Convergence of Code and Speech

As part of the Women Tech Global 2021 conference, I presented a short talk called “Convergence of Code and Speech”, under the theme of “Future of Work”. In hindsight, the title was a little pretentious, but what are you going to do?

I haven’t gotten the recording of the live version yet, but I always like to do a recorded test run by myself, and here’s what that was like:

Programming code and human speech are often seen as being positioned at different ends of the spectrum. At one end, code is terse, strictly logical, and limited. On the other end, speech is unbounded and often verbose. However, the gap between the two is narrowing.

Code is moving towards speech. Good code these days is code that exceeds simply being functional; good code is concise, but expressive; good code is easily comprehensible; good code refers to human concepts rather than technical components.

Shrinking global borders and the internet are making collaborating on code easier than ever before, but contributors may come from different cultures and languages. Remote work has significantly increased the likelihood of a multilingual devleopment team, so code style guides have emerged that are similar to writing style guides. Artificial intelligence is learning to interpret intention, and many apps now employ fuzzy search algorithms. Our expectations for code have increased such that we now expect code to understand us even when we’re being imprecise or inaccurate. Modern programming languages like React and Python use hierarchical abstraction to hide complexity, moving towards code that prioritizes being understood over being thorough.

On the other side of the spectrum, speech is also moving towards coding. Multilingual teams also encourage simplicity in written communication, removing colloquialisms and cultural references so that a more global audience can understand us. Tools like Grammarly correct for simplicity and conciseness. Verbal speech, too, is becoming increasingly programmatic with the rise of “smart” speakers and personal assistants that enforce a syntax for voice commands. Advances in artificial intelligence and machine learning have enabled us to interact with machines in natural language, and machines to interact with us back. OpenAI’s GPT-3 can produce convincingly human text about a variety of topics with little prompting. Writing tools like Notion, Roam Research, and Obsidian are growing ecosystems that mix human language with programming keywords to enable backlinks or access other features.

What are the implications of this convergence of code and speech? Writers will need to learn to code, and coders will need to learn to write. The two will become synonymous, and equally essential especially for knowledge workers. Tools that enable both coding and writing will become more valuable, so it’s worth investing the time to learn them. Both code and speech have inherent linguistic or cultural biases, so using inclusive language will become even more important.

What are coding and speech converging into? Their intersection is simply expressing thought. She who is able to express thought, both programmatically and naturally, will be best poised to take advantage of this convergence.