Back in Dec 06, 2024 I played with Cursor IDE for 2 minutes and immediately signed up for the pro version. Impact was unescapable. Familiar interface. many same extensions I have been already using. In 2025, Cursor has saved me hours and hours of time. It found bugs I couldn’t see. So thank you the devs.
However, the direction IDE is going doesn’t feel right anymore. About a year ago, cursor was fun to use, AI was fun to interface with. Interface was familiar. I can’t recognize it today. Too many interface changes (stuffs moved left and right, disappeared, and then back). AI tone changed quite a lot, from being fun to work together to a serious, cold robotic automata. Today was tipping point when I asked it to do something, it started with “Customer is asking for…”
instead of “let me look into it.”. We need that fun. Our jobs (sw dev) are already harder, bosses are breathing down our necks, is it done yet, is it done yet, why can’t you do it quicker w/ AI, ask AI to do it, just get something that works ![]()
Such a change in one year reminds me the book: “Tog on Interface” by Bruce Tognazzini. One of the first chapters says customers jump ship when software has changed so much that it doesn’t look like existing software anymore and customers end up considering the new version like one of competitors’ options. I’ll be frank, in the last two months I thought about and tested a few other IDE’s. I like Cursor (my first love :)) and still use it, but my feedback here to show my uneasiness.
Cursor’s value to me was it was David against Goliath, having familiar interface and extensions and dev team that was putting a useful product.
Anyways, it is also possible I’m not in the target customer defintions. I certainly don’t use background agents. Still, improvements are good but maybe schedule them every 3 months (or 1 month) and in an incremental way?
Take us on a journey, not an off-road race w/ sharp turns ![]()
Thanks for helping change the world.