Introduction
When asked about your knowledge of Elon Musk, you’d likely recall his involvement with the Trump administration, Tesla cars, and SpaceX’s mission to Mars. But would you mention his artificial intelligence (AI) work? Probably not.
Elon Musk’s AI Journey
Musk, one of the pioneers in generative AI, might not be the most popular figure in this field, but he’s certainly one of the most controversial. Will his future endeavors take off? Is there something promising worth waiting for?
Musk’s AI journey began in 2015 when he, along with Sam Altman and others, co-founded OpenAI—a nonprofit research laboratory aiming to put AI at the service of all humanity. While funding OpenAI, Musk took on a role as an “AI danger prophet,” leading to his public clash with Facebook founder Mark Zuckerberg in 2017. Zuckerberg saw AI as a positive force, while Musk envisioned it as a potential world threat.
As is common with Musk’s projects, drama soon ensued. He left OpenAI in 2018, believing it was betraying its ideals by seeking lucrative deals with companies like Microsoft to compete against Google’s AlphaGo program.
The Birth of Grok
On July 12, 2023, Musk announced his proposal: xAI, an AI company dedicated to “understanding the universe.” He claimed it wouldn’t just innovate but present an AI that’s “maximally truth-seeking” and free from restrictions imposed on other models.
His chat AI, Grok, was integrated directly into the X social network (formerly Twitter). Initially available to premium subscribers and later in the free app, Grok claims to be “the most intelligent AI in the world,” striving for maximum truth, utility, and a touch of rebellious humor while avoiding excessive censorship common in other models.
Grok is reportedly trained on unique data (real-time X posts from over 500 million users) based on a proprietary language model, Grok-1. In 2024, it added Grok-1.5V for text and image integration, followed by Grok-2 with the promise of an upcoming Grok-3.
Despite Musk’s claim of competing directly with ChatGPT-4, Claude, or Gemini, experts place him several steps behind due to flaws in his reasoning. However, some have praised his skills in generating code and his initiative of opening parts of his code to experts.
Is it Worth It?
Grok functions similarly to other AIs, being quick, providing distinct information, and challenging users at the end of each response to ask something “rebellious.” I appreciate that it doesn’t generate “perfect” texts like university assignments or encyclopedia entries.
I tested Grok by asking its opinion on Donald Trump’s government (something other models like ChatGPT and Gemini typically avoid, stating they lack updated or available information). Grok provided numerous entertaining details, as if reading a critical blogger. “Oh, the orange elephant in the room!” it responded.
I then asked about Mexico’s former President Andrés Manuel López Obrador, and Grok began with, “Well, we’ve moved from the orange elephant to an eagle devouring a snake, with mañaneras included…” It did provide additional information.
Would I switch to another AI? Not at all! But I’ll keep Grok for some light entertainment and to find “alternative” information. However, like other models, I wouldn’t fully trust its data (it’s based on user tweets that aren’t necessarily verified).
Grok has been criticized for its political stance, raising doubts about its ideological bias (lacking objectivity). Some say conversing with this AI would be like talking to Musk: in the same ironic, confrontational, and unpredictable tone, making it entertaining but unreliable for me.
Grok’s biggest disadvantage is its still low user base. According to VisualCapitalist figures, by July 2025, ChatGPT (OpenAI) held 82.7% of the generative AI market; Perplexity, 8.2%; Copilot (Microsoft), 4.5%; Gemini (Google), 2.2%; DeepSeek (Chinese AI), 1.5%; Claude (Anthropic), 0.9%; and all other models, including Llama (Mark Zuckerberg’s AI) and Grok (Musk), held less than 1%.
Despite this, Musk doesn’t seem ready to give up soon. In February 2025, he attempted a masterstroke: partnering with his former “enemy,” Zuckerberg, to acquire OpenAI. The deal didn’t materialize.
A Misunderstood AI?
Musk plans to integrate Grok into Tesla for better understanding of driver intentions and in SpaceX for autonomous space missions. His vision is to create an interconnected ecosystem using non-generative AI (not available to users).
How will he achieve this? xAI has recruited former OpenAI and Google DeepMind engineers. In 2024, it raised $6 billion in private investment, demonstrating that despite not being the most popular AI, its founder remains magnetic to investors.
Though not my favorite tech figure, I’m eager to see what the world’s richest man will do to “not give up” on AI, as he never needed to be first to win. When Tesla was launched, the electric car market was nearly nonexistent; today it’s worth more than Ford and GM combined. When SpaceX was founded, rocket technology was exclusive to governments…
In his AI pursuit, who will Musk ally with? Who will he challenge? Will he create the AI that pilots rockets, drives cars, and moderates social media debates? We’ll have to wait and see…
