Heat death? As we get a better understanding of the universe, each new step takes more and more resources. 150 years ago a single physicist or astronomer could on their own still make a world-changing discovery. Now, most major research is done by large groups working together. In the case of particle physics, the research involves literally the largest machines any humans have ever built. And the discoveries are getting smaller. Once one has the basic idea of evolution and DNA, and how RNA functions, in some sense what remains in biology, while quite interesting, just aren't as large or as amazing questions, even if their answers may have a lot of important applications and will continue to help us understand life. And one sees some of this issue also in where practical engineering has gone with science also- between 1885 and 1910 you have the first practical cars, the first radio and the first airplanes. People often like to say were in the midst of a technological revolution, but the turn of that century was far closer to that. Now, while we still have game-changing technologies, they aren't coming as fast. We may be moving into a long plateau.

We absolutely cannot decrease the entropy of a closed system. It's (almost) physically impossible. But it doesn't seem implausible that future engineers might be able to build machines that are efficient enough to approximate processes with zero net entropy increase. Perhaps they could get close enough for all intents and purposes? That's science. Heat death of the universe is what matches the data we have at this moment the best. We might get new data tomorrow.