Albert Einstein—creator of relativity, godfather of quantum physics, bender of space and time—had a little problem that dogged him all his career: lack of vision.

It may seem an unlikely charge to levy against the greatest scientific visionary of modern times, but even Einstein had his limits. Despite the extraordinary intuitive leaps he made, he often found himself unable to see what lay beyond his basic insights. As a result, many of the most stunning ideas associated with the theory of relativity were developed not by Einstein but by other scientists interpreting his work. In quantum physics, too, Einstein set out the fundamental concepts but initially failed to recognize where they would lead. And in his final, grandest search for a theory that unified all of physics, he simply never moved far enough beyond the math and science he had learned during his student years.

More surprising, Einstein resisted the full implications of his work even after those implications were pointed out to him. Repeatedly he sought to undercut many of his colleagues’ interpretations or to explain them away because they seemed too absurd to be true. These rejections recall the words of Arthur Eddington, a brilliant British physicist and one of Einstein’s most tireless champions: “Not only is the universe stranger than we imagine, it is stranger than we can imagine.” One of history’s most expansive minds was no match for the boundless oddity of nature.

Almost as soon as Einstein completed his 1905 paper introducing the special theory of relativity, he found his ideas taking on a life of their own. That paper spelled out how an observer’s motion through space affects his motion through time (to someone traveling at nearly the speed of light, time slows to a crawl), but it said nothing about treating time as a fourth dimension in a continuum of space-time. That concept, which today’s students learn as quintessential Einstein, was actually the work of German mathematician Hermann Minkowski. Einstein was at first nonplussed by Minkowski’s elaboration of his theory, shaking it off as “superfluous erudition.” Only years later did he recognize space-time as integral to special relativity and to the grander general theory of relativity that followed.

After Einstein published the definitive version of general relativity in 1916, he again found that his theory was full of oddities that he neither expected nor accepted. Just months later, Karl Schwarzschild, a 42-year-old physicist serving in the German army during the First World War, successfully applied Einstein’s abstract equations of space and time to a realistic physical problem, modeling the geometry of space surrounding a star. His solution impressed Einstein. Yet Einstein expressed one deep concern: Schwarzschild’s calculations showed that if the mass of a star were compressed into a small enough volume, Einstein’s equations went haywire. Time froze; space became infinite. Physicists call that a singularity, a place where the normal laws of nature break down. Schwarzschild had stumbled onto the first clue that black holes might exist.

For years no one paid much attention to Schwarzschild’s discovery, but in 1939 Einstein attempted to disprove the annoying singularity. He argued that a star could not exist under the conditions described by Schwarzschild because the material within it would have to reach orbital velocities equaling the speed of light. But Einstein assumed that the star had to remain stable, whereas the universe is full of objects that explode or collapse violently. In that same year, J. Robert Oppenheimer—the physicist who would soon direct the Manhattan Project—and one of his students showed that highly massive stars could implode under their own gravity, getting denser and more extreme until their gravity trapped even light. That is exactly how astronomers now believe most black holes form.

Not long after Schwarzschild’s discovery came another, even more troubling prediction from general relativity. Like most scientists of the time, Einstein was convinced that the universe (stars included) was static and eternal. So it came as a shock when, in 1922, an obscure Russian physicist and meteorologist named Alexander Friedmann showed that Einstein’s masterpiece theory described a universe that should either collapse on itself or fly apart. Einstein initially rejected Friedmann’s analysis as “suspicious,” then reconsidered and judged that the results might be mathematically correct but physically irrelevant. To fix what appeared to be a flaw in general relativity, Einstein adjusted his equations, adding a factor he called the cosmological constant—a kind of antigravity force—so that the equations yielded an unchanging cosmos.

Not until nearly a decade later did Einstein acknowledge his error. In 1929 the American astronomer Edwin Hubble discovered that all the galaxies appear to be hurtling away from one another at tremendous speeds. The universe was not static; it was expanding, just as general relativity had suggested. Two years later Einstein publicly denounced his cosmological constant. If he had trusted Friedmann —if he had trusted his own theory—back in 1922, he might have predicted that the universe was expanding, and he wouldn’t have had to scramble to adjust his theory after Hubble’s discovery.

“Einstein was furious with himself,” says Carlo Rovelli, a physicist at the University of the Mediterranean in Marseille, France. “He could have said, ‘My theory says this, so therefore I predict that the universe is expanding.’ But he didn’t have the courage to say that.”