Something that a lot of people don't realize is that dead ends have value. It is only when you start building that you know how workable something is, how efficient. In order to build only the best possible thing, we would need precognitive abilities that violate casual relationships. So we have to actually do the math. When a project fails because some other project worked better, they are both part of the same overarching project, carried out by the whole civilization, to investigate multiple possibilities. And one was found to be better. We're going to have a lot more dead ends on our road to the stars, immortality, and the tech singularity. We have to learn not to regard that money and time as wasted, so long as we are willing to change course when the dead end becomes obvious. Research is, ultimately, search.