Probably not very helpful for programming. At least yet, maybe if voice command lines like Alexa or MCP Agents advance you could run a Stenographer to voice like 100% faster than what you could achieve just by running text-to-speech. But also at that point you could just speak, I guess.