To keep up at work, I have to use AI - in turn, I have found that I just have tmux with lazygit, perhaps neovim to check on a file out of curiosity and my agent(s) open. I only ever use my IDE to launch a local instance of what I'm working on, none of the features or plugins really matter anymore. I wonder if the future of IDEs is bleak.
I think the UI metaphors will evolve over time. I do think terminal UI can get you very far, but as things 1) become more complex, and 2) happen in the cloud/remote machines and collaboratively across multiple participants, I think terminal makes less sense.
These days, something like the GPT/Claude desktop apps are pretty good, where the interaction with agents is front-and-center, but IDE-like UIs can appear when necessary.
I feel like the ultimate IDE is WYSIWYG and visual, while allowing arbitary natural language edits, plus some affordances to view the underlying code artifacts when necessary.
IDE is not dying but shifting from a dump text editor to an orchestration hub for local and cloud AI agents. Cursor is still a useful thing, where just the editor feature may fade over time :)
Yea. I haven’t opened VS code in I don’t know how long. Among other things, I just shipped a full mobile app for Android and iOS and never had to touch XCode or Android Studio.
Those IDE tools that help humans also help LLMs, and I’m sure there’s a better UI than a prompt box for users to view and edit code (though maybe it will be accessed from a prompt box, and the next step is IDE tool calls and UI building blocks).
This is me, too. I find IDEs to be generally a pain in the butt, so stopped using them (except at work, where using Visual Studio is basically required) many years back.
The terminal (or, honestly, multiple terminal sessions) is a much better solution for me.
The future is brighter than ever with IDEs and GUIs.
Using the terminal and other things like lazygit is currently limiting and doesn’t really make sense.
I think we will go back to IDEs as AI improves vastly (think cheaper and better computer use models) and we will stop using and shoving everything into terminals all the time.
These days, something like the GPT/Claude desktop apps are pretty good, where the interaction with agents is front-and-center, but IDE-like UIs can appear when necessary.
I feel like the ultimate IDE is WYSIWYG and visual, while allowing arbitary natural language edits, plus some affordances to view the underlying code artifacts when necessary.
The terminal (or, honestly, multiple terminal sessions) is a much better solution for me.
Using the terminal and other things like lazygit is currently limiting and doesn’t really make sense.
I think we will go back to IDEs as AI improves vastly (think cheaper and better computer use models) and we will stop using and shoving everything into terminals all the time.
It just feels so wrong.