> Current AI tools are unable to keep large codebases in context, making it impossible to reason about system-wide impacts.
they have trouble even with small database, recently I was experimenting with generating a small script
Code-gen tools generated nice testcases from description (this was usable!).
But code they generated failed on exactly this testcases, feeding back crashes/test failures resulted in no improvements. I tried several (Claude, Gemini, ChatGPT, Mistral). None was able to do this.
they have trouble even with small database, recently I was experimenting with generating a small script
Code-gen tools generated nice testcases from description (this was usable!).
But code they generated failed on exactly this testcases, feeding back crashes/test failures resulted in no improvements. I tried several (Claude, Gemini, ChatGPT, Mistral). None was able to do this.
LLMs will never be ready.
These AI agents will go after both mid-level and senior-level engineers and set the price of them in a couple of years time.
What is also certain is that there will be far less engineers hired.