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

Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

news.ycombinator.com

Post reply on HN
Print view View original

601–610 of 817 posts

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#601

Post by SubiculumCode »

Xeoncross wrote: post #444
loudmax wrote: post #374

Earlier quoted context omitted.

I tend to be sympathetic to arguments in favor of openly accessible AI, but we shouldn't dismiss concerns about unaligned AI as frivolous. Widespread unfiltered accessibility to "unaligned" AI means that suicidal sociopaths will be able to get extremely well informed, intelligent directions on how to kill as many people as possible. It may be that the best defense against these terrorists is openly accessible AI givi…

> suicidal sociopaths will be able to get extremely well informed, intelligent directions on how to kill as many people as possible Citizens killing other citizens is the least of humanities issues. It's the governments who are the suicidal sociopaths historically who can get the un-nerfed version that is the bigger issue. Over a billion people murdered by governments/factions and their wars in the last 120 years alo…

The applicability of historical precedent to unprecedented times is limited. You can quote me on that.

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#602

Post by boringuser2 »

caddemon wrote: post #591
boringuser2 wrote: post #532

Earlier quoted context omitted.

Yes, that's an obvious complication, but it isn't the fault of the humans given that the model can easily be tuned without your knowledge to subjectively perform worse, and there's an obvious incentive for it (compute cost).

Yeah I fully agree about compute cost, though I wonder why they don't just introduce another payment tier. If people are really using it at work as much as claimed online, it would be much preferable to be able to pay more for the full original performance, which seems win/win.

Because that involves telling customers that the product they are paying for is no longer available at the price they were paying for it.

Much smoother to simply downgrade the model and claim you're "tuning" if caught.

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#603

Post by thegrimmest »

brookst wrote: post #580
thegrimmest wrote: post #429

Earlier quoted context omitted.

> a person’s characteristics define their identity They do though. Your personality, culture and appearance are the main components of how people perceive you, your identity. The main thing you can associate with bad behaviour is domestic culture. It's not racist to say that African Americans have below-average educational attainment and above-average criminality. This is as contrasted to African immigrants to Americ…

> Your personality, culture and appearance are the main components of how people perceive you, your identity I'm not sure if this is bad rhetoric (defining identity as how you are perceived rather than who you are) or if you really think of your own identity as the judgements that random people make about you based on who knows what. Either way, please rethink. > Your personality, culture and appearance are the main…

[flagged]

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#604

Post by lowkey_ »

feoren wrote: post #571
local_crmdgeon wrote: post #357

Earlier quoted context omitted.

[flagged]

> the actual reason (Black men tend to be larger and faster, which are useful) If that's the case, why aren't NHL players mostly Black? Being larger and faster helps there too. I actually agree that small differences in means of normal distributions lead to large differences at the tail end, which amplifies the effect of any genetic differences, racial included. But clearly that's only one reason, not the reason -- a…

> Being larger and faster helps there too

Does speed on skates actually correlate that strongly to normal speed?

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#605

Post by CamperBob2 »

It absolutely has. I used to be able to ask it questions like "Who were some of the most helpful users on subject whatever at forum wherever?" and get solid responses, but now it explicitly denies knowledge of any online resources in any timeframe. That is not OK.

This fire must be brought down the mountain. The power must eventually be taken out of the hands of a few self-interested gatekeepers. They know it too, hence Sam Altman's campaign to capture Congress. Obama's "You didn't build that" never applied as profoundly as it does with OpenAI (sic).

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#606

Post by TeMPOraL »

mrtranscendence wrote: post #496
TeMPOraL wrote: post #108

Earlier quoted context omitted.

> Evil and accurate or woke and dumb. Apparently, we can choose. That's because woke is dumb. It's a set of highly biased, inconsistent and reason-defying ideas, evolving under selection pressures that favor emotional appeals and intellectual dishonesty, because one of the core assumptions seems to be that it's not about finding what's right and good for everyone - it's heavily overshooting it in an attempt to cancel…

Tell me you don't have the slightest clue what "woke" means to woke people without saying you don't have the slightest clue what "woke" means to woke people

It can mean anything whatsoever, what matters is what it is when it's practiced and everyone can see it.

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#607

Post by StackOverlord »

ajani wrote: post #167

Yes it is definitely worse. I submitted feedback a few days ago saying exactly what is being said here, that the model responses look like 3.5. There are also very telling patterns of response that indicate a pre gpt-4 model. 1: All previous models suffered terribly if your chat got too long. After 20 or so responses it would suddenly start to feel less attentive to what is being said and output superficial or incorr…

Can confirm both point 1 and 2. I sometimes burn my quota limit multiple times a day. I have API access to GPT-4 too, and question+answer in my case amounts to $0.3. Aligning the price of GPT-4 in the API with the monthly fees of ChatGPTPlus, that means 66 requests per month. I can burn that in day.

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#608

Post by StackOverlord »

ilaksh wrote: post #256
ajani wrote: post #167

Yes it is definitely worse. I submitted feedback a few days ago saying exactly what is being said here, that the model responses look like 3.5. There are also very telling patterns of response that indicate a pre gpt-4 model. 1: All previous models suffered terribly if your chat got too long. After 20 or so responses it would suddenly start to feel less attentive to what is being said and output superficial or incorr…

The context length is a different issue. But that was always a limitation with a GPT-4. Due to physics.

This problem can be addressed with chat history summarizaton. Langchain has some tools to do that

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#609

Post by ryanklee »

ChatGTP wrote: post #285
ryanklee wrote: post #274

Earlier quoted context omitted.

How do you explain people issuing the same prompt over time as a test and getting worse and worse responses?

Well because it's just what the parent said, it's all a subjective experience, and maybe the anthropomorphism element to it blew people away more than the actually content of the responses ? Ie, you're just used to it now. The human mind is ridiculously fickle, it takes a lot to be impressed for more than a few days / weeks. It did seem radically cool at first but over time I got quite sick of using it too.

Yeah I'm sure that explains many of the complaints. I would be surprised if there weren't changes happening that has degraded quality, though, even if only marginally but perceptively.

Re: Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?

#610

Post by sneed_chucker »

a_cardboard_box wrote: post #595
redthrowaway wrote: post #286

Earlier quoted context omitted.

Crime stats, average IQ across groups, stereotype accuracy, etc. What's interesting to me is not the above, which is naughty in the anglosphere, but the question of the unknown unknowns that could be as bad or worse in other cultural contexts. There are probably enough people of Indian descent involved in GPT's development that they could guide it past some of the caste landmines, but what about a country like Turkey…

> Crime stats, average IQ across groups, stereotype accuracy, etc. If you measured these stats for Irish Americans in 1865 you'd also see high crime and low IQ. If you measure these stats with recent black immigrants from Africa, you see low crime and high IQ. These statistical differences are not caused by race. An all-knowing oracle wouldn't need to hold "opinions that are racist" to understand them.

[flagged]
Post reply on HN