Era of AI Programmers
I've been occasionally trying AI at work on and off for a while. So here is my assessment as someone who's been into programming for more than a 15 years and have learnt programming the hard way (books, blogs, documentation, reading source code).
- AI bankrupts your thinking ability.
- AI has got really good at entry level noob tasks.
- Want to create 1,500 words content? easy.
- Want to create engaging landing pages? easy.
- Want to create CRUD app? easy.
- AI struggles at really hard problems. You'd never be able to code enterprise grade product with AI (just yet).
- At work, I could write algorithms to generate trading signals. Impossible to do it with AI.
- Write performant code taking into consideration current vs future scale.
- Sometimes the smallest code change can have the biggest impact. AI can never develop the human intelligence to make that small change.
- While AI can never be in the Arena, AI programmers would struggle to sense what's wrong with their app without thinking from first principles.
- Sometimes AI tries to cheat, so developers who just "Accept" are introducing bugs into code every day.
- Everyone's got their unique coding style and introduce a personality into the codebase. AI spits out code with zero personality, often robotic.
- AI defenders claiming that there would be more jobs, are dismissing the fact that there will be more noise than signal. For customers it's going to be extremely hard to determine signal and their subconscious mind would fall back on big-tech while making purchase decisions.
- Industrial scale manufacturing created the problem of plenty, often leading to landfills. In this AI era, we are already seeing exponential growth in creation of vibe coded projects often getting flipped/shut down leading to long term decline in trust.
- eg: Product hunt is mostly AI apps and submissions and is already dead.
- Finally, there aren't those many jobs to feed 8 billion people. Not everything needs to be made at industrial scale.