
Learning from TikTok's Recommender and Trade-offs

The rise of taste, human authenticity and judgment in an AI world | Adam Mosseri (Head of IG)
Topic 7 of 9
Published Jul 9, 2026 · Clip 48:03–52:07 (4:04)
Adam expresses admiration for TikTok's recommendation engine, particularly its exploration-based ranking that can break small and niche talent by testing content users might not know they like, versus exploitation-based ranking. Instagram has invested heavily in catching up—optimizing not just engagement but originality, breakout content, and recency—and believes it has line of sight to best-in-class recommendations. He also frames all major platform debates as trade-offs: never seeing uninteresting content means only lowest-common-denominator stuff, and privacy vs. safety involves real trade-offs on both sides.