Microsoft closes AI Frontiers after AI reorg

Following a corporate reorganization that merged Microsoft's AI Frontiers research sub-organization into Microsoft AI, the author argues that the new leadership's exclusive focus on scaling rather than research is misdirected and incompatible with innovation. Consequently, the author and founding leadership members have left the organization, leading to the closure of AI Frontiers. The author contends that research yields substantial returns, such as significant training and inference cost savings, which make a scaling-only strategy short-sighted. Open article →