Connie has just enough anti-venom to save one of the two poison victims before her. Now, faced with their pleading faces, but realizing it makes no difference to the total welfare, Connie finds herself totally uninterested in the question of who to save. It strikes her as no more normatively significant than the choice between a $20 bill or two tens.
To many -- myself included -- such indifference seems inappropriate. We think that which person survives is a matter of normative significance, so that Connie is (in her thoughts) making a kind of moral mistake.
I take this scenario to exemplify the worry that consequentialists see people instrumentally, as mere "receptacles" of value, or that they neglect the separateness of persons. Critics are assuming, in effect, that consequentialists must follow Connie in treating the welfare of distinct persons as a mere number, free-floating and fungible.
But once we realize that the fitting consequentialist agent would desire each good (separately), we can see the mistake in this way of thinking. The problem with Connie is that she doesn't appreciate that each individual's welfare is a distinct intrinsic good. She, in effect, only sees a single token good -- the aggregate welfare -- whereas a more plausible consequentialist view holds that the aggregate is merely an abstraction from a great plurality of distinct intrinsic goods (namely: each distinct person's welfare).
Since the fitting consequentialist will have distinct intrinsic desires for each person's welfare, they will not react with indifference, but rather ambivalence, when faced with tradeoffs like Connie's. They will be pulled in both directions, torn by the distinct importance of the two lives (only one of which can be saved), and whichever one they do save, they will still see something regrettable about the loss of the other.
In this way, the consequentialist can fully appreciate the separateness of persons. They make tradeoffs between lives, seeing that a greater benefit to one outweighs a lesser cost to another, but that does not entail seeing the two as fungible like money. For the benefit to one does not cancel the loss to another, which is instead seen as a unique and irreplaceable source of regret. But it is not as regrettable as it would have been to forsake the greater (and also unique) benefit to another.
In short: We can make tradeoffs between distinct intrinsic values, recognizing that some may be more important than others, without thereby turning them into merely instrumental values (fungible means to some further end of "aggregate" value). This is demonstrated by the distinction between indifference and ambivalence -- or, more generally, between tradeoffs where the cost is cancelled by the gain, vs. those where the cost remains distinctly regrettable, and is merely outweighed by the gain.
[For a more developed version of this argument, see my paper, 'Value Receptacles'.]