Torres is stupid.

Torres’s arguments often rely on conflation of what can be dubbed “reasonable” and “unreasonable” speculation. For example, to assume that (post)humans will one day colonize and perhaps terraform Mars is a reasonable speculation; to assume that (post)humans will one day develop “weapons that could destroy entire galaxies” (p. 82) is an unreasonable speculation. There are no indications whatsoever that the latter is physically possible. For instance, the binding energy of the stellar subsystem of the Milky Way is about –6 ´ 1050 J; if “destruction of the Galaxy” means unbinding all its stars, this would imply the hyperweapon capable of releasing at least that much energy, equivalent to completely converting more than 3,000 Solar-mass stars into usable energy (as per E = mc2 ).3 This is significantly more than the output of the most energetic known natural processes, including supernovae, gamma-ray bursts, and the most violent QSOs; even more importantly, we cannot envision – if our current astrophysics is correct and there is no bulk antimatter anywhere in the universe – the kind of energy-releasing process necessary for this task. Even if our distant posthuman descendants manage to discover and control such a process, the timescale for such a destruction would of necessity be measured in multiples of 107 or 108 years, which sounds a bit unreasonable in the tactical sense. And even if a military operation spanning 100,000,000 years or so could be palatable to a hypothetical future immortals, its efficiency would be doubtful, since not only would the adversary have enough time to prepare, but the unbinding might not automatically mean the destruction of adversary’s material and industrial capacities. (And similar objections could be posed to other construals of “destruction”, e.g., attempting to collapse all baryonic matter into the central supermassive black hole.)