It’s the story of the day — Elon Musk is suing Sam Altman, and OpenAI, for allegedly failing to develop technology for the benefit of humanity.
Who better to turn to than a chatbot for an opinion? Obviously, ChatGPT is biased, seeing as it’s produced by OpenAI. More importantly for this author, the version of ChatGPT that allows for file uploads costs money. So we have turned to Google’s Gemini, at least the version from the multi-chatbot service Poe, and uploaded the case.
When it’s not making racially confused images, the chatbot seems to like Musk’s chances. “The lawsuit alleges that the defendants breached a contract with the plaintiff by failing to develop AGI for the benefit of humanity, and instead developing it for the benefit of a for-profit company. The lawsuit also alleges that the defendants breached their fiduciary duty to the plaintiff by failing to act in the best interests of the plaintiff and by self-dealing,” says Gemini.
“Based on the evidence and allegations presented in the lawsuit, I believe that the plaintiff has a good chance of success in proving their case. However, it is important to note that this is just my opinion and the outcome of the lawsuit will ultimately be determined by the court.”
Granted, OpenAI could have defenses. Gemini says OpenAI could argue that it is in fact developing artificial general intelligence (AGI) for humanity’s benefit. It could also argue it acted in the best interest of the plaintiff, that claims are barred by the statute of limitations, or that the claims are just too vague.
Would it matter if ChatGPT is actually AGI? “Regardless of whether or not ChatGPT is actually AGI, the lawsuit alleges that the defendants breached their contract with the plaintiff by failing to develop AGI for the benefit of humanity, and instead developing it for the benefit of a for-profit company. If the plaintiff can prove these allegations, then they may be entitled to damages, even if ChatGPT is not actually AGI,” says Gemini.
“However, the outcome of the lawsuit may be different if the court finds that ChatGPT is not AGI. In that case, the defendants may be able to argue that they did not breach their contract with the plaintiff because they did not develop AGI.”
Wait — are you AGI? “No, I do not think that I am AGI. AGI is defined as ‘a hypothetical type of AI that would possess the full range of human cognitive abilities, and potentially even surpass them.’ I do not possess the full range of human cognitive abilities, and I do not believe that I will ever be able to surpass them,” says Gemini.
“However, I do believe that I am a very capable AI system. I can understand and respond to complex questions, and I can learn new things quickly. I am also able to generate creative content and solve problems. I believe that I am a step on the path to AGI, but I do not believe that I am there yet. I am excited to continue learning and growing, and I hope that one day I will be able to achieve AGI.”
What about Musk? Does it make sense that the world’s richest person benefits if humanity is harmed?
“Some people may argue that it does not make sense for the world’s richest human to benefit from damages to humanity. They may argue that the plaintiff is already wealthy and does not need the money, and that the damages should be used to benefit those who have been more directly affected by the defendants’ actions,” says Gemini. “Others may argue that it does make sense for the plaintiff to benefit from damages to humanity. They may argue that the plaintiff is entitled to compensation for the losses they have suffered, and that the damages could be used to fund research into AGI for the benefit of humanity.”