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

Microsoft eyes $10B bet on ChatGPT

semafor.com

Post reply on HN
Print view View original

381–390 of 541 posts

Re: Microsoft eyes $10B bet on ChatGPT

#381

Post by Q6T46nT668w6i3m »

CJefferson wrote: post #346
mannerheim wrote: post #258

Earlier quoted context omitted.

AlphaGo and AlphaZero are Google/Deepmind, not OpenAI. They've released papers, and there's a model that reproduces the paper. https://en.wikipedia.org/wiki/Leela_Zero

Firstly, you are right about deepmind vs openai, woops. While they released papers, I personally wouldn't have been happy reviewing the paper, because it was basically impossible for anyone to check (given how long Leela Zero took to get up to the same level). Releasing the network via bittorrent would have been very low cost. I am impressed by Leela Zero, and independant duplication is still important, but it would…

DeepMind has an excellent track record of helping other researchers reproduce their work. OpenAI does not.

Re: Microsoft eyes $10B bet on ChatGPT

#382

Post by justapassenger »

mannerheim wrote: post #341
londons_explore wrote: post #327

Earlier quoted context omitted.

Nearly everything openAI has showcased at least has a publicly published paper on how to reproduce what they have done. Sure, it isn't as good as source code and model weights... But having a paper probably reduces by 90% the time+money expense to reproduce the work.

Depends on what models we're talking about. I've heard reproducing ChatGPT would require on the order of ten million dollars' of compute.

That's cheap for an AI model training at big tech. Your average ads model engineer at those companies likely uses more compute each quarter.

It mostly suggests that we're in another AI hype bubble. MS and other big tech companies can easily replicate OpenAI results and do same type of research.

Re: Microsoft eyes $10B bet on ChatGPT

#383

Post by Q6T46nT668w6i3m »

la64710 wrote: post #153

Question to AI/ML folks : Is there no comparable open source model? Is the future going to be controlled by big corporations who own the models themselves? If models are so computationally intensive to produce does it mean than the more computational power a company has the better its models will be?

It depends what you mean by “open source.” You can find countless open source implementations. The issue is training. Meta has released their LLM weights. I believe Google has too.

Re: Microsoft eyes $10B bet on ChatGPT

#384

Post by adrr »

World177 wrote: post #377

> Additionally, a structure with the ownership of OpenAI would be put in place. Microsoft would entail a 49 percent stake while other investors would take over the other 49 percent. The remaining 2 percent will reportedly go to OpenAI’s non-profit parent firm. When OpenAI was funded as a non-profit in 2015, it raised $1 billion from investors that included YC Research. [1] Sam Altman was also the former president of…

Was the $1b donated to the non-profit? That’s what I would want to know. Every time I hear non profit, I just assume it’s for tax avoidance.

Re: Microsoft eyes $10B bet on ChatGPT

#385

Post by localhost »

orra wrote: post #349
localhost wrote: post #238

Earlier quoted context omitted.

RE: Open Source models, there is the AnthropicAI model making the rounds on Twitter[1] and Stability.ai (makers of Stable Diffusion) are working on one [2]. If we use recent history as an example, OpenAI announced DALL-E in Jan 5, 2021 [3], announced v2 and a waitlist for public use in July 20, 2022, and Stable Diffusion shipped an open source model on August 22, 2022 [4] using ~$600K of compute (at retail prices on…

Stable Diffusion's code and model weights can be downloaded, but it's proprietary, not Open Source.

This is the text of the license from this repo [1]. Seems pretty open to me?

About this license

The Responsible AI License allows users to take advantage of the model in a wide range of settings (including free use and redistribution) as long as they respect the specific use case restrictions outlined, which correspond to model applications the licensor deems ill-suited for the model or are likely to cause harm.

[1] https://huggingface.co/CompVis/stable-diffusion-v-1-4-origin...

Re: Microsoft eyes $10B bet on ChatGPT

#387

Post by mannerheim »

justapassenger wrote: post #382
mannerheim wrote: post #341

Earlier quoted context omitted.

Depends on what models we're talking about. I've heard reproducing ChatGPT would require on the order of ten million dollars' of compute.

That's cheap for an AI model training at big tech. Your average ads model engineer at those companies likely uses more compute each quarter. It mostly suggests that we're in another AI hype bubble. MS and other big tech companies can easily replicate OpenAI results and do same type of research.

Well, there's not only monetary cost on compute, you need humans interacting with it for RLHF, it takes time as well, you need engineers, and at the end of it, if you're just copying the leader, you're always going to be behind and not going to be coming up with the latest and greatest.

I'm just disappointed the compute cost alone seems difficult to surmount for open source/community projects.

Re: Microsoft eyes $10B bet on ChatGPT

#388

Post by esperent »

swagmoney1606 wrote: post #379
World177 wrote: post #377

> Additionally, a structure with the ownership of OpenAI would be put in place. Microsoft would entail a 49 percent stake while other investors would take over the other 49 percent. The remaining 2 percent will reportedly go to OpenAI’s non-profit parent firm. When OpenAI was funded as a non-profit in 2015, it raised $1 billion from investors that included YC Research. [1] Sam Altman was also the former president of…

One thing that really bothers me is how much of a contradiction "OpenAI" is. There's virtually nothing open about it.

It's a good reminder that names don't mean anything. Words don't mean anything ("don't be evil" comes to mind).

Only actions matter.

Re: Microsoft eyes $10B bet on ChatGPT

#389

Post by impulser_ »

moneywoes wrote: post #187
s_dev wrote: post #181

Earlier quoted context omitted.

Nah -- this sounds like Google buying YouTube or Android for billions while they were in their infancy but had proven tech. People did say at the time Google overpaid but they were some of the wisest investments Google ever made. I hope Microsoft don't buy them and OpenAI just becomes another FANG.

The google acquisition is a great comparison. Years from now it will look like a steal

No it's not. YouTube had a working product with a working business model.

OpenAI has neither.

Re: Microsoft eyes $10B bet on ChatGPT

#390

Post by nimbius »

swagmoney1606 wrote: post #379
World177 wrote: post #377

> Additionally, a structure with the ownership of OpenAI would be put in place. Microsoft would entail a 49 percent stake while other investors would take over the other 49 percent. The remaining 2 percent will reportedly go to OpenAI’s non-profit parent firm. When OpenAI was funded as a non-profit in 2015, it raised $1 billion from investors that included YC Research. [1] Sam Altman was also the former president of…

One thing that really bothers me is how much of a contradiction "OpenAI" is. There's virtually nothing open about it.

well clearly it was open for business.
Post reply on HN