News for hackers. Stuff that matters.
Re:Alpaca RLHF-ed to beat ChatGPT by r3trohack3r on Tuesday May 23, 2023 @09:41PM (#36053031)
Whenever I see a claim about GPT I get temporarily excited until I learn it’s GPT3.5 and not GPT4.

4 isn’t just marginally better at most tasks I use it for, it’s operating at an entirely different level to the point where I have little (no?) day-to-day use of 3.5 at this point.

Re:Alpaca RLHF-ed to beat ChatGPT by danielmarkbruce on Tuesday May 23, 2023 @09:46PM (#36053064)
From what I see practically everyone is making this comparison and it is bs. As you stay, 4 is an entirely different beast to 3.5.
Re:Alpaca RLHF-ed to beat ChatGPT by napsterbr on Tuesday May 23, 2023 @09:46PM (#36053066)
I'm assuming you use gpt4 via ChatGPT plus. Does the message cap bother you? I heard it's something like 25 messages per 3 hours. That sounds so low I don't even bother subscribing.

I guess this doesn't apply if you use it via the api.

Re:Alpaca RLHF-ed to beat ChatGPT by anonylizard on Tuesday May 23, 2023 @09:51PM (#36053123)
I was initially deterred. But in practice, when using it for professional purposes, I never encounter it.

Your coding speed is unlikely to be that fast, requiring 25 code segments in 3 hours. GPT-4 outputs something, you need time to double check, test, additional googling etc. Its still a massive speed boost.

Using it recreationally (Especially chatting) will result in a lot more requests.

Re:Alpaca RLHF-ed to beat ChatGPT by taf2 on Tuesday May 23, 2023 @10:21PM (#36053319)
The only time I hit it I usually realize I need a mental break anyway so usually it’s plenty. Suppose it depends on how your using it but for me I ask it for code of things I could write but would rather focus my energy on the bigger problem then a single function to merge two objects while keeping the order sorted of a joined list.. that kind of thing it’s great for
Re:Alpaca RLHF-ed to beat ChatGPT by bugglebeetle on Tuesday May 23, 2023 @10:56PM (#36053528)
I hit the limit all the time now since they’ve made GPT-4 so much worse. The amount of coaxing it and back and forth it needs has increased dramatically in the past month.
Re:Alpaca RLHF-ed to beat ChatGPT by tulip4attoo on Tuesday May 23, 2023 @10:32PM (#36053407)
I only hit the limit when I using some bots to interact with it. Never hit it after 2 months when normally using it.
Re:Alpaca RLHF-ed to beat ChatGPT by skybrian on Tuesday May 23, 2023 @10:21PM (#36053315)
I've never hit the limit because GPT 4 is slow (like a dialup modem) and I don't like waiting for it. Usually I do something else while it's writing a response.

I haven't used it a whole lot.

Re:Alpaca RLHF-ed to beat ChatGPT by r3trohack3r on Tuesday May 23, 2023 @09:51PM (#36053119)
It does bother me, I’ve been hit by it 3 times now (I use it as a daily driver, for code you spend enough time between prompts working that it’s rare to go through that volume)

When I hit the limit, I work on the problem myself and wait until 4 resets instead of relying on 3.5. 4 is so much better that I don’t trust 3.5 with my work anymore.

Re:Alpaca RLHF-ed to beat ChatGPT by furyofantares on Tuesday May 23, 2023 @09:48PM (#36053090)
It sounds very low and somehow it very rarely bothers me. It sure is annoying when it bothers me, but it's a lot higher in practice than the number feels.
Re:Alpaca RLHF-ed to beat ChatGPT by danielmarkbruce on Tuesday May 23, 2023 @10:50PM (#36053496)
In practice it rarely comes up. I don't even know if it's actually enforced. I'm pretty certain I do more than 25 every 3 hours.
Re:Alpaca RLHF-ed to beat ChatGPT by ineedasername on Tuesday May 23, 2023 @11:53PM (#36053886)
I have no idea what happened. My original 1-month Plus subscription to test GPT-4 was, as usual, limited to 25/hour and tied to my personal gmail address. But then recently-- in the last week-- I resubscribed and I accidentally did so under my work email, which has a more specialized and restricted TLD, and under that I don't have any quota limits for the GPT-4 version of ChatGPT.

It might seem a small thing, having to space out prompts 25 ever 3 hours when you might not have used more than 100-200 in a day anyway, but the net result is liberating. I experiment, explore the limits, and get whimsical with it to a much greater extent than when I can to consciously think about each prompt as a rationed resource.

Re:Alpaca RLHF-ed to beat ChatGPT by endofreach on Tuesday May 23, 2023 @11:59PM (#36053924)
For me tbh, i kinda like the limit. I use GPT-4 a lot lately. Hitting the limit reminds me, i got too lazy writing code myself or i got way too deep into it. Then i just close the tab & remember, that i still love writing code the (not quite yet) old way.
Re:Alpaca RLHF-ed to beat ChatGPT by bugglebeetle on Tuesday May 23, 2023 @10:51PM (#36053502)
GPT-4 barely performs above 3.5 now. They’ve resource-constrained or otherwise hobbled it to support all the corners of Microsoft products they’re stuffing it in. The amount of errors and logic degradation after the May update is incredibly obvious for all but the most trivial use cases.

It’s going to be very funny if being turned into a next generation Clippy is what makes them lose out to their competitors

Re:Alpaca RLHF-ed to beat ChatGPT by NicoJuicy on Tuesday May 23, 2023 @11:06PM (#36053590)
Microsoft hosts their own models through azure? Why would that resource constrain OpenAI?
Re:Alpaca RLHF-ed to beat ChatGPT by bugglebeetle on Tuesday May 23, 2023 @11:49PM (#36053849)
That’s just my speculation. GPT-4 has recently become more “snappy” at a significant degradation in reasoning and overall performance. My guess would be that they crippled it in someway to make performance better in the Microsoft integrations vs. it being a good but slow service in the ChatGPT UI.