"Pull requests are up 111% in two years".
Would be more honest to say that the number of pull requests "detected" by linear are up XXX%.
Because it only works if you setup git repo tracking and use it properly. And at that point it is not obvious if more teams are using linear and using it correctly, or if the number of PR really increased that much!
Why not just write the code yourself? To me it seems like methodically constructing the pull request by hand is probably faster than lazily prompting and re-prompting.
Writing code directly takes longer to warm up. Usually, I'd keep tens of thousands of lines in my head. In the past, I spent a lot of time designing error propagation and execution contexts. (Talented people might figure it out right away, but unfortunately I don't have that kind of talent.) So I'd have to think about things like Result<T> and how far to propagate errors—and worry about whether my approach would conflict with the existing codebase.
But these days, AI just generates code following the existing patterns of the codebase. In the past, staring at a blank screen meant going through a checklist of things to design—starting from policies and writing everything down step by step. Now, I just ask AI and it gives me a template—which is great. Then if the AI makes a mistake, I fix it manually.
Of course, I still hand-code sometimes—but only in the areas I enjoy. Most of the time, I use AI coding. Both are fun, and they complement each other in interesting ways. Doing both together is actually enjoyable.
For me, coding is either a flow state (when I have huge swaths of code to write) or deeply thinking about just a few lines (mostly to figure if I caught everything). The latter case is when I procrastinate the most, because I usually have an idea of a solution, but don’t want to do the work of actually verifying if it’s correct. Using AI then is skipping the enjoyable part (writing the code) to the unpleasant part (formulating hypothesis and verifying them).
Well you got me. I’d agree about that workflow, but something more rapid like a few seconds generating code and then a couple minutes reading it can be more engaging to me than writing it out yourself. I think it’s near time we all stop having such strong opinions about the matter either way personally.
It’s worthwhile to me under specific conditions, mainly that the codebase has existing patterns and abstractions that the LLM utilizes, and they’re actually good. It’s not a common case but I’m lucky that one of the codebases I’m in is one I built myself mostly by hand, and LLMs are really effective in it at the moment. My main concern is making sure the generated code maintains the high standard, but it’s definitely saving me time.
I’ve kind of changed my mind on prompting, it’s definitely a skill. It’s a skill based on your own skills in the domain. I’m at the point where I can get the LLM to generate the same code (roughly speaking) that I would have written. So it’s basically generating the same thing I would write, just faster. So it’s like reading your own code. Using it as a crutch to do things you aren’t capable of is where people run into trouble. That’s where the massive amounts of code review come into play. For me, I’m only ever reviewing 100-300 loc changes at a time. Often less. Because I know what I’m doing and can break things down into manageable diffs.
Can’t see myself going back, but also can’t see myself doing it without the experience I have without LLMs. Which is a bit of an issue for new developers. Not sure what the solution is for that.
That plus if you’ve ever stared at your code and then searched StackOverflow to see if you could find a better way of doing it, it’s like having that running continuously.
> Time spent on customer requests, docs, and projects held steady [..] AI has so far changed how teams execute far more than how they decide what to build
I think the measurement for this may be flawed. We do mostly use AI to decide how to build. But what we build is influenced by AI-driven research into a problem or task. That's largely done in coding and desktop AI tools, not Linear Asks/AI.
I'm working on accelerating my team's work by implementing AI-driven code pipelines with guardrails to eliminate as much unnecessary review time as possible. Also making a chatbot for turning repetitive tasks & PRs into buttons, and an "architectural guidance" chatbot that gives advice tailored to our business, software/system architecture, cloud, standards, etc. This puts AI and automated jobs in the center of both how (automated task) and what (architecture guidance).
But this has a not-so-great implication for Linear. With my tools, a human never has to touch a ticket, so we could use any ticketing system with an API or CLI. Linear is a great product because they made a great interface. What happens when I replace their interface with a chat bot?
"Pull requests are up 111% in two years". Would be more honest to say that the number of pull requests "detected" by linear are up XXX%. Because it only works if you setup git repo tracking and use it properly. And at that point it is not obvious if more teams are using linear and using it correctly, or if the number of PR really increased that much!
These days, my work has become: generate code for 20 minutes, then spend an hour reading it.
And then more hours cleaning it up and re-prompting.
Why not just write the code yourself? To me it seems like methodically constructing the pull request by hand is probably faster than lazily prompting and re-prompting.
[delayed]
Writing code directly takes longer to warm up. Usually, I'd keep tens of thousands of lines in my head. In the past, I spent a lot of time designing error propagation and execution contexts. (Talented people might figure it out right away, but unfortunately I don't have that kind of talent.) So I'd have to think about things like Result<T> and how far to propagate errors—and worry about whether my approach would conflict with the existing codebase.
But these days, AI just generates code following the existing patterns of the codebase. In the past, staring at a blank screen meant going through a checklist of things to design—starting from policies and writing everything down step by step. Now, I just ask AI and it gives me a template—which is great. Then if the AI makes a mistake, I fix it manually.
Of course, I still hand-code sometimes—but only in the areas I enjoy. Most of the time, I use AI coding. Both are fun, and they complement each other in interesting ways. Doing both together is actually enjoyable.
For me, coding is either a flow state (when I have huge swaths of code to write) or deeply thinking about just a few lines (mostly to figure if I caught everything). The latter case is when I procrastinate the most, because I usually have an idea of a solution, but don’t want to do the work of actually verifying if it’s correct. Using AI then is skipping the enjoyable part (writing the code) to the unpleasant part (formulating hypothesis and verifying them).
That's gotta be at least 10x or 20x more efficient than the old way of doing things.
It could be. Or it could be 1x, or 0.2x. You don’t have enough information to make that judgment.
I was joking. The workflow doesn't seem particularly fast or engaging in my opinion. AI code generation doesn't seem worthwhile to me.
Well you got me. I’d agree about that workflow, but something more rapid like a few seconds generating code and then a couple minutes reading it can be more engaging to me than writing it out yourself. I think it’s near time we all stop having such strong opinions about the matter either way personally.
> I think it’s near time we all stop having such strong opinions about the matter either way personally.
I don't think this is reasonable given how abusive the pro-AI rhetoric has been for years now.
I’m entitled to my opinion about your opinion just like you’re entitled to your opinion about my opinion about your opinion.
If you don't accept my opinion as your own you're going to lose your job and become a member of a permanent underclass.
It’s worthwhile to me under specific conditions, mainly that the codebase has existing patterns and abstractions that the LLM utilizes, and they’re actually good. It’s not a common case but I’m lucky that one of the codebases I’m in is one I built myself mostly by hand, and LLMs are really effective in it at the moment. My main concern is making sure the generated code maintains the high standard, but it’s definitely saving me time.
I’ve kind of changed my mind on prompting, it’s definitely a skill. It’s a skill based on your own skills in the domain. I’m at the point where I can get the LLM to generate the same code (roughly speaking) that I would have written. So it’s basically generating the same thing I would write, just faster. So it’s like reading your own code. Using it as a crutch to do things you aren’t capable of is where people run into trouble. That’s where the massive amounts of code review come into play. For me, I’m only ever reviewing 100-300 loc changes at a time. Often less. Because I know what I’m doing and can break things down into manageable diffs.
Can’t see myself going back, but also can’t see myself doing it without the experience I have without LLMs. Which is a bit of an issue for new developers. Not sure what the solution is for that.
That plus if you’ve ever stared at your code and then searched StackOverflow to see if you could find a better way of doing it, it’s like having that running continuously.
> Time spent on customer requests, docs, and projects held steady [..] AI has so far changed how teams execute far more than how they decide what to build
I think the measurement for this may be flawed. We do mostly use AI to decide how to build. But what we build is influenced by AI-driven research into a problem or task. That's largely done in coding and desktop AI tools, not Linear Asks/AI.
I'm working on accelerating my team's work by implementing AI-driven code pipelines with guardrails to eliminate as much unnecessary review time as possible. Also making a chatbot for turning repetitive tasks & PRs into buttons, and an "architectural guidance" chatbot that gives advice tailored to our business, software/system architecture, cloud, standards, etc. This puts AI and automated jobs in the center of both how (automated task) and what (architecture guidance).
But this has a not-so-great implication for Linear. With my tools, a human never has to touch a ticket, so we could use any ticketing system with an API or CLI. Linear is a great product because they made a great interface. What happens when I replace their interface with a chat bot?