As chatbot competitors like DeepSeek continue to drive forward AI innovation, Elon Musk has revealed that the latest iteration of Grok AI is “scary smart”, and is able to outperform other leading AI chatbots.
Musk claims Grok 3’s synthetic training data method will be the key to its success. However, with the chatbot’s use still limited to X users, it’s uncertain whether its enhanced capacities will have any real impact on its market share.
The Tesla boss disclosed the news via a video conference at the World Government Summit in Dubai, in between taking jabs at his former colleague and AI sparring buddy, Sam Altman, and unveiling some major cost-cutting measures he was taking at DOGE.
Threat from DeepSeek AI
The Chinese AI chatbot maker DeepSeek has been dominating the headlines in recent months, largely due to its open-source large language model (LLM) which has the potential to democratize access to advanced AI technologies.
In what has been described as a “wake-up call” for the US tech industry by President Donald Trump, the development has left many chatbot makers and tech bros shaking in their boots.
This just in! View
the top business tech deals for 2025 👨💻
However, not one to be threatened by emerging developments in the field, xAI founder Elon Musk claims his upcoming chatbot, Grok 3, will be able to outsmart competitors like ChatGPT and DeepSeek, due to its superior reasoning capabilities.
In a session titled ‘Boring Cities, AI, and DOGE’ Musk revealed that Grok 3 is due to be released in a couple of weeks and that it’s “scary smart”.
“Grok 3 has very powerful reasoning capabilities, so in the tests that we’ve done thus far, Grok 3 is outperforming anything that’s been released, that we’re aware of, so that’s a good sign,” – Elon Musk at the World Governments Summit in Dubai
But what makes Grok’s latest model so smart? In the video conference, Musk claimed that Grok 3 was trained on synthetic data, and is capable of reflecting on mistakes by combing through the data to achieve logical consistency.
In contrast, leading US chatbots like Gemini and ChatGPT are trained predominantly on real-world data. This training method is understood to help the chatbots grasp human nuances and complexities better, but can open chatbot developers up to accusations of data theft, and could also compromise chatbot performance by limiting the amount of data a company has access to.
Elon Musk’s and Sam Altman’s Sparring Match Continues
Not one to miss up on an opportunity, Musk also seized the spotlight to take a stab at his AI adversary, OpenAI CEO Sam Altman. The billionaire criticized Altman’s management of OpenAI, comparing the Silicon Valley company to a nonprofit aimed at saving the Amazon rainforest becoming a “lumber company that chops down the trees”.
Musk has been openly critical of OpenAI’s decision to transition from a nonprofit to a for-profit model to support the high costs of developing AI and artificial general intelligence (AGI). A lawsuit the Tesla and X owner filed against the company may even be going to trial, according to a recent statement from a US federal judge.
How Will Grok 3 Compare to the Global Chatbot Competition?
As it stands, Grok AI is still a small player in the AI game. It’s significantly less popular than rivals like ChatGPT, which accounted for 62.5% of the market share for AI tools as of November 2024.
Grok does possess a number of attributes that separate itself from the competition, however. Its native X integration gives that chatbot access to real-time information from the social media platform, and its unique programming enables it to answer provocative prompts in a rebellious and cheeky tone.
Due to these USPs, Musk’s AI chatbot is popular among X users, and those who may find the responses from larger AI models a bit generic and trite.
However, the chatbot has frequently been embroiled in controversy, from responding to prompts with political disinformation to promoting biased content due to its access to X data. Rivals like ChatGPT and Gemini boast far more parameters also, making their responses generally much more accurate.
All in all, even with the use of synthetic training data, it’s unlikely that the soon-to-be-released Grok 3 will be able to hold a candle to bigger competitors. However, as foreign chatbot makers like DeepSeek continue to disrupt the global stage, committing to constant innovation is the only way that US AI developers will be able to sustain a competitive edge.
