Jetbrains has been pushing Junie real hard, my guess is that they've been giving out too many cheap tokens to try to stay relevant as Claude and friends pull people away from the IDE.
I still use Visual Studio in my day job, where Claude's output while VERY helpful, requires my ownership of everything, meaning I am inspecting every single line it changes. If a change is bigger than I think it should, I kill it before I commit.
I know I don't HAVE to use an IDE for that, but if the change is small enough, I am faster than asking Claude to understand the subtext behind my personal context of the product, and the IDE is supremely helpful for making a quick change across a handful of files.
> ...inspecting...
I got used to sublime merge to context-switch editing vs reviewing, staging changes line-by-line
I think this is changing fast. Pressure is increasing on devs for output, and most devs I know are no longer inspecting lines. I have devs in my business unit who claim to not have looked at code for months, except on certain rare occasions. I am not a developer and this week I've been given access to the repo to build my own apps and extensions. There's a "review" between commit and deploy, but there's no chance the Tech Lead can manually review everything, so that's getting done by AI too.
I know this horrifies a lot of devs, but these tools are shockingly good and we are not seeing an increase in bugs. In fact our automated detections (also AI assisted) are reducing the number of customer reported critical bugs.
I really think the days of inspecting every line are over.
I agree that they have gotten shockingly good. It's been a long time since I've seen them do something that is objectively wrong. Once we get closer to the "too cheap to meter" cost level things will change radically again.
It is. I have seen Claude chasing after endless amount of edge cases that are just irrelevant in real usage especially for the kind of users we are supporting. At some point you need to stop reasoning about all those cases because it has zero benefit.
I have seen millions wasted because someone trusted an ai scripts calculation of a metric from the bottom of the org that led the top of the org to make a wrong decision only to laugh about ai. There is value but ffs read the god damn code. You can have the cake and eat it too. If the volume of code is so large you cannot read it, maybe it isnt worth shipping?
Or are you one of the ones pushing the real code reading on to others which seems to be common. Yes i can have agents vibe out 10 features and have my coworkers suffer fixing it in reviews.
What IS useful are the AI reviews. They catch bugs, not all are bugs but they do catch some. It is almost like they are better at finding logical issues across millions of tokens but not good at writing streamlined logic.
The number of times ai gives me a 800 line dif only to replace it with a 5 line dif after i read it and notice it grossly overcomplicated the ask and scoped in a bunch of nonsense from training data.
Basically they give you a way to view the code the llm has generated/changed easily, annotate that code for the llm, and manage multiple agents and at once.
> I am inspecting every single line it changes.
Honestly, when I'm looking at ClaudeCode's output 75% of time i'm doing it to learn and understand and not just check for correctness. I'm confident enough to admit I don't know everything and I've learned a lot from reading Claude's code.
Heavy user of WebStorm and Datagrip since at least 2019.
I've tried it, and it seemed fine, but not compelling enough to even spend the free usage I get with my All Products Pack.
The Jetbrains integration is nice, but if you rely on the tool you're probably not going to use the IDE much anyway.
I would like to see one more ide-integrated, like I think running commands like ‘grep’ with the shell is really for the birds (creates a risk that some other command line might be run, the wrong files might be accessed, all that) and rather there should be a specialized toolbox.
[1] … I reject vibe coding. Token costs be damned but I always like to have a talk before it starts like “I think…, maybe you should…, does this make sense?, do you have any questions for me before you start?” and later “what are you doing in the code in the selection?”
It is alright to get an LLM review on code I write myself, but getting extra tokens is expensive, and it was not clear if I could configure it to use one of the API keys I have from GLM, MiMo, etc.
I tried Air as well. It was alright, but I found it a bit more cumbersome to use then Pi. I tried configuring Pi to be accessed though ACP, but it felt like going through a hoop to have a worse experience. Then again, I am not someone that manages multiple agents in parallel, at most I have one agent implementing something in a different repository while I am doing my own things.
Air could maybe be useful for me if I could plug in the LLMs I actually use directly, it is too tied to ChatGPT, Claude, etc.
Junie is boring, and that's perfect (for me).
The anomaly is the cash flow of investing activities, which is not something you can put such activity as you have described under.