Elon Musk announced the publication of the source code of Grok, his generative AI rival to ChatGPT. A choice of open source which implicitly allows the American entrepreneur to once again attack OpenAI, whose directions he does not like.
In the conflict between Elon Musk and OpenAI, there is the legal aspect. There is also the whole communication battle. At the beginning of March, the company founded by Sam Altman struck hard in this area by disclosing extracts of emails sent by Elon Musk. Objective: to expose its double discourse on the orientations of society.
Grok will go open source
This time, it is the turn of the American billionaire to seek to deal a blow to his former company, from which he walked away from 2018 due to differences of views. In a tweet published on March 11 on X (formerly Twitter), he announced the opening this week of the source code of Grok, his in-house chatbot.
In Elon Musk’s plans, Grok is seen as the rival to ChatGPT, which is OpenAI’s text-generative artificial intelligence. This project was made public in November 2023, four months after the launch of xAI, which is the company specializing in AI responsible for carrying out this project – and incidentally to compete with OpenAI and the rest of the industry.
Since then, Elon Musk has sought to promote Grok by suggesting that his new AI is much more objective, frank and accurate than other systems. In other words, Grok would tell the truth more and would also be able to answer delicate questions, without a “woke” filter – a subject that has become obsessive for him.
A tackle against OpenAI
The opening of Grok’s code (open source) is also another way to differentiate itself from ChatGPT, by implicitly recalling the distance operated by OpenAI in this area, according to Elon Musk. According to the entrepreneur, the company should even rename itself to ClosedAI to illustrate this “betrayal” of initial commitments.
This is also reflected in the complaint: “ Today, OpenAI’s website continues to state that its charter is to ensure that artificial intelligence ‘benefits all of humanity’. In reality, OpenAI has transformed itself into a closed-source subsidiary of the world’s largest technology company: Microsoft. »
These criticisms are not new. They were already issued four years ago, following the agreement between Microsoft and OpenAI. Since then, the person concerned has often taken a position against his company – by calling for a moratorium on the future of ChatGPT or by demanding a halt to the integration of ChatGPT into Bing, Microsoft’s search engine.
OpenAI had attracted criticism in the open source field, due to a reversal on its initial vision. “ We were wrong. We were totally wrong » For example, judged Ilya Sutskever, one of the co-founders of OpenAI and the chief scientist of the project, in March 2023, shortly after the announcement of GPT-4.
In the eyes of the other founders of ChatGPT, Elon Musk’s maneuvers above all reflect bitterness and resentment at no longer being in a company that is experiencing meteoric rise. One piece of evidence among others would be the insistence with which the person concerned seeks to point out that without him, there would have been no ChatGPT.
In the emails that the company published, Elon Musk appears to be maneuvering to try to take the reins of OpenAI in several ways, including by mobilizing Tesla, another company of the billionaire, dedicated to the electric automobile market. He even tried to put the financing of the company in the balance, without success.
Subscribe to Numerama on Google News so you don’t miss any news!