Elon Musk has filed a lawsuit against OpenAI and its co-founders Sam Altman and Greg Brockman for allegedly violating the agreement they signed when the three men founded OpenAI.
The lawsuit provides an interesting insight into the origins of the corporate, which has turn into a market leader in the event of cutting-edge AI models. It is predicated on OpenAI's two principal goals, which were documented in the corporate's “Founding Agreement.”
The three men agreed that OpenAI could be “a nonprofit organization developing AGI for the good thing about humanity, slightly than a for-profit company in search of to maximise the profits of its shareholders.”
They also agreed that OpenAI “could be open-source, would only keep in mind countervailing security considerations, and wouldn’t keep its technology secret for proprietary business reasons.”
Musk feared, and still fears, that the profitable pursuit of AGI and the exclusive control of the technology by corporations like Google could be harmful to human society.
The idea behind the agreement was that OpenAI would create an open and secure AGI that may be freely available to humanity, eliminating the chance of for-profit corporations using the technology to the detriment of humanity.
The infringement
The undisputed leading AI model today is OpenAI's GPT-4, but there may be nothing remotely open about it, whatever the name of the corporate that developed it. Unlike GPT-3, OpenAI has not even published a technical document detailing how GPT-4 works.
While the founding agreement stated that the corporate would pursue open source technology, only OpenAI and certain Microsoft have access to GPT-4's code, weights and training data.
Running an organization like OpenAI costs money, so it's no surprise that it charges users fees to make use of its products. But the exclusive licensing agreement with Microsoft means its models at the moment are getting used to generate profits for Big Tech. Exactly the alternative of the founding principles that Musk says everyone agreed to.
The lawsuit claims that “OpenAI, Inc. has been transformed right into a closed-source, de facto subsidiary of the most important technology company on the planet: Microsoft.” Under its latest board, it is just not only evolving an AGI, but is definitely refining it “To maximize profits for Microsoft and never for the good thing about humanity.”
Do we’ve AGI?
Microsoft's license agreement applies only to OpenAI's pre-AGI technology. If an OpenAI model is found to have reached the AGI level, the license agreement prohibits Microsoft from accessing it.
Musk's lawsuit makes a daring claim by stating that “GPT-4 is an AGI algorithm” so Microsoft shouldn’t have access to it. The lawsuit asks the court to rule that that is true. I'm undecided how the judge and jury are presupposed to try this when there is no such thing as a industry-accepted definition of what exactly AGI is.
The lawsuit also confirms the rumors about what OpenAI's engineers have been working on. The lawsuit claims that “OpenAI is currently developing a model called Q* (Q Star) that has an excellent stronger claim to AGI.”
Ultimately, it’s the OpenAI Board that decides what counts as AGI under the Microsoft License Agreement. After Altman was fired, rehired and reassigned, that call now rests with the brand new board members appointed with Microsoft's blessing.
Was this why Ilya Sutskever was faraway from the board and kept quietly in some back room? Has anyone heard from him in months? What did Ilya see?
Don't expect the brand new board to rush to say that AGI has been achieved, even when that’s the corporate's explicitly stated goal.
AI fears
One of the stated reasons for the lawsuit is Musk's ongoing fears that AGI could fall into the fallacious hands.
The lawsuit recounts how Musk met with DeepMind co-founder Demis Hassabis in 2012 to debate these fears, a few of which Hassabis shared.
It also states: “After meeting with Mr. Hassabis and DeepMind investors, one in every of the investors remarked that one of the best thing he could have done for humanity was to shoot Mr. Hassabis on the spot.”
When Musk discussed his fears with then-Google CEO Larry Page, Page said that a super-intelligent AI that may replace humans was merely “the subsequent stage of evolution” and claimed that Musk was pro-human along with his views Views of a “specialist”.
Is Musk being overly dramatic or should we be apprehensive?
The lawsuit is value a read in the event you desire a behind-the-scenes have a look at how much of the industry got to where it’s today. However, it is just not likely that the case can be resolved any time soon.
OpenAI could well confirm Musk's AI fears and use this as a principal reason to not open source its models. For OpenAI's already busy legal team, representing the corporate's profit motive could also be a tougher task.
