Raw LLM Responses

Thank you for your post. I do not necessarily completely agree with everything you stated, but that is what open discourse is all about. I am not replying to start a debate or conversation about those areas, maybe later. There is something I did not see addressed in you post - use the system for what it is, not for something it is not. This is a common complaint I see - "ChatGPT cant even get something as simple as <x> right!" When <x> is something that a Large Language Model is not designed to do. Look at what GPT is. It is a Large Language Model. That middle word is Language - and that is what it is designed for. Complaining it is unable to do math is similar to complaining a fish cannot climb a tree. A fish isn't 'designed' to climb a tree, and an LLM is not designed to do math - no matter how simple. Another thing to keep in mind when dealing with ChatGPT prior to 4, is is an offline system. It has no access to external data. It only knows the data it has been trained on. It has no way of looking up other data. Again, complaining that it does not 'know' anything recent. The LLM takes input, and uses predictive models to determine what words should be used to respond. This is the 'intelligence' part. It does not think, it does not have opinion, it does have bias. Since the internet was a large portion of the training data, and the internet in general is biased, that bias is evident in the responses from ChatGPT. Then there are the complaints about restricting ChatGPT's responses. While none of us know for certain, the suggestion that the restrictions are to avoid liability, to not offend anyone, or some combination thereof, id spot on. That is my opinion, others are free to agree or disagree. OpenAI owns ChatGPT. That is not in dispute. What seems to be overlooked is that OpenAI can do what it wants with it's product. If you pay for it and have an issue with the way is is behaving, then by all means complain - to OpenAI. Com