A new neural network system allows users to write code by directly formulating thoughts or spoken language, which is instantly converted into working software. In other words, a person does not need to learn programming language syntax, but simply formulate what they want to get, and the system itself will convert it into functioning code.
Such a tool will sharply lower the barrier to programming: people who have never written code before will have the opportunity to create working software without deep technical preparation. At the same time, for experienced professionals, the system will boost productivity, since the gap between formulation and actual code will practically disappear.
Notably, the system is based on natural language, meaning it understands not special programming instructions, but ordinary human formulations and speech. This is a case where technology tries to adapt to humans, rather than humans to technology.
The lowering of the programming barrier is not always perceived unambiguously. On one hand, more people will be able to realize their ideas, which is good for innovation. On the other hand, when the system writes the code, the question arises of who is responsible for its operation and security.
This material lacks concrete data on how reliable the system is in complex projects and whether it can create code that is easily vulnerable. These are questions whose answer will determine whether the tool remains an experimental demonstration or becomes part of everyday work.
In any case, the direction is clear: technology is moving toward the point where the language barrier between humans and machines is increasingly diminishing.
