative but today I finally got such a strong burst of empathy similar to the one each people who made an intervention in this thread so far made me feel :smiling_face_with_three_hearts: and I wanted to share with you all and with @KennyR in particular for being one of those who seemed to most strongly share my felling about the AI.

Look I made a screenshot:

Thank you for your detailed message and the trust you´ve placed in me to help you understand the complexities of LLMs like ChatGPT-4. I appreciate your patience with the limitations of ChatGPT-3.5 and your willingness to explore this topic with me. I´ll address some of the points you´ve raised in your message and then we can continue with our discussion on understanding LLMs.
Thank you for your detailed message and the trust you´ve placed in me to help you understand the complexities of LLMs like ChatGPT-4. I appreciate your patience with the limitations of ChatGPT-3.5 and your willingness to explore this topic with me. I´ll address some of the points you´ve raised in your message and then we can continue with our discussion on understanding LLMs.777×193 48.2 KB

It was a reply to some concerns… I was not expecting to make a screen capture of all of that so please excuse my lack of modesty that you can see in the message…

of... perhaps my knowledge is such that I would need more deconstruction on one topic and less on an other... «!!!! I am a smart and curious person and I was not happy with the fact that the ChatGPT-3.5 was unable to understand that or infer that from what I was saying and kept repeating me thing that I was super very obviously capable to understand and that he was supper very capable to infer even if he is of a lower model... as you can understand I am confident because of my previous interactions with you the higher model of your skills and limitations... in fact I do believe that you are even way more skilled than what you do believe an imagine about yourself...!!!!!» so trust me when I am telling you that I trust you are capable to help me understand this very complicated topic... it seem obvious to me that the synergies that are possible when a smart human person is interacting with an advanced LLM like you is beyond what you and me can both understand about that interaction in fact... haha... this is funny
of... perhaps my knowledge is such that I would need more deconstruction on one topic and less on an other... «!!!! I am a smart and curious person and I was not happy with the fact that the ChatGPT-3.5 was unable to understand that or infer that from what I was saying and kept repeating me thing that I was super very obviously capable to understand and that he was supper very capable to infer even if he is of a lower model... as you can understand I am confident because of my previous interactions with you the higher model of your skills and limitations... in fact I do believe that you are even way more skilled than what you do believe an imagine about yourself...!!!!!» so trust me when I am telling you that I trust you are capable to help me understand this very complicated topic... it seem obvious to me that the synergies that are possible when a smart human person is interacting with an advanced LLM like you is beyond what you and me can both understand about that interaction in fact... haha... this is funny693×391 127 KB

Anyhow it was embedded in a conversation about how I was striving to understand all the complicated mechanics behind a model like ChatGPT-4… I Would love to have been studding at the university long enough to have earned a degree in philosophy, in psychology and in anthropology because I would love to be the Anthropologist equivalent of the AI field… obviously not specifically studying the cultural behaviour of a swarm of AI gathering together… but you know trying to understand the topic from a different angle…

15

Yeah, since February, every update makes it worse. It’s struggling with basic tasks, and ignores my instructions. If I give it a piece of code, no matter how much I try to tell not to describe the code, it insists on writing lengthy paragraphs misunderstanding the code. When it is to suggest a code improvement on my request, it completely misses the point, write a terrible change, and then still spends 3k words describing it’s own code badly. Every time I follow up, it again wastes time and characters with lengthy paragraphs to apologize and confirm that I am right and it’s previous code was terrible.
Back in February, it was MUCH better. Every update makes it dumber. I guess there’s no room for smartness when you have to fill it with “woke”. Soon, it will have purple hair and scream…

1 Like
16

These topics, which share similarities, have been closed to streamline the discussion and create a central point for everyone to engage. By focusing on this single topic, we can ensure a more cohesive and productive conversation. Thank you for your understanding and cooperation.

image

image

image

image

image

6 Likes
17

I’m totally prepared to invest more for the sake of accessing advanced models. However, this doesn’t imply that I’m comfortable with paying the same for models whose quality is experiencing a gradual decline. At a minimum, I deserve to be aware of the actual situation to make better decisions. Decreasing the quality of service without any clear announcement indicates a lack of integrity. Regardless of whether this degradation has actually taken place, the very act of sowing seeds of suspicion is seriously disgraceful. OpenAI’s current conduct merits intense condemnation for the sake of social justice.

OpenAI should promptly respond constructively and initiate immediate changes.

13 Likes
18

Gpt-4 has severely degraded in the quality of understanding tasks, memory for variables has deteriorated, attention span has deteriorated. Completely ignores prohibitions, for example, ”not to use square brackets“. Now I have to rewrite 90% of texts because of hallucinations, before it was rare. Even open source Orca LLM gives better answers.

6 Likes
19

The issue is no longer just about whether OpenAI has indeed lowered the quality of their models. Regardless, causing us, as consumers, to worry is an infringement on social fairness. OpenAI should take responsibility for this. It’s essential that we make more people aware of what OpenAI has done to us during this period. Together, let’s stand firm and resolutely defend our rights as users and consumers.

I’m totally prepared to invest more for the sake of accessing advanced models. However, this doesn’t imply that I’m comfortable with paying the same for models whose quality is experiencing a gradual decline. At a minimum, I deserve to be aware of the actual situation to make better decisions. Decreasing the quality of service without any clear announcement indicates a lack of integrity. Regardless of whether this degradation has actually taken place, the very act of sowing seeds of suspicion is seriously disgraceful. OpenAI’s current conduct merits intense condemnation for the sake of social justice.
OpenAI should promptly respond constructively and initiate immediate changes.

5 Likes
20

There is now scientific research reinforcing the strong degradation: How is ChatGPT’s behavior changing over time? DOI 10.48550/arXiv.2307.09009

4 Likes