One simple reason that we remember scientists is that we use their names for things. We use Maxwell's equations, Ohm's law, Newtonian mechanics, Euler-Lagrange,

One simple reason that we remember scientists is that we use their names for things. We use Maxwell's equations, Ohm's law, Newtonian mechanics, Euler-Lagrange, the Einstein coefficients, Avogadro's number and so on. And then there are the SI units which are mostly named after people. Perhaps we'll enter a sci-fi scenario where we've forgotten why things are named the way they are. The general public? You learn from the high school syllabus, which has a finite amount of time and content. This is expensive educational real estate! That means you can learn the basics (mechanics, electricity, cell bio, etc), and some current state-of-the-art to add relevance and interest. Will the basics change in 1000 years? No idea. Classical equations "work" well enough to predict common phenomena, even if we come up with some unified theory in the future. Most of the people that we think of as "scientists" were post-Renaissance. Galileo onwards. Newton, Euler, Farady, Darwin, for example: these are new people - barely 500 years old at most. It's helped that we've been increasingly good about writing things down, and science has been heavily romanticised. Quantum mechanics and relativity are babies. A lot of recent discoveries, even Nobel winning science, go largely unnoticed because they're so niche technically that most people have no idea about them. Things like the giant magnetoresistive effect - absolutely world changing, but it's not trivial science and we're probably not going to be using magnetic media in 1000 years. What about algorithms and mathematics? There are algorithms which are provably correct methods for doing things, named after people (e.g. Dijkstra's algorithm). Are these likely to be forgotten by computer scientists? There are lots of extensions and optimisation tweaks, but the core remains the same.