Too often, I’ve been harsh with AI.
Not because I don’t use it or because I fear it. Quite the opposite, I think it’s brilliant. What bothers me is how often it’s abused: sold as a shortcut, a source of low-quality code, or a promise of endless possibilities without any real effort. That’s where my critiques usually kick in.
For me, AI shouldn’t be a substitute for thinking. It should be an incredibly powerful tool. Accessible, yes, but one that adds another layer of consciousness and intention to the work. That’s where its true potential lives.
Until recently, I used AI pretty lazily. A senior on call for every question, a junior on demand for quick scaffolding and throwaway scripts. Helpful, time-saving, but always something I polished before pushing to production.
Starting my new job, though, I’ve discovered new ways to use AI that genuinely helped me hit the ground running.
Beyond Code: AI for Onboarding
The first surprise was non-technical. At my new company, we have an internal AI chatbot trained on Confluence docs and public Slack channels. Basically, a bot with the highest company knowledge at your fingertips.
During onboarding, instead of bothering teammates with trivial questions, or drowning in documentation, I just asked the bot. It gave me answers, context, and even helped me refine what I needed before going to a human. The result? My interactions with people were more focused, more thoughtful, and less of a time sink for them.
Another use case: people themselves. I’ve always worked in small teams, and suddenly learning dozens of new names and roles was overwhelming. The bot gave me background on who I’d be talking to. That made my first calls smoother, more personal, and much less nerve-racking.
Cursor: A Guided Tour of the Codebase
Then came coding. With Cursor, I suddenly had an expert guide to the codebase. Before touching a single ticket, I asked it for virtual tours of key areas, common patterns, and likely hotspots for my first contributions.
By the time my first tickets arrived, I didn’t panic. I already knew the terrain. Cursor helped me sanity-check my approaches and make small adjustments before diving in. That confidence boost was priceless.
Standards, Reviews, and PRs
One of my biggest worries was conforming to coding standards and team practices. Normally, this means digging through endless examples and second-guessing whether you’re following the “right” pattern or just copying tech debt. With Cursor, I could ask directly. The answers weren’t perfect, but they gave me a strong baseline to build from, and I learned to question and refine them.
When it came time for my first PR, I asked Cursor to play the role of a senior reviewer from my team. It flagged a couple of things that would almost certainly have come up later. Fixing them early meant I went into the real review with confidence. I even used it to draft my PR body, feeding it our template and getting back something clear, structured, and useful for my reviewers.
A Companion, Not a Crutch
These first few weeks, AI hasn’t replaced my work. It’s been a companion. It made the transition smoother, helped me learn faster, and let me show up to my team prepared and confident.
I’m not against AI. I’m against the idea of using it without effort, without curiosity, without brainpower. Because used right, it doesn’t just give you answers, it helps you ask better questions.