Microsoft recently announced its plans to invest $ 1 billion in OpenAI, an emerging AI company co-founded by Elon Musk (which no longer participates) and focused on the development of artificial intelligence at the human level. Normally, this would be an article about robots that can think and how two of the biggest players in the AI industry join forces could lead to amazing things. But this deal is really creepy. So let's talk about that instead.
First, OpenAI was founded as a non-profit organization. Under CEO Sam Altman, the company recently reorganized to become what is essentially a for-profit business nested under a nonprofit organization with supervisory power. In other words, as Wired OpenAI can operate as a for-profit company, but it has to comply with the original statute of developing AGI that benefits all of humanity.
It's quite interesting, the company that OpenAI seems to be trying to catch up with this agreement is Google, another for-profit business once guided by the "don't be bad" principle. Although the motto did not work, Google’s current resources and talent dwindle. of almost all other AI adventures, including OpenAI. Enter Microsoft
The agreement looks like a marriage made in heaven: Microsoft has the coffers and the hardware, OpenAI has Ilya Sutskever and dozens of other researchers who would be the smartest people in most rooms. But, based on everything we've seen, this is not a joint research agreement, nor a development partnership, nor even a promise to work together on the same problems.
seems as an agreement in which OpenAI will develop all or part of its Azure technology so that Microsoft can sell, distribute or choose open source. In return, Microsoft will deliver OpenAI cash over the next decade to eventually equal around $ 1B, but expects to recover everything since OpenAI pays to use Azure or other computer services .
Lol. 19659002] This is what Microsoft's post on the agreement had to say:
Microsoft and OpenAI will jointly build new Azure AI supercomputing technologies. OpenAI will carry its services to run on Microsoft Azure, which it will use to create new artificial intelligence technologies and fulfill the promise of general artificial intelligence. Microsoft will become the preferred OpenAI partner to market new AI technologies
The OpenAI version was a bit different:
OpenAI is producing a sequence of increasingly powerful AI technologies, which requires a lot of capital to computing power The most obvious way to cover costs is to build a product, but that would mean changing our approach. Instead, we intend to license some of our pre-AGI technologies, with Microsoft becoming our preferred partner to market them.
And Microsoft CEO Satya Nadella has just made things more confusing. According to New York Times :
Mr. Nadella said Microsoft would not necessarily invest that trillion dollars at a time. It could be distributed over the course of a decade or more … it will be returned to its own business, since OpenAI acquires computing power.
What is really happening here? Who knows. The New York Times writer Cade Metz had an opinion that could explain it:
With the agreement, both OpenAI and Microsoft are looking for some public relations.
– Cade Metz (@CadeMetz) July 22, 2019
And AI expert Stephen Merity had an absolutely satisfactory thread on the subject in which he apparently points out that the only thing that keeps OpenAI in compliance With its non-profit open source roots is the fact that  promises not to act differently now that it is ready to market. Big tech promises huh, how much are they worth?
What is OpenAI?
A nonprofit organization that took advantage of goodwill while offering equitably for years preparing a change to a for-profit organization that now seeks to obtain closed technology licenses through a third party by segmenting technology under a banner of pre / post technology "AGI"? https://t.co/X6JIIuvcAd[19659002◆—Smerity(@Smerity) July 22, 2019
But, without a doubt, the most entertaining agreement that can be found on Twitter was three. – The interference of Google AI expert Francois Chollet (opinions are his own, according to his biography):
Many people in the AI community are confused by the OpenAI pivot of nonprofit to for purposes profit, his cult, public relations beyond the parody of "capturing the light of all future value in the universe" and his billion-dollar association with Azure … https://t.co/F6un60jlc5 [19659002◆—FrançoisChollet(@fchollet) July 24, 2019
And on top of that, you are now part, in the eyes of the world, of a techno-cult cult …
– François Chollet (@fchollet) July 24, 2019 Is OpenAI a techno-cult of the end of the world? Probably not. But all this meaningless, full of public relations, makes no sense of the kind of closed-end marketing junk that makes idealistic developers leave great technology to work in the academic world or in the nonprofit sector. For many OpenAI fans, this is like their favorite punk band selling and becoming used car dealers. I hope it is rather a techno-cult of the end of the world:
This is your best tweet.
– Tristan Greene (@ mrgreene1977) May 12, 2019