Kevin Scott described Microsoft's partnership with OpenAI as a strategic "bet" on Sam Altman's company during an episode of Reid Hoffman's "Possible" podcast.
The Microsoft chief technology officer explained that collaborating with OpenAI would motivate Microsoft to "build better infrastructure." Scott, who has been Microsoft's CTO since 2017, acknowledged the acceleration of artificial intelligence progress early in his tenure and recognized the need for Microsoft to enhance its infrastructure, a thought that began in 2018.
In an email to Microsoft cofounder Bill Gates and CEO Satya Nadella, Scott expressed concerns about Google's AI capabilities, emphasizing the importance of advancing Microsoft's infrastructure. This email, recently disclosed as part of the Department of Justice's antitrust case against Google, was titled "Thoughts on OpenAI" and was sent just before Microsoft announced its $1 billion investment in OpenAI.
Scott explained that Microsoft couldn't do everything alone and saw OpenAI as a partner that could help forecast the benefits of investing in computing. He believed that by working with OpenAI, Microsoft would be pushed to improve its infrastructure while enabling OpenAI to excel.
Reports suggest that Microsoft is developing its own large language model, named MAI-1, with approximately 500 billion parameters, in competition with Google and OpenAI. Scott appeared to confirm this in a LinkedIn post, indicating that Microsoft occasionally needs to develop custom models, whether by training from scratch or fine-tuning existing models.
Microsoft has not yet responded to requests for comment regarding these developments.