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

Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

openai.com

Post reply on HN
Print view View original

351–360 of 433 posts

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#351

Post by pdntspa »

azuanrb wrote: post #278

It’s interesting that many comments mention switching back to Claude. I’m on the opposite end, as I’ve been quite happy with ChatGPT recently. Anthropic clearly changed something after December last year. My Pro plan is barely usable now, even when using only Sonnet. I frequently hit the weekly limit, which never happened before. In contrast, ChatGPT has been very generous with usage on their plan. Another pattern I’…

I am using my claude pro plan for at least 8 hours a day, 5 days a week, to maintain a medium-sized codebase, and my total weekly usage is something like 25% of my limit.

What the hell are people doing that burns through that token limit so fast?

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#352

Post by pxc »

unethical_ban wrote: post #101
pxc wrote: post #28

Earlier quoted context omitted.

The range of attitudes in there is interesting. There are a lot of people who take a fairly sensible "this is interactive fiction" kind of attitude, and there are others who bristle at any claim or reminder that these relationships are fictitious. There are even people with human partners who have "married" one or more AIs.

IIRC you'll get modded or banned for being critical of the use case. Which is their "right", but it's freaking weird.

Yes. My experience is that it doesn't require scolding, mocking, or criticizing anyone to get permabanned. Just being up front about the fact that you have concerns about the use case is enough for a permaban, even if you only bring that up in order to demonstrate that such a position does not stem from contempt for LLM-as-companion users. :-\

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#353

Post by palmotea »

fc417fc802 wrote: post #339
palmotea wrote: post #330

Earlier quoted context omitted.

> including the part where alcohol isn't considered to be a drug for some strange reason. It's not a strange reason. IIRC, most cultures have a culturally understood and tolerated intoxicant. In our culture, that's alcohol. Human culture is not some strange robotic thing, where the expectation is some kind hyper consistency in whatever narrow slice you look at.

I don't object to alcohol being tolerated. But I do think that distinguishing it from other drugs is odd. Particularly when the primary reason given for regulating other drugs is their addictiveness which alcohol shares. We tolerate a recreational drug. Lots of people regularly consume a recreational drug and yet somehow society doesn't split at the seams. We should just acknowledge the reality. I think people would…

> I don't object to alcohol being tolerated. But I do think that distinguishing it from other drugs is odd.

The point I was making is that it's not odd, unless you're thinking about human culture wrong (e.g. like its somehow weird that broad rules have exceptions).

> Particularly when the primary reason given for regulating other drugs is their addictiveness which alcohol shares.

One, not all addictive drugs are equally addictive. Two, it appears you have a weird waterfall-like idea how culture develops, like there's some kind identification of a problematic characteristic (addictiveness), then there's a comprehensive research program to find all things with that characteristic (all addictive substances), and finally consistent rules are set so that they're all treated exactly the same when looked at myopically (allow all or deny all). Human culture is much more organic than that, and it won't look like math or well-architected software. There's a lot more give and take.

I mean here are some obvious complexities that will lead to disparate treatment of different substances:

1. Shared cultural knowledge about how to manage the substance, including rituals for use (this is the big one).

2. Degree of addictiveness and other problematic behavior.

3. Socially positive aspects.

4. Tradition.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#354

Post by fragmede »

Wowfunhappy wrote: post #305
satisfice wrote: post #302

Earlier quoted context omitted.

Have you ever stopped to realize that, from the Victorian’s point of view, they have been proven completely right about what would happen if ladies started showing their ankles?

...sorry, I'm dense apparently, what did they predict vs what happened?

The Victorians were accidentally right about ankles, which is funny in hindsight. Once one arbitrary rule breaks, people start noticing the rest are kind of fake too, and it turns out "modesty" was load-bearing for a whole governance model.

Ankles -> knees -> jazz -> voting -> rock -> no-fault divorce -> Tinder -> polyamory discourse on airplanes. it's a joke, but also sort of how cultural change actually propagates. The collapse did happen, just not of morals. Of enforcement. After that, everything is just people discovering the rules were optional all along. Including money.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#355

Post by ComputerGuru »

Scene_Cast2 wrote: post #65
NewsaHackO wrote: post #49

>We brought GPT‑4o back after hearing clear feedback from a subset of Plus and Pro users, who told us they needed more time to transition key use cases, like creative ideation, and that they preferred GPT‑4o’s conversational style and warmth. This does verify the idea that OpenAI does not make models sycophantic due to attempted subversion by buttering up users so that that they use the product more, its because peop…

As someone who's worked with population data, I found that there is an enormous rift between reported opinion (and HN and reddit opinion) vs revealed (through experimentation) population preferences.

Not true. People can rationally know what they want but still be tempted by the poorer alternative.

If you ask me if I want to eat healthy and clean and I respond on the affirmative, it’s not a “gotcha” if you bait me with a greasy cheeseburger and then say “you failed the A/B test, demonstrating we know what you actually want more than you.”

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#356

Post by palmotea »

noosphr wrote: post #189
palmotea wrote: post #135

Earlier quoted context omitted.

That's why laws against drugs are so terrible, it forces law-abiding businesses to leave money on the table. Repeal the laws and I'm sure there will be tons of startups to profit off of drug addiction.

The Politician's syllogism in action: That is terrible. Se have to do something. This is something. We must do it. It terms of harm current laws on drugs fail everyone but teetotaller who want everyone else to have a miserable life too.

> It terms of harm current laws on drugs fail everyone but teetotaller who want everyone else to have a miserable life too.

You think teetotallers have miserables lives? Come on.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#357

Post by jghn »

pdntspa wrote: post #351
azuanrb wrote: post #278

It’s interesting that many comments mention switching back to Claude. I’m on the opposite end, as I’ve been quite happy with ChatGPT recently. Anthropic clearly changed something after December last year. My Pro plan is barely usable now, even when using only Sonnet. I frequently hit the weekly limit, which never happened before. In contrast, ChatGPT has been very generous with usage on their plan. Another pattern I’…

I am using my claude pro plan for at least 8 hours a day, 5 days a week, to maintain a medium-sized codebase, and my total weekly usage is something like 25% of my limit. What the hell are people doing that burns through that token limit so fast?

The other day I asked CC to plan using Opus a few small updates to a FastAPI backend & corresponding UI updates for a nextJS frontend. Then I had it implement using Sonnet. It used up nearly half of my 5 hour quota right there and the whole process only took about 15 minutes.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#358

Post by ComputerGuru »

Hard_Space wrote: post #251
QuadrupleA wrote: post #247

Been unhappy with the GPT5 series, after daily driving 4.x for ages (I chat with them through the API) - very pedantic, goes off on too many side topics, stops following system instructions after a few turns (e.g. "you respond in 1-3 sentences" becomes long bulleted lists and multiple paragraphs very quickly. Much better feel with the Claude 4.5 series, for both chat and coding.

> you respond in 1-3 sentences" becomes long bulleted lists and multiple paragraphs very quickly This is why my heart sank this morning. I have spent over a year training 4.0 to just about be helpful enough to get me an extra 1-2 hours a day of productivity. From experimentation, I can see no hope of reproducing that with 5x, and even 5x admits as much to me, when I discussed it with them today: > Prolixity is a side…

You can’t ask GPT to assess the situation. That’s not the kind of question you can count on a an LLM to accurately answer.

Playing with the system prompts, temperature, and max token output dials absolutely lets you make enough headway (with the 5 series) in this regard to demonstrably render its self-analysis incorrect.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#359

Post by ComputerGuru »

thedudeabides5 wrote: post #147

will this nuke my old convos? opus 4.5 is better at gpt on everything except code execution (but with pro you get a lot of claude code usage) and if they nuke all my old convos I'll prob downgrade from pro to freee

Not that I’m aware. Models can be fairly seamlessly switched even mid-conversation, so this is unlikely to affect history.

Re: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT

#360

Post by ssl-3 »

SoftTalker wrote: post #332
mFixman wrote: post #311

Earlier quoted context omitted.

I've been impressed by how good ChatGPT is at getting the right context old conversations. When I ask simple programming questions in a new conversation it can generally figure out which project I'm going to apply it to, and write examples catered to those projects. I feel that it also makes the responses a bit more warm and personal.

ChatGPT is what work pays for so it's what I've used. I find it grossly verbose and obsequious, but you can find useful nuggets in the vomit it produces.

ChatGPT can very much be that way.

It can also be terse and cold, while also somewhat-malleably insistent -- like an old toolkit in the shed.

It's all tunable.

Post reply on HN