The market has been awaiting Grok 2’s successor. While Musk remains confident, we must still ask – will it surprise us?
Elon Musk has been under the spotlight since the US elections. And this tech mogul seems unstoppable. He addressed the World Governments Summit in Dubai virtually. Musk had plenty to say regarding the next steps of xAI, OpenAI, and the US economy.
At the center of this discussion is Grok 3.
To fuel the anticipation, Musk said Grok 3 is the best chatbot amidst its competitors currently present in the market.
He was further questioned by the UAE AI Minister Omar Al Olama on what the techies and market competitors expect from this model. Musk described it as “scary smart” with strong reasoning abilities, leaving less room for doubts about its potential.
Musk also emphasizes that its base logic is better than Grok 2, its predecessor. Moreover, Grok 3 remains in the final refinement stage and could be released in another week or two.
His confidence concerning Grok 3 stems from tests that the company is currently undertaking. During this, the model was trained on ample synthetic data and additional data processing techniques to improve its “logical consistency.” If the data is incorrect, the model will rethink where it went wrong and remove it.
However, there’s more to what meets the eye.
This was refuted by an xAI engineer who ranked Grok 3’s coding abilities behind OpenAI’s o1-pro, o1, o3-mini, and other models in a social media post on X. For Benjamin De Karker, who was part of the human data team development for Grok 3, the unreleased AI model ranked fourth.
Later, he claimed to have been “threatened” by xAI to delete the post or face the consequences. Kraker continues to voice his disappointment and states he would rather resign to safeguard his dignity.
This has sparked a debate regarding free speech and openness. If so, what is the precedent that big organizations like xAI are setting?