Feedback from a new user

hello, this will be a hot take, but I’ll provide some context first. I’m a “senior” software developer, been getting paid to code professionally for over 12 years now. I had been using ChatGPT Plus exclusively for some time, finally decided to bite the bullet and installed Cursor last month and it blew me away (I still have my OpenAI subscription though, I love it).

What had initially kept me away was the whole 500 requests limit pricing model. I don’t like granular data like that, and what drew me in was the new pricing model where I could get meaningful work done without worrying about micromanaging the number of requests I’ve made. This will upset the long time users, I know, but this is literally the reason I joined. I don’t want to see some chart or gauge or number; I just want to use the app and service, and if I’ve hit a limit on some model, I’ll switch to another or even use Auto until I need something specific. I don’t mind using it this way and quite enjoy it.

Also, using it in this way works. It keeps me from being lazy, and it encourages me to keep things within context and to make more efficient requests and use of the AI offerings.

Really the only thing driving me nuts right now is how the chat window agent will run half of the terminal commands in its own chat-window-embedded terminal and some in a tabbed terminal. I wish it would run everything in the external-to-the-chat-window IDE tab.

That is all, thanks for reading, have a nice day.

2 Likes

Working ‘with numbers’ gives you a prediction of how long you can work. Without this, you’re just gambling or playing with a toy without ensuring it will work for the next hour or two because of limits.
Without transparency and even visible negotiation with the clients, all other problems you described are so tiny…