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

Microsoft and OpenAI end their exclusive and revenue-sharing deal

bloomberg.com

Post reply on HN
Print view View original

881–890 of 915 posts

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#881

Post by bigyabai »

scottjg wrote: post #726
AuthAuth wrote: post #611

Earlier quoted context omitted.

>the M1 series GPU using 120W could beat an RTX 3090 using 420W You're cooked if you actually believe this

I very recently ran the numbers on these GPUs for an upcoming blog post. The token generation performance is bad, but the prefill performance is _really_ bad. For a Qwen 3.6 35B / 3B MoE, 4-bit quant: - parsing a 4k prompt on a M4 Macbook Air takes 17 seconds before generating a single token. - on an M4 Max Mac Studio it's faster at 2.3 seconds - on an RTX 5090, it's 142ms. RTX 5090 uses more power than an M4 Max Mac…

That's just a 4k context too. At a realistic context window of 16-32k tokens, the comparison becomes downright unfair.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#882

Post by hirako2000 »

whatever1 wrote: post #859
hirako2000 wrote: post #430

Earlier quoted context omitted.

A company can dilute just like that?

Of course. Public companies do this too. Where do you think the stocks awards that they give to employees come from? They come from just printing more shares every quarter, diluting every shareholder.

I was told stock to employees are left aside shares, or they sell the stock the company owns if they already went public.

I'm not convinced what you believe is true. Dilution is possible undoubtedly, but perhaps if majority shareholders approve it. And even then, likely regulated by a suite or other constraints.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#883

Post by fuck_google »

square_usual wrote: post #855
mlmonkey wrote: post #854

Earlier quoted context omitted.

This is what I've heard on the "street". Building a CUDA-compatible stack for AMD's hardware requires highly-paid SWEs. It's a very niche field, and talent is hard to come by. But AMD does not want to pay these specialized SWEs the market rate. Their existing SWEs would be up in arms saying, basically, "what are we, chopped liver??", or so the thinking goes. So AMD is stuck with a shitty software stack which cannot c…

> highly-laid SWEs Freudian slip?

[dead]

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#884

Post by bigyabai »

munk-a wrote: post #640
bigyabai wrote: post #303

Earlier quoted context omitted.

Apple is basically in the same boat as AMD and Intel. They have a weak, raster-focused GPU architecture that doesn't scale to 100B+ inference workloads and especially struggles with large context prefill. TPUs smoke them on inference, and Nvidia hardware is far-and-away more efficient for training.

Apple is in a much better boat than AMD or Intel. They have a gigantic warchest and can just snap up whoever looks like a leader coming out of the bubble burst.

Counterpoint: Apple's opportunity to invest in GPGPU architectures was ~2017 when Apple Silicon was in it's design stages. Apple always knew they were surrendering a large market segment by depreciating CUDA and OpenCL, they just never knew how big the market segment would get. Their liquid cash is wasted mid-bubble, and waiting for it to pop is not a realistic timeline anymore.

Arguably, Apple isn't even in a boat right now. At least AMD and Intel both ship hardware that synergizes with CUDA - Apple jumped off that ship, their hardware doesn't even come up in infrastructure discussions where AMD, Intel and Nvidia are taken for granted.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#885

Post by whatever1 »

hirako2000 wrote: post #882
whatever1 wrote: post #859

Earlier quoted context omitted.

Of course. Public companies do this too. Where do you think the stocks awards that they give to employees come from? They come from just printing more shares every quarter, diluting every shareholder.

I was told stock to employees are left aside shares, or they sell the stock the company owns if they already went public. I'm not convinced what you believe is true. Dilution is possible undoubtedly, but perhaps if majority shareholders approve it. And even then, likely regulated by a suite or other constraints.

Yep every quarter the Board has to approve it. Check the financial statement of AMZN for example.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#886

Post by bigyabai »

zozbot234 wrote: post #791
bigyabai wrote: post #773

Earlier quoted context omitted.

Not even a cluster of Mac Pros could run a dense 5T parameter model with RDMA, to my knowledge.

SOTA models are reportedly MoE, not dense.

A 5T MoE model is still bottlenecked by streaming weights from SSD, in addition to compute bottlenecks during prefill and decode.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#887

Post by mrandish »

everforward wrote: post #403
lesuorac wrote: post #318

Earlier quoted context omitted.

Why? FaceBook largely requires an Apple iPhone, Apple computer, "Microsoft" computer, "Google" phone, or a "Google" computer to use it. At any point one of those companies could cut FaceBook off (ex. [1]). The Metaverse was a long term goal to get people onto a device (Occulus) that Meta controlled. While I think an AR device is much more useful than VR; I'm not convinced that it's a mistake for Meta to peruse not be…

I think this is sane washing their idea in the modern context of it having failed. I think at the time, they thought VR would be the next big thing and wanted to become the dominant player via first mover advantage. The headsets don’t really make sense to me in the way you’re describing. Phones are omnipresent because it’s a thing you always just have on you. Headsets are large enough that it’s a conscious choice to…

> the web interface is like right there staring at them.

True but the an app gives Facebook much more user data for targeting which dramatically increases revenue per ad. Persistent user data that's largely unconstrained by privacy safeguards is the holy grail. The mobile browsers are also controlled by Apple and Google, so despite the web being 'open', when one of them makes even minor changes to increase browser privacy defaults, it can have major impact on Facebook's revenue.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#888

Post by measurablefunc »

fsniper wrote: post #793
measurablefunc wrote: post #389

Earlier quoted context omitted.

People want to believe in magic so they will find excuses to do so. Computers have been proving theorems for a long time now but Isabelle/HOL didn't have the marketing budget of OpenAI so people didn't care. Now that Sam Altman is doing the marketing people all of a sudden care about proving theorems.

Isabelle/HOL (a specialized software to do math proofs) doing proofs is not the analogue to LLMs (with the common accepted degeratory description: automated plagiarism machine) being capable of doing proofs. It's not the marketing, it's what the intention and the capability matrix is coming up to. I would be excited the same when Isabelle/HOL writes poetry.

Like I said, you want to believe in magic & will find any excuse to do so b/c you don't really understand how computers actually work. Good luck.

Re: Microsoft and OpenAI end their exclusive and revenue-sharing deal

#890

Post by marcellus23 »

wg0 wrote: post #729
marcellus23 wrote: post #678

Earlier quoted context omitted.

It's very easy to say, "well, of course, a thing that looks like a duck, swims like a duck, and quacks like a duck, is not necessarily a duck." But when you're presented with something indistinguishable from a duck in every way, how do you determine whether it's a duck? You can't just say "well I know it's not a duck". It's dodging the question.

Well. AI doesn't walk or quack like a duck. Ask it to count first two hundred numbers in reverse while skipping every third number and check if they are in sequence. Check the car wash examples on YouTube.

If I picked a human off the street and asked them to "count first two hundred numbers in reverse while skipping every third number and check if they are in sequence", I bet most would screw up.

my point is not that current LLMs are sentient, or even that LLMs ever could be. My point is that it's very difficult to come up with a way to test consciousness, and it makes me a bit nervous to see people suggesting that something could never be conscious just because it's technological and not biological.

Post reply on HN