Musk vs. OpenAI – Another Hot Mess

OpenAI blows itself up yet again.

• In its rush to discredit Mr Musk’s lawsuit, OpenAI has revealed that it never had any real intention of being open and called attention to the fact that the company has strayed very far indeed from its original mission statement.
• The problem for OpenAI is that its image for honesty and its position as a philanthropic trailblazer has been shattered handing a large PR opportunity for its multitude of competitors to promote their products over OpenAI’s.
• Furthermore, developers will look at this and wonder if they can trust a company that has played fast and loose with the truth since almost the day it was founded.
• Mr Musk has a case but I suspect that he will lose because OpenAI’s statements are worded loosely enough to provide wiggle room for its lawyers but there is now little doubt that the company wilfully misled the market with regard to its intentions.
• This is clear from OpenAI’s actions over the last few years as well as the emails it has disclosed to rebut Mr Musk’s case.
  • First, Openness: In its launch statement (see here) OpenAI says “Researchers will be strongly encouraged to publish their work, whether as papers, blog posts, or code, and our patents (if any) will be shared with the world. We’ll freely collaborate with others across many institutions and expect to work with companies to research and deploy new technologies.”
  • However, just 3 weeks after the company was launched, it was already expressing doubts internally about how long it would remain open.
  • This is clear in an email from one of the founders to Mr Musk that OpenAI has made public in its defence.
  • This quickly came to pass as OpenAI is now clearly acting in contravention of its statements as GPT-4 is closed with no disclosure of its structure, how it was trained or what data was used to train it.
  • This makes sense complete sense commercially, but it gives an impression of being dishonest with the public.
  • Furthermore, in a May 2016 article in the New Yorker, there are allegations from a meeting between a Google researcher and OpenAI where the company clearly stated that it didn’t plan on releasing all of its source code.
  • This was followed by “But let’s please try not to correct that” from Mr Altman.
  • There are no legal ramifications here, but anyone who has signed up to OpenAI’s public vision will now feel like a complete fool.
  • Second, profit: In its launch statement (see here) OpenAI says “Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. Since our research is free from financial obligations, we can better focus on a positive human impact.”
  • OpenAI has now created a for-profit subsidiary which has created substantial internal conflicts within the company which have not been resolved despite the absurd events of late 2023.
• On top of this, the company remains full of conflicts of interest as well as conflicts between the profit and non-profit parts of the company that could spill over once again into warfare.
• A good example of one of these conflicts is the deal that Microsoft has with regard to its access to OpenAI’s IP.
• Microsoft has a perpetual license to all of OpenAI’s IP except IP which is considered to be artificial general intelligence (AGI) which is generally considered to be the point at which the machines become self-aware and surpass humans in almost all tasks.
• If this is ever created it will be one of the most valuable pieces of IP of all time, but it is the board (where Microsoft has a seat but not a vote) that decides what is AGI and what is not.
• In my opinion, this issue is moot because I think we remain as far away from AGI as ever but many in the AI industry will strongly disagree with me.
• However, it demonstrates just how easily quickly OpenAI could be at war with its biggest supplier and partner.
• Consequently, I think that OpenAI remains a time bomb waiting to go off and even if it beats Mr Musk, this will fix nothing.
• A proper implosion of OpenAI will damage Microsoft greatly and as a result, I think the risk of owning Microsoft shares is rising making them increasingly unattractive.
• I don’t own Microsoft and would have sold it a while ago if I did.