Frontier models from OpenAI/Anthropic/Google make tons of mistakes. I work with them every single day and you really have to verify everything if you don’t want

Frontier models from OpenAI/Anthropic/Google make tons of mistakes. I work with them every single day and you really have to verify everything if you don’t want to end up with overengineered spaghetti code full of bugs.