Programmer Yevgeny Rusakov, who worked at Rozetka, Mastercard, and Silpo, admitted to Threads that he spends most of his time writing prompts for artificial intelligence.
Programmer Yevgeny Rusakov, who worked at Rozetka, Mastercard, and Silpo, admitted to Threads that he spends most of his time writing prompts for artificial intelligence.
«You know, when I read posts about 90% of the code being written by AI and programmers becoming prompt operators, I laughed at that. I take my words back, I’m a prompt programmer 90% of the time. It’s not funny anymore,» he wrote .
Other social media users are actively commenting on his post.
«One of my fears is working on the same project with such a programmer and then deleting those 100,500 lines of generated code. Because AI is simply not in the context of what is already written in the system, what the system is integrated with, and what are the agreed-upon rules for writing code in the team,» notes user Dana Hrychun-Pylypets.
«That’s what a programmer is for, to bring AI into context. You’re just confusing it with vibe coding, where you don’t look at the code that the neuron gives, and stupidly force it to do everything. But that’s an extreme. And so you just indicate the boundaries of what’s needed, indicate what to rely on, etc. You set the framework and describe the task in detail, and everything is in chocolate,» explains designer Ivan Kireev to Dani.
«Unfortunately, the understanding of what good (or even acceptable) code is is rapidly being lost,» notes a user with the nickname friedensreich26, co-founder of Modular.UA.
«Actually, it’s not bad, if you were able to assemble an algorithm, an architecture in your head, then why write? We are waiting for a future in which thoughts will be directly translated into code,» noted STO Maksym Rudenko.
Нас чекає мертвий інтернет, бо ця стаття теж написана нейромережами