One such atheist is Richard Carrier. He maintains that this principle is “not merely illogical, it's demonstrably false. We get an 'ought' from an 'is' all the time.”
Carrier's basic argument is that there are objective facts about what people desire, as well as objective facts about how to achieve these desires, and that “wherever both are an empirically demonstrated fact, the imperative they entail is [also] an empirically demonstrated fact.” So, for instance, consider the following example:
1. You want your car to run well
2. In order for your car to run well, its oil must be changed with sufficient regularity
3. Therefore, you ought to change your car's oil with sufficient regularity
Here, we have two “is” premises which (we can assume) state empirical facts, and an “ought” conclusion. And since the conclusion is supposedly derived from factual statements, Carrier believes it is “factually true independent of human opinion or belief.” He thinks he has solved the problem and that therefore the principle that one cannot derive an “ought” from an “is” should “never be uttered again.”
Well, sorry, but I'm going to utter it once again: One cannot derive an “ought” from an “is,” and it isn't hard to understand why. In deductive arguments, there cannot be anything in the conclusion that isn't already contained in the premises: All deduction does is “extract” some of the information included in the premises; it cannot add anything new. It follows that if the premises of a valid argument contain no “oughts,” then the conclusion cannot contain any “oughts” either (except in a trivial sense which is discussed in the note below).
And in fact it's easy to see what's wrong with Carrier's argument. That “you ought to change your car's oil with sufficient regularity” only follows if it is also the case that you ought to want your car to run well. In other words, there must already be an “ought” in the premises if the conclusion is to follow. Compare:
1. Manson wants to kill lots of innocent people
2. In order to kill lots of innocent people, one needs weapons
3. Therefore, Manson ought to get weapons
Is the conclusion here “factually true independent of human opinion or belief”? According to Carrier's argument, it is! But who seriously thinks that it is a moral truth that Manson should acquire weapons?
Nor is it difficult to come up with a counter-example to the first argument. Suppose Mary is an extreme environmentalist and disagrees that we ought to want our cars to run at all. Can we demonstrate to Mary the “fact” that you ought to change your car's oil regularly?
No one has ever successfully bridged the is-ought gap. If you want to maintain that there are objective moral truths, you ought to find some other way.
For anyone who might be interested, the trivial sense in which one can derive a conclusion that contains “ought” from premises that do not is as follows: Given that from p you can conclude p or q, one can validly argue something like,
1. The earth is round
2. Therefore, the earth is round or you ought to go fly a kite.
But this is irrelevant, as it is not a way to derive moral principles (the conclusion is true, but it does not actually say that you should go fly a kite).
The quotations are from Richard Carrier, “Moral Facts Naturally Exist (and Science Could Find Them),” in The End of Christianity, ed. John Loftus, pp. 334-335