A height gap in \(\mathrm{GL}_d(\overline{\mathbb{Q}})\) and almost laws (Q6924839)

This article is devoted to providing a more elementary proof of \textit{E. Breuillard}'s theorem [Ann. Math. (2) 174, No. 2, 1057--1110 (2011; Zbl 1243.11071)], a result that can be regarded as a non-abelian analogue of Lehmer's problem on the Mahler measure of an algebraic number. The theorem states that if a finite set \( F \subset \mathrm{GL}_d(\overline{\mathbb{Q}}) \) generates a group that is not virtually solvable, then its arithmetic (or normalized) height -- measuring its arithmetic complexity -- is bounded below by a positive constant \( \varepsilon_d > 0 \) depending only on \( d \) and independent of \( F \).\N\NSome consequences of Breuillard's theorem include the existence of a uniform lower bound on the exponential growth of non-virtually solvable subgroups of \( \mathrm{GL}_d(\mathbb{C}) \), a strong form of the classical Tits alternative ensuring the existence of free subgroups in linear groups, lower bounds on the girth of finite groups of Lie type and more recently, applications to the geometry of arithmetic locally symmetric spaces.\N\NThe proof builds on Breuillard's results on products of matrices while avoiding deeper tools from Diophantine geometry, such as \textit{Y. Bilu}'s equidistribution theorem [Duke Math. J. 89, No. 3, 465--476 (1997; Zbl 0918.11035)], \textit{S. Zhang}'s work on small points [J. Algebr. Geom. 4, No. 2, 281--300 (1995; Zbl 0861.14019)] and the geometry of Bruhat-Tits buildings. Its central idea is the use of word maps in compact Lie groups whose images lie arbitrarily close to the identity.