CONCEPT
Formulative Thinking
Licklider's category for the cognitive work that happens before a problem has been specified — the messy, associative, exploratory process of figuring out what the question actually is.
A formulated problem has been specified with enough precision to be solved — variables identified, constraints stated, equations written. Machines have been solving formulated problems since the first vacuum-tube calculators. Formulative thinking is something entirely different: the researcher sensing a pattern she cannot yet name, the engineer knowing something is wrong without locating the fault, the writer feeling the shape of a book before finding its structure. Every interface before 2025 demanded that the human formulate before engaging the machine. The natural language interface was the first to accept formulative input — messy, partial, exploratory — and respond with interpretation rather than error messages.
In The You On AI Field Guide
Licklider's first stated aim for the symbiosis was 'to let computers facilitate formulative thinking as they already facilitated the solution of formulated problems.' The sentence is easy to read and hard to appreciate. It describes a capability that did not exist, that would not exist for decades, and that required not merely faster machines but