Comment by Twey

I don't think there's _fundamentally_ anything hard about voice coding, but our current systems are terribly designed for it. We need programming languages that are much more keyword- rather than symbol-heavy, or more efficient systems of pronouncing symbols, or higher-level editing primitives, or maybe all of the above. There's some overlap with both stenography (which I know some people use for code, but I've never got around to setting up a theory for) and also structured editing projects like Hazel. Reckon we might see more of it in the future, if we're to make the ubiquitous and malleable computing dreams a reality.