Live data from Hacker News

Introducing ChatGPT and Whisper APIs

openai.com

151–160 of 696 posts

Re: Introducing ChatGPT and Whisper APIs

#152
post #15

Earlier quoted context omitted.

What the cost cutting measures suggest is that AI like this could maybe soon be run on consumer hardware. That combined with actually open source language models could be huge. OpenAI won't allow for that for obvious reasons, but this confirms that the optimizations are there, and that's exciting enough news on its own.

I mean, meta's new LAMA model runs on a single A100 in the 13B parameter variant (which performs similar to GPT3 65B).

"Performs" on paper until they give a demo.

Re: Introducing ChatGPT and Whisper APIs

#153

Earlier quoted context omitted.

It's amazing that "being two months out of date" in AI means that you are already a dinosaur

I scan the SD subreddit and am subscribed to 3 big ai art youtubes just to stay up to date. With things moving this fast, alot of info is out of date and can be very burdensome to comb through the good stuff later. I try and set aside 30mins twice a week to apply the new techniques to help cement them in my mind and see their strengths and weaknesses. ControlNET really changed the game and now OffsetNoise (check out…

What are the youtubers?

Re: Introducing ChatGPT and Whisper APIs

#154

Earlier quoted context omitted.

It's amazing that "being two months out of date" in AI means that you are already a dinosaur

Agriculture reduced the global human economy/resource production doubling time from 100,000s of years to 1000s of years. Industrial revolution dropped it from 1000s to 10s or even 1s. If AI follows the same path it becomes 0.1 - 0.01 years. Your 401k wouldn't need 40 years to build a comfortable retirement, only 4 weeks.

Assuming the supply of labor or automation sufficient to provide a “comfortable retirement” also takes 4 weeks to come online.

Re: Introducing ChatGPT and Whisper APIs

#155
post #9

>Through a series of system-wide optimizations, we’ve achieved 90% cost reduction for ChatGPT since December This is seriously impressive. A MILLION tokens for 2 dollars is absolutely fucking insane. I hope that the gains reached here can also be found by open source and non-controlled AI projects. If so, that could be huge for the advancement of AI.

It's tokens processed , not generated.

If you have 10K tokens in your conversation, the next reply means 10K + len(reply) extra tokens. I estimate 125 rounds of conversation fit in 1M tokens, for $2.

Re: Introducing ChatGPT and Whisper APIs

#156
I hope this pricing impacts ChatGPT+

$20 is equivalent to what, 10,000,000 tokens? At ~750 words/1k tokens, that’s 7.5 million words per month, or roughly 250,000 words per day, 10,416 words per hour, 173 words per minute, every minute, 24/7.

I uh, do not have that big of a utilization need. It’s kind of weird to vastly overpay

Re: Introducing ChatGPT and Whisper APIs

#157
post #90

Earlier quoted context omitted.

You are like 2 months out of date. Stable diffusion now has a massive ecosystem around it (civitai/automatic1111), that when used well, completely crushes any competitors in terms of the images it produces. Midjourney is still competitive, but mostly because its easier to use. Dalle2 will get you laughed out of the room in any ai art discussion.

Easier to use is often all that it takes. In Midjourney you get fantastic results just by using their discord and a text prompt. To get some similar results in Stable Diffusion you need to set it up, download the models, understand how the various moving parts work together, fiddle with the parameters, donwload specific models out of the hundreds (thousands?) available, iterate, iterate, iterate...

Why do you say that? Couldn’t you just use dreamstudio.ai?

Re: Introducing ChatGPT and Whisper APIs

#159

Earlier quoted context omitted.

You are like 2 months out of date. Stable diffusion now has a massive ecosystem around it (civitai/automatic1111), that when used well, completely crushes any competitors in terms of the images it produces. Midjourney is still competitive, but mostly because its easier to use. Dalle2 will get you laughed out of the room in any ai art discussion.

It's amazing that "being two months out of date" in AI means that you are already a dinosaur

ControlNet became popular with in the last couple of weeks and LoRA fine-tuning slightly before that and both things have completely changed the landscape too. Even a month out of date and you are a dinosaur at the moment.

Re: Introducing ChatGPT and Whisper APIs

#160
post #56

What is the developer experience of using OpenAI's Server-Sent Events endpoint from something else than their Python and Node.js libraries? The SSE endpoint is required for use cases like chat so the end user doesn't have to wait until the whole reply has been generated. I started implementing a simple SSE client on top of C#/.Net's HttpClient but it's harder than I first assumed.

We had to restream server events from openai -> our backend -> client. It was pretty simple.
Post reply on HN