Ask HN: Do you struggle with flow state when using AI assisted coding tools?
21–30 of 55 posts
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#22For me, flow state can be achieved through the true multitasking that agentic coding allows. Start an agent on a task, then set up another agent for another task. While one is working, the other usually needs feedback or tweaking. I run Claude Code in yolo mode in a dev container where it can cause little damage, so I’m not wasting my time approving tool calls (or managing an allowlist of tool calls). Sometimes I hav…
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#23For me, flow state can be achieved through the true multitasking that agentic coding allows. Start an agent on a task, then set up another agent for another task. While one is working, the other usually needs feedback or tweaking. I run Claude Code in yolo mode in a dev container where it can cause little damage, so I’m not wasting my time approving tool calls (or managing an allowlist of tool calls). Sometimes I hav…
Is isolating files they work on an issue for you? Does this simply not pop up (because you avoid it by having a good understanding of what will be touched) or are you doing a lot of merging?
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#24My best "hack" for this has been to use Freedom[1] to create a blocklist of all my go-to time sucks (including, sadly, HN). This at least stops me from getting pulled in too deeply.
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#25Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#26This is Claude Sonnet 4
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#27No, I'd say they actually help me stay in flow. The autocomplete mode takes care of annoying boilerplate (mostly) and if I'm stuck on something the chat mode can often provide a suggestion that helps me unstuck (even if not usable directly). These two help me keep the state on for longer. Agent mode just produces dogshit that takes me longer to clean than if I wrote it myself to begin with. I don't use that.
Fair enough, but for most frontend or UI/UX tasks I do you claude code which is similar to Agent mode and that's when I have these constant interruptions in the flow. I believe this will be worse in the near future as agents become more powerful.
Contrary to popular belief you won't be left behind if you don't use these "agents". If they ever get good enough that they don't need babysitting anymore and make you 10x productive, then you can simply adopt their use then. Any current skillset regarding prompting won't stay relevant.
Though I will caveat that I work on more niche subjects, not frontend, so my perception of their usefulness may be skewed.
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#28I've had a few instances where I've been able to go into flow state with 3 concurrent sessions on fully disjoint projects. Feels like context switching is easier when there's almost zero cognitive overlap between the projects in terms of actual features, but the stacks are all similar.
But for sure I haven't even come closer to these glorious flow sessions fully locked into coding some clever bit of logic.
So, here for the suggestions too.
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#29Most of us are doing relatively simple things like using with the largest/slowest models possible instead of downgrading to a smaller model. Reason: our tools lack the ability to escalate to more expensive models as needed and switching models manually is tedious. Ideally, a cheap locally running model would be first in line and respond quickly for quick things.
And then there's the whole asynchronous vs. synchronous thing. With Codex, it runs in a browser and it allows you to create a pull request whenever it is done. You can work on multiple pull requests with it even and it might work in parallel.
What's good about the Codex experience is that your input is only needed at the end. What's bad is that it takes ages. Even for simple stuff. Like a simple follow up question results in it boiling the oceans to startup the container again.
Slow AI is exactly like slow builds: frustrating and likely to distract you. If it's going to take a minute you are going to do something else. And that might not be work related. So it breaks your flow because you are sitting on your hands and filling your short term memory with garbage. Context switches are expensive (and break flow). Our brains don't do that well. And then you forget to switch back so you lose time.
I don't use reasoning models that much for this reason. It's easier and faster for me to manually patch up my code with whatever the LLM says I should fix. And on larger repositories the chance of a good PR drops sharply. So, even if it takes more context micromanagement to feed it all the detail it needs, this can be faster and more effective. And I get an answer in a few seconds instead of in a few minutes.
Re: Ask HN: Do you struggle with flow state when using AI assisted coding tools?
#30Then it costs me a lot of energy to get back working on the problem and fix all the mistakes that the AI made.