In this seminar, we will explore algorithmic problems that lie at the edge of our current understanding—problems that admit surprisingly efficient algorithms (e.g., quasipolynomial-time) but for which no polynomial-time solution is known, and no strong hardness evidence rules one out.