AI can produce garbage, it can produce excellent results. As long as you know what you are doing, and provide perfect specification, it can come with results, that you alone can't think of, or they can take a lot of time for 'manual' research and implementation. Some people still find issues, and don't trust, but the topic is big here: what's the model, how good is the specification, what's the used process/workflow, what are the agents, what's the technical background of the person leading the agents. Not that easy to say: AI does not work, why people want to use it after all, when it produces garbage
Life changes, customer demand it. We have no choice, and we need to adapt. Of course, there will be places that code will still be written manually. Banking, military, and other highly confidential areas have no other way to do it.
How did people feel in the beginning of the industry revolution? Have they felt the same way back then ?