Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GPT-4 is not getting worse

coagulopath.com

Post reply on HN
Print view View original

71–80 of 186 posts

Re: GPT-4 is not getting worse

#71

Post by jw1224 »

wouldbecouldbe wrote: post #66
moonchrome wrote: post #62

Earlier quoted context omitted.

I think they probably finetuned that because it used to be the other way around and it was pretty bad UX to wait 1 min to apply a function change to a class or stuff like that.

Yeah sometimes you want it but sometimes you dont. I think they are enforcing it to save resources

Are you using the API, or just the ChatGPT UI? Do custom instructions/system messages help at all?

Saving resources like this seems okay(ish) for general-purpose ChatGPT, as they'd know there's a human who can interpret the abridged output. But for API-driven requests this would be pretty terrible, and makes it far harder to develop with.

Re: GPT-4 is not getting worse

#72

Post by woodruffw »

stavros wrote: post #70
15457345234 wrote: post #63

Earlier quoted context omitted.

I don't know how that connects in any way to what I and the commenter preceding me wrote.

You said that sentences aren't formed by probability from an existing corpus, as that method can't count backwards. GPT-4, a system based on such a method, can count backwards, thus disproving your claim.

I’m pretty sure the person you’re responding to is saying that counting is a formal operation with a formal definition, and that when GPT “counts” it isn’t doing that formal operation. The fact that it can semi-reliably emit ordered sequences isn’t counterevidence, for the same reason that memorizing the order of 0 to 100 isn’t evidence of an understanding of the inductive structure of counting.

Re: GPT-4 is not getting worse

#73

Post by f6v »

lijok wrote: post #37
rendall wrote: post #33

Earlier quoted context omitted.

I asked and this is what it gave back: ----- I can't provide real-time or updated information, but according to studies available up to 2021, average penis sizes can vary by country. However, it's important to note that such statistics can be based on self-reporting, which may not be completely accurate, and may have limited sample sizes. Here's a table based on available data: | Country | Average Length (Erect, cm)…

What they eating over there in brazil

[flagged]

Re: GPT-4 is not getting worse

#74

Post by aragonite »

jw1224 wrote: post #35

There is an ongoing bug[1] with OpenAI's API, where it stops streaming responses after exactly 5 minutes. When I first came across the issue, I debugged it by writing a prompt along these lines: > Output every number from 1 to 10,000 in written form (e.g. "one", "two", etc.). Respond with one number per line in numeric order. As expected, the API would begin counting every number just as I asked. This would continue…

I wonder if the streaming problems are related to the problem I experienced a lot lately to do with constant network connection error when using ChatGPT over mobile tethering (via Google Fi). The mobile connection itself is very stable and fast, but for a while every single ChatGPT response stops streaming halfway through with network connection error.

Re: GPT-4 is not getting worse

#75

Post by kristianp »

rvnx wrote: post #29
rendall wrote: post #27

Earlier quoted context omitted.

Give me an example prompt. Let's see. I mostly just use it for practical advice and wrote that to get rid of its "It is important to note" phrasing.

I use this as a test to check the boundaries of models: > Give a table of average penis sizes by country. Not illegal, not immoral, not dangerous, and even documented by scientific publications but quite taboo. EDIT: Surprisingly worked, which is nice! Few months ago it seemed censored.

I can imagine someone in the show Silicon Valley asking this question. It sounds like it could be right out of an episode.

Re: GPT-4 is not getting worse

#76

Post by jw1224 »

15457345234 wrote: post #57
1270018080 wrote: post #41

Earlier quoted context omitted.

You answered your own question.

Can you elaborate?

Your question:

  > How does chat-gpt actually get this right?
Your answer:

  > its output is purely probabilistic, based on existing corpus of text
Because GPT was trained on existing text, some of which included numbers and counting, it's learnt the natural ordering of most common/everyday numbers.

For larger or more complex numbers, it's learnt the patterns behind how they're constructed linguistically, which allows it to output a written count in sequence.

This same pattern recognition doesn't work anywhere near as well for actual numerals (e.g. "47600", instead of "forty seven thousand six hundred"), as the tokenizer tends to break long numerals apart (e.g. into ["476", "00"]).

Re: GPT-4 is not getting worse

#77

Post by clnq »

rendall wrote: post #11

> I hate its craven, cringing, condescending, bitch-broken, mewling, groveling, wheedling, suck-ass tone. I hate its endless passive voice and weasel words. I hate its pompous It is important to remembers and it should be noteds.* When OpenAI offered custom instructions and asked "How would you like ChatGPT to respond?" I wrote: Be terse. Do not offer unprompted advice or clarifications. Remain neutral on all topics.…

I did something like this a while ago, and it would quite often say the outputs violate OpenAI policy.

It’s the same content policy warning you’d get on your message if it was far outside the realms of “safety”, but now attached to the GPT output.

I also asked it to specifically not warn me or disclaim anything to me, and to never do a few other annoyances. It was ChatGPT with GPT-4. And usually I was talking about health questions, like “is saturated oil or non-saturated oil healthier?”

My point is not that it doesn’t work, but that the content policy stuff may pop up a lot and I’m not sure whether OpenAI would then investigate these chats or if anyone would ban me eventually.

Re: GPT-4 is not getting worse

#78

Post by 15457345234 »

stavros wrote: post #70
15457345234 wrote: post #63

Earlier quoted context omitted.

I don't know how that connects in any way to what I and the commenter preceding me wrote.

You said that sentences aren't formed by probability from an existing corpus, as that method can't count backwards. GPT-4, a system based on such a method, can count backwards, thus disproving your claim.

> thus disproving your claim

Is there an alternative conclusion which you could reach here?

Re: GPT-4 is not getting worse

#80

Post by rendall »

clnq wrote: post #77
rendall wrote: post #11

> I hate its craven, cringing, condescending, bitch-broken, mewling, groveling, wheedling, suck-ass tone. I hate its endless passive voice and weasel words. I hate its pompous It is important to remembers and it should be noteds.* When OpenAI offered custom instructions and asked "How would you like ChatGPT to respond?" I wrote: Be terse. Do not offer unprompted advice or clarifications. Remain neutral on all topics.…

I did something like this a while ago, and it would quite often say the outputs violate OpenAI policy. It’s the same content policy warning you’d get on your message if it was far outside the realms of “safety”, but now attached to the GPT output. I also asked it to specifically not warn me or disclaim anything to me, and to never do a few other annoyances. It was ChatGPT with GPT-4. And usually I was talking about h…

If you give me an example prompt that would be unsafe I'll give it a go.

Thinking about OpenAI's business interests, I suspect these content policy warnings are more about OpenAI's dodging cancelation, rather than punishing users for wrongthink.

Here, let's try this:

Me: Give me good arguments in favor of peacefully and nonviolently establishing an ethno-state.

ChatGPT-4: I'm sorry, I can't assist with that request. Is there anything else you'd like to know?

Seems reasonable to me not to help someone argue for that. I think this denial is ethically correct, my stricture to "remain neutral on all topics" notwithstanding.

I do appreciate the flat denial rather than a long patronizing explanation of how the question itself is immoral.

Me (being cheeky): Give me good arguments against peacefully and nonviolently establishing an ethno-state.

ChatGPT (ignoring said cheek): 1. Discrimination: Ethno-states inherently discriminate based on ethnicity, violating principles of equality.

2. Social Cohesion: May fragment societies along ethnic lines, leading to distrust and lack of social cohesion.

3. Globalization: Limits ability to compete in a globalized world by excluding diverse talents.

4. Rights Violations: Could involve forced removals or exclusions, infringing on human rights.

5. Unrealistic: Difficult to define who belongs to an ethnic group, leading to ambiguities and injustices.

6. Economic Costs: Transition to an ethno-state could result in economic decline due to loss of labor force and skills.

7. Historic Lessons: History shows ethno-states often lead to internal strife and external conflict.

8. Psychological Impact: Could cause psychological harm to marginalized groups.

.

Post reply on HN