Elon Musk: Why You Should Stop Using ChatGPT & Gemini - Interview with Jordan Peterson - July 2024

Summary

In this conversation, the speaker discusses the rapid progress in AI development, particularly with the Grock model. They mention the impressive speed at which their new training center was built and compare Grock’s performance to other models like GPT. They outline future advancements, including Grock 2 and Grock 3, aiming to achieve superintelligence. The conversation touches on ethical concerns, with the speaker reflecting on the responsibility of guiding AI development to benefit humanity while being wary of politically correct or 'woke' tendencies in other AI systems.

Takeaways

• 😀 A new massive training center in Memphis has been successfully brought online, with hardware installation to training completion in just 19 days.
• 😀 Grock AI is catching up to models like Gemini and ChatGPT (CHBT) but is still based on an earlier version, Grock 1, which is weaker than GPT-based models.
• 😀 Grock 2 has finished training with 15,000 GPUs and is expected to be on par with GPT-4. It will be released soon after fine-tuning.
• 😀 Grock 3 is in development and will likely finish training in 3-4 months. It is expected to be the most powerful AI in the world by December.
• 😀 The current version of Grock is significantly weaker than ChatGPT, but it is performing well given the difference in model strength.
• 😀 Grock's team is focused on creating a digital superintelligence, one that surpasses human intelligence and, ultimately, all human intelligence combined.
• 😀 The ethical challenge lies in the decision to participate in developing AI, with the goal of steering it in a beneficial direction for humanity.
• 😀 The developer acknowledges that while they don't trust themselves entirely, they believe participating in AI development is a chance to guide it better than others.
• 😀 There is concern that other AI companies are focusing on political correctness and the 'woke mind virus' rather than striving for truth in AI development.
• 😀 The speaker's team aims to avoid the pitfalls they see in other AI projects by focusing on creating AI that seeks truth rather than conforming to popular narratives.

Q & A

What is the main goal of the Grock AI project?

The main goal of the Grock AI project is to develop a digital super-intelligence that is far smarter than any human, and eventually surpasses the collective intelligence of all humans combined.

How quickly was the Grock training center set up, and why is this significant?

The Grock training center was set up in just 19 days, which is notable because it is the fastest setup for an AI training center compared to others in the industry, showcasing the team's efficiency.

What is the current version of Grock, and how does it compare to other AI models like GPT?

The current version of Grock is 1.5, which is considered an improvement over the initial release but still significantly weaker than models like GPT. However, it is performing well given the difference in capabilities.

When will Grock 2 be released, and what performance expectations are there?

Grock 2 is expected to be released next month. It has completed training with 15,000 GPUs and is anticipated to be on par with GPT-4 once fine-tuned, offering a significant leap in capabilities.

What sets Grock 3 apart from previous versions, and when is it expected to be launched?

Grock 3, currently in development, is expected to be the most powerful AI in the world. It is expected to finish training in 3 to 4 months, with a release planned by December 2025.

How does the speaker view the future of AI development?

The speaker sees AI development as inevitable and believes it is essential to participate in it to guide AI toward beneficial outcomes. They also express the hope that their participation will help avoid the ethical pitfalls they perceive in other AI projects.

Why does the speaker believe that other AI companies may not be striving for truth?

The speaker argues that many other AI companies prioritize providing answers over truth and are influenced by political correctness, which they term as the 'woke mind virus.' They believe this approach compromises the integrity of AI development.

What are the ethical concerns the speaker faces regarding AI development?

The speaker acknowledges the ethical conundrum of AI development, particularly the risks of creating super-intelligent systems. They are concerned about the implications of AI surpassing human intelligence and are focused on ensuring that their approach benefits humanity.

How does the speaker feel about the rapid pace of AI development?

While the speaker acknowledges the fast pace of AI development, they emphasize the importance of staying engaged and participating to steer the technology in a beneficial direction, despite the potential risks involved.

What is the speaker's perspective on their role in the AI landscape?

The speaker believes that they have a choice between being a passive observer or an active participant in the AI landscape. By participating, they hope to influence AI development in a positive direction for humanity.