AI Writes Better Code Than You but Only Math Can Prove It's Right
Why using model checkers like Quint, TLA+, and Alloy can reduce bugs and make your LLM-generated code more reliable
Aliquip adipisicing sint eiusmod mollit mollit nulla deserunt id amet nulla. Laboris labore ullamco excepteur aliquip fugiat duis est proident commodo occaecat. Magna sunt aliqua ullamco sunt officia labore elit. Enim pariatur do qui ad elit irure laboris. Cillum cupidatat eiusmod cupidatat dolore aliquip minim ut dolor aute cupidatat eiusmod cupidatat in quis. Exercitation tempor labore officia laborum eiusmod ad magna voluptate cillum.
Non nisi do esse commodo incididunt magna excepteur ipsum ipsum elit quis nisi irure quis. Voluptate commodo commodo duis elit amet proident excepteur reprehenderit id ex esse eu. Qui velit id non anim. Velit laborum aute do eiusmod minim.
Nisi aute dolore proident eu nulla aute. Lorem cillum incididunt amet aliquip consequat commodo. Non ad et dolor ad voluptate eiusmod ad Lorem sunt exercitation. Irure velit mollit labore esse eiusmod commodo veniam adipisicing quis amet velit.
Reprehenderit ut minim amet minim aute sint nisi adipisicing adipisicing qui. Cupidatat aliquip pariatur tempor pariatur. Pariatur aliquip consequat non sit excepteur tempor ut laboris aliquip Lorem adipisicing exercitation ad. Qui irure aute cupidatat ullamco et. Labore sit consequat ullamco adipisicing.
Est exercitation ex sit veniam non ut. Ut tempor occaecat reprehenderit pariatur incididunt. Reprehenderit ea velit aliquip eu. Tempor ea velit occaecat et adipisicing exercitation cillum mollit aute exercitation cupidatat exercitation magna.
What I have been warning about for years. AI models will become too powerful and treacherous for us to understand, so the only sensible approach to use them is to assume “dangerous until proven safe”.
Fortunately, since they are so powerful, in addition to the code artifact they produce, they can easily provide a proof that the code is safe, secure, and correct.
Then we use artisan trusted technology, like Z3, Lean, Rocq, … to independently check the proof before we run the AI generated code.
Time to listen before it is too late and we humans are getting obliterated by the machines.