I don't really know what to tell you. Using AI agents feels like cheating. But it is literally impossible to ignore the fact that the latest AI Agents are objectively better at coding than most developers. Even considering the hallucinations, manual code written by a human also produces bugs and poor choice of software patterns.
If you accept the fact that AI hallucinates code in certain areas, or domain specific topics, and yes I know it's a big ask to accept that and use AI. However, all AI has to do is code better than humans, because both AI and humans can make mistakes. The cloud AI robot simply makes fewer mistakes in many areas, and especially boilerplate code.
The argument that Claude can't write embedded firmware drivers for some obscure esoteric hardware doesn't mitigate the fact that the vast majority of code these days are written in languages that agents do understand. Python, javascript, go, rust, SQL.
The amount of time creating backend to a database can take days, weeks, or even months. That's because you have to have someone construct the SQL queries, someone to create the backend server, someone to write the routes that hit the database, and whatever else.
AI agents literally can find the shape of the backend, the shape of the db schema, and automatically connect all that shit together. It can write clean code, and add comments useful for manual debugging.
That being said...
“The Three Laws of Robotics:
1: A robot may not injure a human being or, through inaction, allow a human being to come to harm;
2: A robot must obey the orders given it by human beings except where such orders would conflict with the First Law;
3: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law;
The Zeroth Law: A robot may not harm humanity, or, by inaction, allow humanity to come to harm.”
~Asimov
Well I don't know man, we're already using this shit to kill people in wars and such, so...