🤔🤔🤔

Let's break it down.

This is about the lawsuit in which Musk claims that OpenAI violated its original mission, had no right to abandon its non-profit structure, and owes him a lot of money (he calculated $134 billion) for betrayed expectations, among other things.

The crux is this: if Musk can prove that GPT-4o is AGI, he will thereby prove that OpenAI concealed AGI from the public and illegally profited from it.

Specifically, if the court suddenly recognizes this fact, OpenAI's licensing agreement with Microsoft would become void (it only applies to pre-AGI, not AGI technologies), and OpenAI's founding agreement would require them to make their models open and freely accessible.

And apparently, they would owe Musk money because they allegedly deliberately failed to acknowledge AGI. Musk claims this is the reason the board fired Altman in 2023, and simultaneously the reason the board was replaced with a commercial one: so that members would have no financial incentive to declare AGI due to their own financial interests.

But why specifically GPT-4o, and not GPT-5, o3, etc.?

First, it came out earlier. Second, the leap compared to previous models was indeed enormous, and Musk cites a Microsoft article titled Sparks of AGI as evidence.

Third, because of deep empathy. This is what GPT-4o was remembered for by users, and this is what subsequent generations cannot replicate. According to Musk, OpenAI deliberately deprived GPT-4o's successors of this ability and took the model out of service despite petitions and user requests.

The next court hearing is scheduled for late April; the investigation continues.