Do you comprehend that if it weren't for Microsoft OpenAI wouldn't exist, nor their products which changed the world?
You don't make AI in a garage with a home PC and beer. You need gigantic amount of data, compute, and talented people to organize all this, which are exceptionally rare, and also they need to eat and feed their families too.
Oh yeah, I do comprehend how such major technologies of comparable importance are managed in NON-PROFIT organizations, Linux Foundation is for one.
If you're happy to give up total control of tech to one corporation, while the CEO of the company, using the influence of the OpenAI, attracts resources for his personal hardware start-up[1], — our views on how a not-for-profit company should be run are starkly different.
You can work on and compile Linux on a home PC, even more the case with their other projects. Try training GPT-4 on a home PC.
The problem is you have no idea what you're talking about. This is not about how important something is, but what resources it requires as a BASIS MINIMUM for it to happen at all. Linux is important, but it's not nearly as expensive to create or maintain. If it's so easy for a scrappy non-profit to attract billions to tinker with AI, where I say are those startups? Where? Q.E.D. The most open thing we have, Llama, came from giant Meta.
Regarding the hardware startup, neither me nor anyone else is defending Altman about that, nor it has any relation to Microsoft. If you want to argue the subject of this thread, you're welcome. If you just want to copy paste generic talking points with no relevance, I don't care.
> You can work on and compile Linux on a home PC, even more the case with their other projects. Try training GPT-4 on a home PC.
Sam Altman was happy to join the venture and lead OpenAI knowing that it was the non-profit AI research company from the beginning. As a CEO, he was supposed to lead the company as it was created. As a result of his management, the company first ceased to be open, then in fact non-profit, then safe (according to it's leading scientists), then the number of published research whitepapers fell sharply, then has taken steps to become a gatekeeper for government regulations, and finally, the communication between Sam and the board (which really should be running the company) was lost. If a CEO was unable to manage a company, including attracting investment and resources, without violating the fundamental principles of that company, that CEO should have been dismissed — what has been effectively attempted by the board. Otherwise, it's a power grab and a de facto reformatting of the company into a different entity. Do you comprehend that?
I have no idea what compiling Linux kernel, or running GPT-4 locally, does anything in common in the context of large non-profit organization management. My argument is that the company should have been run in such a way that one corporation could not seize complete control. Never, under no circumstances. In spite of this, the CEO made all the effort to make it so. As a result, we have now in fact a for-profit company with good product, tons of money and GPUs, that belongs to the Microsoft. Was it worth it? Does this justify it for you? If you are not a $MSFT shareholder the answer should be obvious.
> If it's so easy for a scrappy non-profit to attract billions to tinker with AI, where I say are those startups? Where? Q.E.D. The most open thing we have, Llama, came from giant Meta.
Take a good look at HuggingFace Chatbot Arena leaderboard page[1]. Pay attention to the values of different models on different benchmarks. GPT4 is the leader, but not by an unattainable margin. Anthropic, which has spinned off from the very same OpenAI, rapidly catches up, along with other solutions. Even free, open (for commercial usage) models, while being self-hostable, are not an orders of magnitude worse. They are well in range of being practical and useful, including for commercial products. There is no moat[2].
> Linux is important, but it's not nearly as expensive to create or maintain.
> The problem is you have no idea what you're talking about.
The company was supposed to be funded by Elon who pledged a billion. Then he bailed when his coup failed. Sam Altman had to find a sponsor, or have the company go bankrupt in a matter of weeks.
What's hard to comprehend here?
As for the models topping your Hugging Face chart, they're literally all funded (by the billion) by private enterprises and the US govt. It's baffling to me how this is supposed to be an argument in your favor. Even the "not by an unattainable margin" comment makes no sense, unless you think getting from 40% to 60% is as easy as getting from 60% to 80%, and then getting from 80% to 100%. Which you clearly seem to think.
Anthropic's models are closed and funded by the money of Sam Bankman-Fried, of all people, and more recently Amazon. They're trying to raise billions more from Google and others. That's your idea of being open and not needing corporate investment for some reason? LOL, oh my god.
The "free" model you're citing is MADE BY META, a company worth almost a trillion dollars, and which has the largest social media presence in the world, from which to mine data.
You have absolutely not even the faintest idea what you're talking about.
You're either pretending or incapable of understanding the point that is being reiterated in this thread.
The OpenAI company, that was supposed to be non-profit open research think lab, has been taking investments of dozens of millions from different funds, while producing highest quality research material for the public good. Up to the point when the CEO sold it for Microsoft's $1B deal that solved their infra & GPU resource problems for… completely locking the company as an unofficial division of Microsoft Research. This deal, essentially, entirely distorted the raison d'être of the non-profit company, closed the "open research" part, and made it a for-profit extension for MS/Bing services.
Here, perspectives fork in two ways: either you believe it's a great development that you can benefit of by using closed paid OpenAI/Microsoft commercial services, OR you believe that metamorphosis of the leading open AI research shop into it's exact antipode is embodiment of disappointment. From your boorish, superficial remarks, I can see which side you are on.
I don't care if Sam Altman took money from Microsoft, or Sam Bankman-Fried (before he was charged), or US govt — that entirely misses the point. I do care that Sam Altman orchestrated a complete relegation of the research lab to an overpowered for-profit "Clippy" assistant for a single huge corporation, depriving us from the open research. He had to work on attracting investments from several corporations and other entities to avoid centralisation of control. But we see that he acted according to his own personal agenda, not in the interests of the company's charter. In that perspective, I don't give a damn if he could have raised a billion or not as a non-profit company. I'd rather it was $300-500M that would have kept the company open than a $1B and what we have today. I'd rather the company didn't survive this crisis than what we have today.
I gave you the benchmarking of other LLM models that are several months (up to a year) behind GPT4 in performance development. I don't care if they are founded by megacorps or not, if they do what non-profit "Open" AI no more does, i.e. publishing open research and supplying us open models. Comprehend? Your argument of "if it weren't for Microsoft OpenAI wouldn't exist" is irrelevant, if OAI no longer produces open research that it supposed to; it better wouldn't otherwise. Let others for-profit multi-billion corps take over, if we can't have better.
To further refute your $MSFT shill demagoguery, narrowly focused on compute accelerators, I'll add that Abu Dhabi's TTI (UAE gov money for academia) has produced and released open model Falcon 180B, while French nonprofit research lab Kyutai has just raised €330M. In Europe, for a minute, not even SV! Apparently, funding open research is possible? Q.E.D. I'd rather Ilya and Andrej move to this lab than peddling "Laundry buddy" GPT Bing API tokens BS disgrace, but it's up to them.
And you can stay in your delusions where only OpenAI+Microsoft is capable of leading AI research, lead by an ego inflated CEO. Don't forget to scan your eye pupils for $worldcoin.
OpenAI, the original non-profit, DIED. Elon Musk pulled the funding, it DIED. It's DEAD, get it? It does not exist without money. I don't know how to get it through your head. There's no such thing as a serious AI company without billions of funding. Period.
So. Given that, to keep the team together, Sam made the for-profit subsidiary and hosted it in the hollowed out non-profit, Sam went out to get investors and Microsoft chimed in the most.
While the new hybrid OpenAI tries to respect the spirit of the original OpenAI... WHICH IS DEAD (GET IT?)... it has to also balance between non-profit mission, and its investors, thanks to which it exists at all.
That's kind of like maybe you think you're a world-leading poet and novel writer, but your boss doesn't care, he wants those toilets clean. That's what he's paying for. You can then use that money to spend your free time writing poems.
Not sure I have to respect the rest of your BS with an answer. "Oh the other megacorp LLMs are just few months behind GPT4". Newsflash! OpenAI also exists within... time, genius. Time is passing there, too. And GPT-5 is ready and few Fortune 100 companies are testing it currently. They're not sitting still and waiting for the others to catch up.
I never said other companies can't compete, in general, but they're no better than OpenAI. Amazon, Google, Inflection, Facebook are no better than OpenAI. They're not. And I never mentioned anything about Worldcoin, you're basically desperate to strawman this, but you are exceptionally bad at it.
Did I say you don't have the slightest, faintest, most distant idea what you're talking about? Anyway this is my last message.
