Martin touched on this point in his video on AI Code Generation ruclips.net/video/Z_OSq0eh2xM/видео.html (answer: no, a human programmer must verify the code). Jeff Crume, the channel's security guy, also cited that oversight is needed *especially* for open source/public code since the model *could* be influenced to generate malicious code by corpus poisoning.
Any certain tools or libraries?
I think this video was an Ad of some IBM product for modernization of COBOL systems
ansioso e no aguardo deste projeto maravilhoso muito bem contextualizado com a atual conjuntura global tecnológica
What about AI hallucination? Can we count on AI generated code?
Surely the models will never improve 🙄
Martin touched on this point in his video on AI Code Generation ruclips.net/video/Z_OSq0eh2xM/видео.html (answer: no, a human programmer must verify the code). Jeff Crume, the channel's security guy, also cited that oversight is needed *especially* for open source/public code since the model *could* be influenced to generate malicious code by corpus poisoning.
I think, Generative AI will end most of developers' jobs. What are your thoughts?