Note on the number of hinges defined by a point set in \(\mathbb{R}^2\) (Q2658385)

Let be given \(n\) points in the Eucidean plane. A \textit{hinge} is a multiset of two distances measured among some 3 of the \(n\) point. There are at most \(3\binom{n}{3} \) distinct hinges. What is the minimum number of distinct hinges in sets of \(n\) points? This problem was first raised by \textit{A. Iosevich} and \textit{J. Passant} [Proc. Steklov Inst. Math. 303, 129--139 (2018; Zbl 1414.05101); translation in Tr. Mat. Inst. Steklova 303, 142--154 (2018)]. If the points make a \(\sqrt{n}\times \sqrt{n}\) grid, there are about \(cn^2/\log n\) hinges with some constant \(c>0\). The paper under review shows that the number of hinges is always at least \(cn^2/\log^3 n\), which only a \(\log^2 n\) factor off the grid construction mentioned. The proof heavily uses the methods of \textit{L. Guth} and \textit{N. H. Katz} [Ann. Math. (2) 181, No. 1, 155--190 (2015; Zbl 1310.52019)], which made a breakthrough on the Erdős distance problem.