He's saying that what is believed to be the truth at one point in time often ends up being false from another point in time. And this is inescapable since we never know as much as we think we do. In the late 19th century it was believed that physics was basically done, and all that remained was refinement to ever more decimal points. Then came along the early 20th century when quantum mechanics and relativity completely revolutionized the field and largely overturned stuff that had been believed to be true for centuries.

Science can do a decent job of disproving a hypothesis because even a single contradiction should be enough to suffice in good science, but it's far less efficient at proving anything true even if it seems to always be true. For instance mathematical relationship describing the gravitational attraction between large bodies seemed to always work, but it turns out it was merely a rough approximation that completely fails in various cases such as when one body has a particularly large gravitational pull, or when very high relative velocities are present. And even modern understanding is, at best, another rough approximation because we can already see endless examples in the cosmos of examples that defy current understanding and require further refinements in a direction that's currently unknown.

---

Basically at any point in history if you look at the bleeding edge science from a century before, it looks naive in many ways. In each era people always think they have finally moved beyond this, but we never have and it's entirely possible we never will since it's likely this universe has surprises awaiting us that we can't even yet imagine. Think about how utterly bizarre it is that time itself is a relative variable meaning with tech capable of reaching sufficiently high velocities you can literally travel into the future, relative to people at rest (such as all of Earth for example). It's nonsense, but it's completely real.