@MackensieJack Great! I'm tackling the scalability and security of the products I'm building, as you know with the AI usage and outages/disasters that can happen. But yeah, learning a lot as well. Keeping up is hard though with everything going so fast
Reflections after working with ai coding agents for 7 days / 10+ hours a day:
1/ there is definitely a hard cap of what you can do no matter how good the model is
That is determined by the time itself which is 24 hours in a day. No ai model can give you more time.
2/ definitely these things do a great job but screw things badly if you have no idea what you are doing
Sometimes they make you spin in circles for hours
3/ ai agents cannot replace developers. Not not not in the next 2-3 years. They simply canโt. You need a developer.
4/ ai in the hands of an experienced developer accelerates the first 60% of a project massively. Then then gain massively decreases.
5/ you can obtain small useless tools very fast. Complex ones require days or weeks to build.
For the record I have been a developer all my life and have over 15 years of experience developing software.
I have CS degree if that matters for a fact.
Keep developing apps and use ai, itโs still hard to develop production ready apps.
@jamonholmgren When we're talking about 'tech debt' of an established project, it's not as simple as building a new feature, and mustn't leave any tech debt behind.
It's not that simple in reality. Refactoring is involved. Priorities change. Requirements change. There are so many factors.
@Star_Knight12@ThePrimeagen@theo Is it true? I'm not yet into coding other than web and app development, and I'm learning Go currently. What's your advice for us juniors?
You might believe you should spend less time thinking about code because of AI.
I strongly disagree! Weโre watching this play out live where tons of AI generated code becomes a liability.
At the end of the day, an engineer needs to be responsible / on call for code that gets shipped to production. If you donโt understand the system youโre trying to debug, youโre probably going to have a bad time.
Yes, AI can help with all of this, if you set up the proper systems. You can have agents triage prod logs, look at errors, etc. You can speed up parts of the investigation, but an engineer needs to make the call. There might be serious customer or financial implications from that change.
I expect the trend continue for trimming dependencies, vendoring code so you can modify it directly, preferring simpler systems with fewer abstractions, and spending waaaay more time thinking about system design and code maintenance.
Iโve said this before, but itโs a great time to get familiar with CS fundamentals and some of the history behind what great software looks like. Many parts will be different in the coming years as AI progresses, but also a lot more than people realize will stay the same.
@krishdotdev Oh wow, so you're the one from 2026, launching MVPs in 15 days, creating a mess.
How sweet of you man! Already thinking of seniors in 2027. You're so kind.
@moshhamedani But, wait, you're utterly wrong here.
We only use "agentic coding" when we've already done that task multiple times ourselves and now we hand those repetitive tasks to Agent. So, why one problem, wait and waste time? You can run multiple agents to solve those repetitive tasks.