Critical observations and verification issues
According to B. Stroustrup, the human factor in programming will remain essential – programmers will continue to write code themselves and use abstraction.
Read more “Neurobics” test: can you recognize 5 movies starring Matt Damon?
The C++ creator highlighted several main problems encountered when using AI for code writing:
Excessive and insecure code: AI-generated code often contains more errors and security vulnerabilities, making it very difficult to check and verify.
Departure of senior code masters: Senior software developers, whose expertise is necessary to verify AI-created code, are beginning to retire or leave the market because they do not want to work only as code checkers for code that changes with every modified prompt.
Read more Hollywood actress Susan Sarandon: supporting Palestine cost me roles
Unsuitability for critical systems: AI is not suitable for writing code that performs critical security or high-performance functions.
B. Stroustrup emphasized that although perhaps 70–80% of the world’s code does not require such strict security or speed-based accuracy, it is precisely the remaining 10–20% of the most complex code that is his main interest.
Read more EK leader: only Greenland and Denmark can decide on Greenland’s future