The fact is that this sector is constantly evolving. What if frameworks, what if new languages … but now it seems that one more variable enters, voice.
We read about it in an article on spectrum.ieee.org, where they comment on two platforms that perform the passage from voice to code:
– Serenade, a digital assistant who writes the commands without requiring that we necessarily dictate each instruction word for word. It has a speech-to-text engine developed specifically for code, unlike Google’s speech-to-text API, which is designed for conversational speech. It has machine learning models trained to identify and translate common programming constructs into syntactically valid code. It raised $ 2.1 million in a seed funding round in 2020.
– Talon, what gives more control over each line, and requires us to have more understanding about each task that is programmed into the machine. It has several components: speech recognition, eye tracking, and noise recognition. Talon’s speech recognition engine is based on Facebook’s Wav2letter automatic speech recognition system, while the eye-tracking and noise-recognition capabilities simulate navigation with a mouse, moving the cursor across the screen based on movements. of eyes and clicks based on mouth noises.
If we compare them, Serenade follows a more natural way of speaking codeWhile Talon seems like another language, it requires more effort.
There are other open source speech coding platforms, such as Aenea and Caster, which are free, but both rely on the Dragon speech recognition engine, which users will have to purchase. Caster offers support for Kaldi, an open source speech recognition toolkit, and Windows Speech Recognition, which comes pre-installed on Windows.
Be that as it may, voice coding is still in its infancy, more oriented to those who have some limit when it comes to typing, since at the moment it is not possible to speak naturally about what we want and see the corresponding code on the screen.