Did Elon Musk’s Tesla AI and Grok Crush OpenAI, and Is This the End of ChatGPT?

TL;DR

Elon Musk’s Grok and Tesla AI efforts challenge OpenAI, but the available evidence does not establish the end of ChatGPT. After leaving OpenAI’s board in February 2018, Musk launched xAI, whose 12-person team built Grok 1; Grok 2 followed on August 14, 2024, alongside Tesla’s Dojo supercomputer. Read on to see how Musk’s split with OpenAI shaped this rivalry.

Transcript

Rock 3. Let me tell you, it's insane. It's super fast, really smart and scary good at getting things right. I bet Chad GBT is sweating right now. Love him or hate him, Elon Musk sure knows how to grab attention when he launches something. And now he's teaming up with Donald Trump for something massive. We go by all the news announcements. Looks lik... Read More

Key Insights

• Elon Musk's AI journey began in 2014 when he became one of the first backers of DeepMind, an AI research company that Google acquired that same year, before he co-founded OpenAI in December 2015.
• OpenAI launched with funding drama from day one: Elon put in less than $45 million while other investors brought in over $90 million, and Elon pushed to claim a $1 billion funding commitment rather than the planned $100 million so they would not look hopeless.
• The core split between Musk and Sam Altman was about vision: Musk wanted AI to serve people and challenge big tech while retaining capitalist stakes, whereas Altman and Greg Brockman believed AI should ultimately belong to everyone.
• Musk stepped down from OpenAI's board in February 2018, citing conflicts with Tesla's own AI plans, after Tesla hired OpenAI's top researcher Andre Karpathy to lead autopilot and Musk suggested folding OpenAI into Tesla as a cash cow.
• xAI built and released Grok version 1 with just 12 team members after only about 2 months of training, demonstrating an unusually lean and fast development approach that XAI said would improve rapidly each week.
• Grok 2 dropped on August 14, 2024 with sharper reasoning and image generation aimed directly at GPT-4, alongside a smaller Grok 2 Mini for faster replies, taking xAI from zero to a legit ChatGPT rival in around a year.
• Neuralink is a tiny brain implant called the Link that uses thin electrodes to read and stimulate brain signals, aiming short-term to help paralyzed people control computers and prosthetics, and long-term to link human brains directly to advanced AI.
• Tesla built its own supercomputer called Dojo running on in-house D1 chips designed specifically for neural network training, avoiding the usual Nvidia GPUs and pushing self-driving vision AI in new directions.

Questions & Answers

Q: Is Elon Musk’s Grok AI the end of ChatGPT?

The available evidence does not show that Grok will end ChatGPT. It shows that xAI became a serious competitor by building Grok 1 with 12 team members and releasing Grok 2 with improved performance and image generation on August 14, 2024.

Q: How did Elon Musk first become involved in AI?

Musk’s AI involvement began in 2014 when he became one of DeepMind’s first backers, the same year Google acquired the company. In December 2015, he co-founded OpenAI as a nonprofit focused on building artificial general intelligence.

Q: How much money did Elon Musk contribute to OpenAI?

Musk contributed less than $45 million, while other investors provided over $90 million. He also proposed announcing a $1 billion funding commitment instead of the planned $100 million total so OpenAI would not appear underfunded.

Q: Why did Elon Musk and Sam Altman disagree at OpenAI?

They differed over OpenAI’s direction, ownership, and control. Musk wanted AI to benefit humanity while retaining a market role, and he sought majority ownership, substantial board influence, and the CEO position when a for-profit model was discussed; OpenAI’s leaders believed control by one person conflicted with the mission.

Q: Why did Elon Musk leave OpenAI?

Musk stepped down from OpenAI’s board in February 2018, citing a conflict with Tesla’s AI work. Tesla was investing heavily in self-driving technology and had hired OpenAI researcher Andre Karpathy to lead Autopilot, while Musk had also suggested folding OpenAI into Tesla.

Q: What is xAI, and how did it grow out of TruthGPT?

Musk announced plans for TruthGPT in April 2023, describing it as a maximum truth-seeking AI intended to understand the nature of the universe. That idea led to xAI, which aims to develop less-biased, reality-grounded, curious, and truth-seeking AI.

Q: How quickly did xAI develop Grok?

xAI built and released Grok 1 with a team of just 12 people after about two months of training. Grok 2 arrived on August 14, 2024 with stronger reasoning and image generation, alongside the smaller Grok 2 Mini for faster responses.

Q: How does Tesla’s Dojo supercomputer support its AI work?

Dojo is Tesla’s supercomputer for training its self-driving AI. It uses Tesla’s in-house D1 chips, which were designed specifically for neural-network training instead of relying on the usual Nvidia GPUs.

Summary & Key Takeaways

• Elon Musk's AI path started in 2014 backing DeepMind, then co-founding OpenAI in December 2015 as a nonprofit aimed at building AGI. From the start there was funding drama, with Musk contributing under $45 million and pushing to claim a $1 billion commitment rather than the planned $100 million total.

• Musk and Sam Altman diverged on vision, with Musk wanting AI to serve humanity while retaining market stakes and Altman favoring openness and shared ownership. Musk left OpenAI's board in February 2018 over Tesla conflicts, pulled a promised donation, and later publicly accused OpenAI of abandoning its mission and becoming a closed, Microsoft-controlled company.

• After signing a March 2023 letter urging a pause on AI beyond GPT-4 and floating TruthGPT, Musk launched xAI, which built Grok 1 with just 12 people. Grok 2 arrived August 14, 2024 with image generation, followed by Grok 3, alongside Neuralink brain implants and Tesla's Dojo supercomputer running custom D1 chips.