The pace of xAI operations is not just faster. Fundamentally structured around eliminating temporal and material constraints, it drives development at a speed unmatched in modern AI ecosystems.
In a recent interview on the Relentless podcast, xAI engineer Sulaiman Gholi provided rare and thought-provoking insight into the unique culture and operating philosophy driving Elon Musk’s latest venture, detailing how the company is achieving breakthrough results by prioritizing fundamental physics and radical autonomy over traditional corporate structure.
Gori’s first experience illustrates the extreme level of responsibility placed on new hires. What is expected is not a typical structured onboarding process, but rather an immediate demand for autonomy and influence. Here’s what he says about his first day: “The first day, they just handed me a laptop and a badge, and I was like, ‘Okay, now what do we do? We don’t even have a team, we’re not told what to do.'” This lack of bureaucratic overhead is intentional, designed to immediately filter out engineers who can identify and solve problems on their own. The result is a highly leveraged environment where ability dictates authority. The best part about working there, he says, is that “no one says no. If we have a good idea, we can usually implement it the same day and show it to Elon or someone and get an answer.” This culture avoids the multiple layers of approval common to established tech giants and accelerates decision-making to the speed of thought.
This relentless focus on speed extends directly to core technical challenges. Gori emphasizes that the xAI approach is rooted in first principles and constantly challenges established engineering assumptions. Many perceived limitations in software development, especially when it comes to latency and speed, are dismissed as “not true.” He argues that many existing stacks in the broader technology industry contain “a lot of stupid stuff” and that by optimizing and eliminating this overhead, engineers can typically achieve “2x to 8x” performance improvements on almost anything. Addressing fundamental efficiency is critical because xAI recognizes that the real bottleneck in achieving artificial general intelligence (AGI) is not software complexity, but the physical constraints of computation: energy and hardware availability.
This realization provides an almost insurmountable competitive advantage through xAI’s integration with the broader Musk ecosystem. Gholi details the strategic decision to leverage the computing power of Tesla vehicles to perform macro hardware simulations. He explains this challenge harshly. “We need a million human emulators. We need a million computers. What do we do?” The answer came by leveraging the hardware already installed in Tesla cars around the world. Gori pointed out that computers for Tesla cars are “actually much more capital efficient” than buying traditional cloud VMs or standard Nvidia chips. Access to a large, distributed network of powerful, capital-efficient computers allows xAI to scale its training and simulation efforts in ways that competitors relying solely on traditional cloud infrastructure cannot. This ability to integrate and leverage seemingly disparate enterprise-wide resources, from building data centers (Colossus) to processors in self-driving cars, is a critical physical moat that protects the explosive progress of xAI.
The pace of iteration through this infrastructure is incredible.
Ghori describes the model development cycle. In this cycle, new iterations are often deployed “every day, sometimes multiple times a day.” The financial impact of this velocity is significant, as calculated based on the immediate value added to core metrics. Ghori jokingly quantified the team’s recent contributions. “I just did the math, and I think it’s about $2.5 million per commit to the main repository. And today we made five.”
This calculation, however exaggerated, emphasizes the idea that every hour saved translates directly into significant value creation, justifying the intense work speed and constant pressure to achieve results. Entire organizations are geared towards rapid acceleration of core metrics, forcing engineers to ignore traditional solutions and achieve high utilization results immediately. Constant pressure allows the team to always focus on the most important path forward, operating on the principle that if tasks can be completed faster, no reason for delay will be tolerated.
