Einstein wrote this line almost a century ago, and the part that doesn’t age is the ranking. Knowledge is below. Imagination sits above. Read it now, and the hierarchy lands harder than it has in any two decades, because the bottom half of the pair—gathering, tracking—has gotten dramatically easier.
The line itself appears in a small 1931 book, Cosmic Religion and Other Opinions and Aphorisms, just after Einstein mentions the 1919 eclipse measurements that confirmed his theory of how gravity bends light. The complete sentence, as Einstein wroteit says: “Imagination is more important than knowledge. Because knowledge is limited, while imagination embraces the whole world.”
He was talking about academic work, not writing, business, or skills frameworks. The imaginative step—the part where you imagine what the world might do—comes first. In Einstein’s narrative, the imaginative or intuitive leap is the first; formal work and empirical checks follow. It probably pushed back the popular image of the lab coat empiricist, as opposed to the idea that science makes its way to the truth. He thought he was imagining the way there, then checked.
For a long time, knowledge was more difficult to acquire. You had to read for him, study for him, find the right person and ask him. The couple’s gathering side didn’t come cheap. The line can therefore be read as an encouragement: don’t underestimate the more dreamy half. Now it reads more like an instruction.
What the WEF found in 2025
It is regularly published by the World Economic Forum The Future of Jobs reportthe most recent editions appear every couple of years., based on a survey of employers responsible for tens of millions of employees. The 2025 edition asked them which skills they consider essential today and which will grow the fastest by 2030.
The main finding: analytical thinking is the most important basic skill, seven out of ten companies consider it essential. Creative thinking, motivation and self-awareness are ranked fourth and fifth. AI and big data lead the fastest-rising list, followed by networks and cyber security, technology literacy and, along with flexibility and curiosity, creative thinking. Technical skills that speak to machines are rising on the list. Likewise, the human capabilities of machines are also the worst.
The same report does a more detailed check. He asks how many people are around 2,800 individual workplace skills today’s generative AI can plausibly replace it. The answer is that none of them showed “very high replacement capacity”. About 69% were classified as ‘low’ or ‘very low’. The areas where artificial intelligence can be easily replaced are the calculable areas – formulation, calculation, summarization, translation. Areas where it doesn’t require nuanced judgment or physical presence.
None of this is anti-AI. Good on the assembly side. You can summarize, search, outline, translate, sketch. I think the step that Einstein pointed to is very bad: imagine what is not in front of you and then look at it. The Forum’s own wording of the same statement is more procedural than poetic – that’s what he’s talking about enhancement of human labor instead of replacing it. The 1931 line and the 2025 survey result say versions of the same thing in different registers.
Here’s how it looks on my desktop
I use AI tools for parts of the job, not all of it. The assembly page is faster than before. Summaries, quick searches, half-finished outlines, talk-aloud sessions where you argue a framework against a model – it all just got a little easier. I spent the morning class flipping through five pages to find a useful study, now I spend it more efficiently. This part is real and pretending otherwise is unfair.
The part that hasn’t changed is the move you’re actually trying to produce in the morning. Sitting with the material until the same facts are arranged into a form that the play can be made of. Angle search continues.
I think that’s what the WEF result actually says below the chart. Analytical and creative thinking top the list, with AI literacy rising fastest below them. The headline is not that the imagination has been compromised. But it’s that they’ve quietly advanced the imagination because they’ve cheapened the layer underneath.
Einstein talked about how physics happens. He was probably right about that, and also, coincidentally, about what a day at work will be like for many people in the next ten or twenty years. First the imagination. Knowledge is the second. The order has not changed since 1931. The second half has now become cheaper and the first half is now visible.
