He said 90% of everything is “crap” in 1953; what Theodore Sturgeon really meant and why his idea matters more in the AI age
“90% of everything is crap” is one of those lines that is frequently repeated but rarely understood in its original context. The phrase is known as Sturgeon’s Law, named after American science fiction writer Theodore Sturgeon. But Sturgeon did not coin it to dismiss science fiction or suggest that most creative work was worthless.He used the argument for almost the opposite reason. For years, Sturgeon had heard critics attack science fiction by pointing to its weakest examples and using them to condemn the entire genre. His response was simple: if 90% of science fiction is poor, then roughly 90% of everything else is poor too.
The argument behind Sturgeon’s Law
At a 1953 science fiction convention in Philadelphia, Sturgeon reportedly challenged the way critics judged entire genres by their weakest material.Mystery fiction could be judged by its finest work. Westerns had acclaimed novels and films. But science fiction, Sturgeon argued, was often represented by its most formulaic or poorly written examples.His point was not that 90% of science fiction had actually been measured and found to be bad. The percentage was rhetorical. Sturgeon was turning the criticism back on those making it.The argument was that judging an entire field by its worst work makes little sense because every field contains mediocre and poor work.
The 10% was always the important part
Sturgeon’s argument became widely known as Sturgeon’s Law, although he also used another formulation: “Nothing is always absolutely so.”The better-known version survived because of its bluntness. But the underlying idea was more nuanced. The existence of bad work does not tell us whether a field is valuable. The important question is what can be found among its best work.Philosopher Daniel Dennett later used a similar principle when discussing how fields should be criticised: if you want to challenge an area of work, engage with its strongest examples rather than its weakest ones.
What changed in the age of AI-generated content ?
Sturgeon’s argument has acquired a different relevance in the internet era, where the volume of published material has grown dramatically.Before digital publishing, producing and distributing a book, magazine article or other creative work involved editors, publishers, printing costs and limited shelf space. A large amount of poor material existed, but much of it never reached a mass audience.The internet removed many of those barriers. Generative AI has lowered some of them further, making it possible to produce large quantities of text quickly.That does not prove that AI has increased the proportion of poor writing. But it does make the filtering problem more visible. Readers can now encounter enormous amounts of material before finding work worth their time.
Your 10% may not be mine
There is another important part of Sturgeon’s thinking that often gets lost. In a later formulation, he suggested that the good part is the portion each person happens to like.That makes the “10%” subjective. There is no universally agreed list of the best books, films, ideas or pieces of writing.What one reader considers exceptional, another may find ordinary.That changes the lesson of Sturgeon’s Law. The goal is not to spend all your time complaining about the 90%. It is to become better at finding the 10% that matters to you.In a world where creating and publishing have become easier, that ability to choose may matter as much as the ability to create.
