William Eckhardt's attack (Eckhardt 1993) on my account of the doomsday argument (Leslie 1992) attaches most weight to the familiar point that humans of the future are not yet alive to observe anything. The doomsday argument goes wrong, he suggests, when it asks us to treat ourselves as drawn randomly from among all humans who will ever have lived. There cannot, he says, be equiprobable sampling from an ensemble part of which currently exists. Hence the argument can establish only the triviality that doomsday cannot come earlier than the time at which we find ourselves. This strikes me as a faulty objection, One might as well argue that humans who are now dead not continue to observe things: they too do not currently exist. Dismissing both humans not alive yet and humans not alive still, we should have to conclude that any urn which the doomsday argument imagined ought really to be an urn filled with the names of those alive at this very moment. The same sort of thing could then be said about absolutely any anthropic arguments which used the observer's position in time. Since, it would be said, anyone now considering such arguments must be alive at this very moment, nothing interesting could be concluded. My article gave various grounds for rejecting the But future people aren't alive yet! line of reasoning. Perhaps the most powerful ground is that in thoughtexperiments such a line of reasoning leads to obviously unwise bets. Imagine, for instance, an experiment with the following plan, which you know was certain to be put into practice. In one year, three persons were to be given one emerald each. In some later year when the three had all died, five thousand people who had not been born when the first three emeralds were distributed would again each be given an emerald. Suppose you yourself get an emerald in the course of this experiment. Forced to bet on whether you are among the three or among the five thousand, you have nothing but those figures to guide you. You have an attractive reason, surely, for thinking you are among the five thousand. What if you instead said to yourself that there was a half chance that the five thousand were not alive yet? What if you then betted, your emerald against another of equal size, that you were among the three? If all the emerald-getters betted like this then admittedly there would at first be three happy winners, but eventually there would be five thousand sorrowful losers. Getting an emerald, you ought not to think it relevant that those as yet unborn are not yet observing anything. You ought to bet that you live at a time occupied by five thousand emerald-getters.