I develop 5-10 times more complex projects now without anticipation to never complete them. And I can do it alone with team of agents running 24/7. Maybe these new projects will not be used by anyone, but I feel more confident in previously unknown for me domains like robotics, agentic runtimes or GraphRAG. So my advice: just keep evolving.
Do more. The LLM is not your replacement, it will help you achieve 10x more. LLM also have quiz mode so just feed it a book and ask to generate questions if you want more.
if anything i feel like using agent to code is a great opportunity to sharpen your critical thinking if you take the time tear apart the agent's assumptions and actively find flaws, areas of improvements etc. i don't think llms have removed any opportunity to practice good thinking (yet), it's just that you have more room to reason in design space rather than implementation space. TL;DR just keep poking at whatever your agents claim to be correct
Not nearly all the code. Maybe 5% - the important code. To keep things sane, explore new areas, get a deeper understanding of the system, delete/simplify agent slop, improve test coverage, or make some other foundational change.
Feels like we’re bending over backwards to avoid just doing this. But it seems for me, at least, the way to have the highest leverage on an agent heavy project, it’s fulfilling, it helps the agent, and it keeps me in flow.
Whenever I am trying to work on a project, I go through it myself first with proper details that what I need to build and all the specifications then tell LLMs the plan so they can code and I can keep my architecture skills intact. For keeping my coding skills intact, for small code changes, I prefer doing them myself, so I don't make myself much dumb.
> What's the silver bullet for both prouctivity and not getting your brain atrophied?
I haven’t stopped using my brain. There’s the secret.
I develop 5-10 times more complex projects now without anticipation to never complete them. And I can do it alone with team of agents running 24/7. Maybe these new projects will not be used by anyone, but I feel more confident in previously unknown for me domains like robotics, agentic runtimes or GraphRAG. So my advice: just keep evolving.
I like to generate a variety of personal crises to deal with since now coding agents seem to do most of my job. That keeps me on my toes.
Do more. The LLM is not your replacement, it will help you achieve 10x more. LLM also have quiz mode so just feed it a book and ask to generate questions if you want more.
if anything i feel like using agent to code is a great opportunity to sharpen your critical thinking if you take the time tear apart the agent's assumptions and actively find flaws, areas of improvements etc. i don't think llms have removed any opportunity to practice good thinking (yet), it's just that you have more room to reason in design space rather than implementation space. TL;DR just keep poking at whatever your agents claim to be correct
I just write code.
Not nearly all the code. Maybe 5% - the important code. To keep things sane, explore new areas, get a deeper understanding of the system, delete/simplify agent slop, improve test coverage, or make some other foundational change.
Feels like we’re bending over backwards to avoid just doing this. But it seems for me, at least, the way to have the highest leverage on an agent heavy project, it’s fulfilling, it helps the agent, and it keeps me in flow.
Whenever I am trying to work on a project, I go through it myself first with proper details that what I need to build and all the specifications then tell LLMs the plan so they can code and I can keep my architecture skills intact. For keeping my coding skills intact, for small code changes, I prefer doing them myself, so I don't make myself much dumb.