On: Google revives controversial cold-fusion experiments

[gridlockd]

> Will it work? Probably not. But that doesn't mean we shouldn't try.

The fact that it "probably will not work" (where "probably" = "almost certainly") is a strong argument for not doing something.

If I told you I had this idea for an over unity generator that "probably will not work", would you invest a million dollars into it?

Research is an investment, resources are extremely limited and they must be allocated prudently.