Elon Musk wants his own AI to “avoid the destruction of humanity”

In this regard, he compared the null desires of AI to exterminate humans with the way in which people protect chimpanzees. “We recognize that humanity could decide to hunt down all chimpanzees and kill them, but we are really glad they exist and we aspire to protect their habitats.”

He also took advantage of the interview to criticize OpenAI, the company of which he is the founder, since he said that at first it was a non-profit organization, but after his departure it became a company where profit incentives could interfere with the ethics of their machine learning models.

It should be remembered that the billionaire also invested $1 billion over several years, but later backed off his contributions when the startup secured an investment from Microsoft.

Musk’s interest in AI seems genuine, as a few days ago, the Financial Times reported that the businessman registered a corporation in the state of Nevada in March under the name X.AI, and of which he is the sole director.

The Financial Times also detailed that Musk has built a team of researchers and engineers, in addition to engaging in conversations with investors and recruiting staff from other AI companies.

On the other hand, he himself confirmed that he has acquired thousands of graphics cards (GPUs) from Nvidia, which are essential for building a natural language model, such as GPT, which powers the popular OpenAI tools.

Musk’s contradictions

Although generative AI is at its peak popularity, it is a development that has been widely criticized by Musk, who has said that it represents one of the greatest risks for the future of humanity.

In fact, at the end of March he signed a letter, along with dozens of technologists, requesting that research and work on these systems be halted, due to the “risks to society and humanity” they represent.

According to the document, the stage of development that AI systems are currently in is dizzying and “no one, not even their creators, can reliably understand, predict or control.”

In this context, they asked all AI laboratories to stop their work in this field for at least six months. “This pause must be public and verifiable”, can be read in the letter, which also highlights that if companies are not capable of doing so, then governments must intervene to regulate this matter.