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

OpenAI and Microsoft extend partnership

openai.com

Post reply on HN
Print view View original

131–140 of 298 posts

Re: OpenAI and Microsoft extend partnership

#131

Post by adastra22 »

awestroke wrote: post #117
WastingMyTime89 wrote: post #85

Earlier quoted context omitted.

It is a for profit company in everything but name. That's my main complain. It has Musk, Thiel and Livingston amongst its initial investors, all known as the greatest philantropists of our time. /s I don't understand why they put this thin veneer on top of what they are doing. Maybe Thiel was burnt with the bad press surrounding Palantir and this is preventive damage control.

It's literally a nonprofit

No, it is profit-capped. And even then only on the same sense that the US government is debt-limited.

Re: OpenAI and Microsoft extend partnership

#132

Post by whimsicalism »

drexlspivey wrote: post #120
dorkwood wrote: post #43

Earlier quoted context omitted.

They keep it, but it goes to the non-profit arm of their business. “But any returns beyond that amount… are owned by the original OpenAI Nonprofit entity.” https://openai.com/blog/openai-lp/

> The fundamental idea of OpenAI LP is that investors and employees can get a capped return if we succeed at our mission Sorry I still don’t get it. If a private equity investor has shares and another investor wants to buy them off of him at 200x they can do that right? Are they obliged to give any excess returns to the non profit? Can’t they just sell the shares at 50x and then buy them back (perhaps through some ot…

My guess is it likely has to do with dividends.

But if your returns from the stocks are capped at 100x your share value, an efficient market would mean you share value never grows 200x.

Re: OpenAI and Microsoft extend partnership

#133

Post by mirker »

jacooper wrote: post #40

Microsoft is handling the AI revolution extremely well. They see the potential brand damage that comes with a such a product, so they probably just hold less than 50% of the company, to be able to say "this is not us". Which then allows openAI to experiment and wow people, while google is anxious about showing 10% of its crazy AI capabilities to avoid potential PR disasters.

Agreed. GitHub and OpenAI are the current branding. Though they did have some twitter bots go bad years ago and maybe they learned from that?

Re: OpenAI and Microsoft extend partnership

#134

Post by whimsicalism »

dandiep wrote: post #107
adpirz wrote: post #61

For people more well versed in this space: does GPT / OpenAI more broadly have a meaningful moat? It seems like there will be a number of these models out there and this isn't as great as say Google's up-till-now advantage in search relevancy.

A couple points which I don't see elsewhere: 1) They have the best quality model. Better quality means more users. More users means more data. Which means higher quality... 2) operationalizing & scaling these these models is non-trivial. I'm not sure what the state of distillation/pruning is for GPT-3, but I imagine they have figured out some proprietary techniques. 3) It's not just publishing a single model, but mak…

I don't view any of those things as a meaningful moat against the other companies with AI labs.

Specifically, training data is not primarily coming from interactions with model. While with RLHF this data might become more important, it is still a very small portion.

Re: OpenAI and Microsoft extend partnership

#135

Post by spyder »

davidy123 wrote: post #90
ThrowawayTestr wrote: post #24

2023 is the year of Bing, calling it now.

That really depends on how much the public accepts results without sources, aka no credit, aka rampant breaking of social norms and copyright, as well as destroying the easy ability to verify something. In many ways, OpenAI and anyone who supports it are trying to pull an Uber here, but shift the Overton on something indescribably larger than transportation licenses. They want to Borg global intelligence (though, of…

OpenAI already has a model to improve factual accuracy and provide citations:

https://openai.com/blog/webgpt/

It's probably not too hard for them to tune ChatGPT and the upcoming GPT4 that way, and I think it's very likely they will do something like that in Bing.

Re: OpenAI and Microsoft extend partnership

#136

Post by whimsicalism »

theropost wrote: post #124
adpirz wrote: post #61

For people more well versed in this space: does GPT / OpenAI more broadly have a meaningful moat? It seems like there will be a number of these models out there and this isn't as great as say Google's up-till-now advantage in search relevancy.

I recently wondered if one of the reasons for Google shutting down Stadia, was to quickly ramp their GPU server stockpile to redirect the resources at GPT modelling, to help catch up.

Google is not constrained by GPUs here and likely will train on TPU pods anyways.

Re: OpenAI and Microsoft extend partnership

#137

Post by whimsicalism »

mejutoco wrote: post #102
lmeyerov wrote: post #87

Earlier quoted context omitted.

Yes & No. GPT ChatGPT: probably not. It's not hard for other big players to enter this space. It's mostly egg-on-face for Google that they haven't given that Google basically invented the model that OpenAI uses and has big versions internally. There's nothing fundamental stopping Google Docs from adding ChatGPT to their UI and getting way more consumer training data than OpenAI can get without a similar play, or for…

> There's nothing fundamental stopping Google Docs from adding ChatGPT to their UI and getting way more consumer training data than OpenAI can get without a similar play. OpenAI could get exactly the same (or more, idk) data by integrating into Teams, considering the Microsoft partnership.

There are definitely more people using Docs than Teams.

I doubt that Microsoft will allow OpenAI to train on teams data from other businesses.

Re: OpenAI and Microsoft extend partnership

#138

Post by whimsicalism »

adpirz wrote: post #61

For people more well versed in this space: does GPT / OpenAI more broadly have a meaningful moat? It seems like there will be a number of these models out there and this isn't as great as say Google's up-till-now advantage in search relevancy.

There are a lot of finicky things that go into training a model as large as this.

But that knowledge will disperse and is already held in many competitor companies. I do not think that OpenAI has a substantial moat here.

Re: OpenAI and Microsoft extend partnership

#139

Post by ethanbond »

ProAm wrote: post #91
gen220 wrote: post #86

Earlier quoted context omitted.

If you want to develop a healthy counterpoint to that cynicism, you should consider reading more about Patagonia. Speaking as somebody who's generally cynical about these moves as well. I'm less read-up on OpenAI. It does feel to me like they've diluted the original non-profit/openness mission to the point of it being an interesting historical quirk, rather than an ongoing, guiding focus.

> you should consider reading more about Patagonia I'd be open to reading more. I think they are largely a good company but feel like this move was more of a tax dodge and to ensure generational wealth than for altruism. But I am open to being wrong about it. But everyone involved in OpenAI doesn't give me warm and fuzzy feelings at all. I admit it's largely cynicism until I know more. But there isnt enough time in t…

Company with decades of going way, way out of norms to operate as an ethical organization at virtually every level does an additional virtuous task: oh ya that’s a tax dodge

Re: OpenAI and Microsoft extend partnership

#140

Post by londons_explore »

Let me translate this blog post out of corpspeak...:

OpenAI has agreed to use Azure exclusively and for free, in return for Azure massively scaling up it's GPU capacity (which OpenAI was hitting before, hence threatening to move). In return, OpenAI gives an extensive license to Microsoft to use ChatGPT in Bing (and other products).

Now why couldn't the blog post have just said that rather than hiding all the details?

Post reply on HN