Why AI Tools like Devin can’t kill coding .
To understand coding, we need to understand language first.
Language is the expression of ideas by means of speech-sounds combined into words. Words are combined into sentences, this combination answering to that of ideas into thoughts.https://www.britannica.com/topic/language
In human languages, a large number of words, sometimes hundreds of thousands, are used to clearly express meanings.
On the other hand, programming languages usually feature fewer than 100 keywords, with additional functionalities being abstracted into libraries built upon these foundational elements.
Though programmers will not often use all the keywords , they do use some of it and predominantly most use from library. It’s perfectly okay for everyone to use the 100 keywords and do what we need instead of relying on libraries.
Language and subtleness.
Human languages are often nuanced; a single expletive, for example, can convey seven distinct meanings depending on the context. In contrast, programming languages typically assign one clear meaning to each word or symbol. “If” in programming is always a conditional, used consistently regardless of context.
Programming languages were designed to eliminate ambiguity, facilitating direct and unambiguous communication with machines. They do have their own syntax, much like human languages. However, the goal is to enable straightforward problem-solving without the nuances present in human communication.
AI aids programmers in transitioning between different programming languages without the need to memorize extensive syntaxes. For newcomers, AI offers easier onboarding through prompts. It doesn’t adhere to a specific language; rather, it adapts to express concepts in any existing programming language.
Devins AI
Consider Devins AI, as demonstrated in various YouTube videos. It interprets instructions given in natural language and translates them into Python code. But are the requirements it receives always crystal clear? Not typically. Natural language is inherently nuanced, leading to unclear requirements. Often, only about 20% of what a human thinks is explicitly stated as a requirement. The remaining 80% is assumed by the product owner to be understood by the developer, which is frequently not the case.
Thinking Devins AI can create its own programming language? That’s like expecting a toy robot to build a spaceship. It’s sticking to Python and JavaScript because, frankly, that’s all it knows. And those manual GitHub and cloud setups? It’s because Devin’s more of a stagehand than a star performer in the tech show. It’s got some neat tricks up its sleeve, but don’t expect any digital wizardry. In the end, Devin’s AI is less of a tech revolution and more of a tech illusion!
