In the field of coding, artificial intelligence has regained authority. But there are some domains where state-of-the-art AI tools fall short of excellence.
Bjarne Stroustrup, creator of C++, one of the most widely used programming languages ​​in history, revealed in a recent podcast one specific domain where AI-generated code is struggling to make an impact.
“I think the area I’m most interested in (programming language design) — the code will still be written by humans, and they will use abstractions,” Stroustrup said on a podcast.
“The examples I’ve seen of attempts to generate code by AI in this domain have not been successful.” He still believes that humans will have the upper hand in this area.
According to Stroustrup, AI struggles in critical areas and the problem is that the more bugs and bloated code it creates, the more security it creates. It is almost impossible to verify these codes. The scariest thing about these problems is that even the most competent engineers don’t want to deal with these problems.
“Senior developers will need to verify it – I’ve seen some of them start to retire, because they don’t want to deal with validating something that changes every time you change signals.”
Stroustrup also talked about the problems caused by AI-generated code because a slight change in the prompt can change the entire codebase in unexpected ways.
This is quite different from human-made changes because when humans make changes, the change is always localized.
“If an AI writes it, you don’t really know where it changed. You have to try to figure it out,” Stroustrup said.
