Elon Musk utilized a personal phone call to Microsoft CEO Satya Nadelal to secure discounted computing resources for OpenAI, according to a report by PC Gamer.
This access to Microsoft's infrastructure played a critical role in OpenAI's ability to train models capable of defeating human professionals in the complex strategy game Dota 2.
During the early stages of OpenAI's development, the computing power required for large-scale machine learning was a significant barrier to entry.
Musk's direct intervention with Nadella provided the necessary hardware scale to advance the project's competitive gaming benchmarks.
A strategic hardware advantage
The report from PC Gamer highlights how this specific arrangement of discounted cloud resources helped facilitate the milestone where OpenAI's AI agent outperformed human players.
This period of development predated the current era of massive-scale LLM training, but the reliance on high-end compute remains a central theme in AI progress.
Musk's ability to bypass standard procurement processes via a personal call to the Microsoft executive ensured that OpenAI had the computational headroom needed for intensive training cycles.
While OpenAI has since transitioned into a different operational model, the foundational success of their gaming AI was built on this early access to Microsoft's Azure capabilities.
