The last few years have seen an unprecedented growth in all the aspects of gravitational-wave (GW) astrophysics. At the frontier is population modeling of GW sources and their potential progenitors: massive stars in binary systems. Various models suggest that bulk of the population of binary black hole mergers have formed in metal-poor environments. This is primarily because low-metallicity stars experience reduced mass loss through winds, allowing them to form more massive black holes.
The pressing question is now: how well do we really understand the ''basics'' of stellar evolution and the ''classical'' formation scenarios in the regime of low metallicity? In my talk, I will challenge the long-standing picture of binary evolution, where a giant star is expected to be rapidly stripped of its envelope when engaged into mass transfer, arguing that this may no longer hold in metal-poor environments. Instead we predict that mass transfer in low-metallicity massive binaries is a less violent event, with a much longer interaction phase (nuclear vs thermal timescale), and often leading to stars that are only partially-stripped of their envelopes. I will discuss the implications of our findings for various binary evolution products (hot stripped stars, supernova progenitors, X-ray binaries, GW sources) and highlight the crucial role of observational constraints from nearby metal-poor dwarfs in the future.