“The difficulty lies, not in the new ideas, but in escaping from the old ones.”
John Maynard Keynes wrote this line in the preface to The General Theory of Employment, Interest and Money. He was preparing readers for a book that challenged established explanations of employment and economic stability. The obstacle, he warned, was not a shortage of concepts. It was the reach of inherited assumptions through every corner of educated thought. The sentence describes intellectual change as an act of release. People can encounter a better model and still interpret it through categories that make the new insight appear unnecessary, impossible, or merely a special case of what they already believe.
Businesses face the same problem when an industry changes. Teams may add digital tools while preserving a paper-era workflow, call a subscription a product sale, or evaluate an emerging market with the cost structure of the mature core. Familiar metrics and reporting lines silently define what counts as evidence. New ideas then receive attention without gaining authority. Escaping an old idea requires naming the assumption, tracing the decisions it controls, and creating a test in which an alternative can produce observable results. Learning becomes structural only when budgets, incentives, and routines can move.
The quotation should not be read as contempt for experience. Old ideas often contain accumulated knowledge about safety, incentives, human behavior, and failed experiments. Novelty is not proof. The task is to distinguish a principle that still explains reality from a convention that survives because it protects identity or power. Leaders can compare predictions with outcomes, invite people who work closest to the problem, and record what evidence would change the prevailing view. Respectful challenge is more reliable than replacing one fashion with another.
Artificial intelligence makes Keynes’s warning newly vivid. Organizations can buy advanced systems while keeping assumptions about supervision, expertise, customer need, and accountability untouched. The result may be faster production of the same mistakes. A useful review begins by asking what must no longer be true for the current model to fail, whose knowledge the model excludes, and which rule exists only because an earlier constraint once made it necessary. Keynes’s line turns innovation inward: before celebrating a new tool, examine the old mental framework that will decide what the tool is allowed to become. One useful ritual is an assumption review held separately from the normal performance meeting. Participants bring a rule, metric, or customer belief that guides investment and identify the historical condition that produced it. They then choose evidence that would weaken the assumption and one reversible test. Separating this work from defending current results gives old ideas a fair examination without requiring change merely for novelty’s sake.
Keynes placed the sentence in the 1936 preface to The General Theory of Employment, Interest and Money.
He anticipated that readers would struggle less with his concepts than with assumptions embedded in prior economic thought.
Mental models guide what people notice, measure, fund, and dismiss, often without being stated.
Experience still matters; the goal is evidence-based revision, not reflexive rejection of every established practice.
Companies often install new technology inside old workflows, incentives, and definitions of value.
Naming inherited assumptions allows innovation to alter decisions instead of merely decorating the existing system.
Explore more "Quotes of The Day"
Discover more notable quotes from influential voices across politics, science, business, technology, sports, and culture. Each quote offers insight into how ideas, beliefs, and decisions shape the world around us.