I would probably show up in their metrics as an active user and one of the 95% but I barely use the product. I have a Pro subscription which I use for personal projects but I do very little, maybe using it once a week for a short session. At work I use Cursor via a corporate account.
I imagine there are lots of people like me who have a subscription to be aware of the product and do some very light work, but the "real" users who rely on the tool might be badly affected by this.
> Even after completing the tasks with AI, the developers believed that they had decreased task times by 20%. But the study found that using AI did the opposite: it increased task completion time by 19%.
This in particular is very interesting to me. I haven't read the study yet but this makes me consider my own use of AI - I often feel like it is speeding me up, but is it really? Can I measure it in a better way?
> [...] Why walk into a store in Soho and see what’s on offer when you can stay home and scroll the entire inventory from the comfort of your couch? Why go to the library to find books about a topic that interests you when you can look it up on Wikipedia in two minutes and move on with your day?
> Instantaneous access to everything obviously comes at a cost. The cost being that we all behave like demented Roman emperors, at once bored and deranged, summoning whatever we want at any time.
Even without social media, we still would have the instant gratification that the author proposes as a problem.
It's a macOS app and I've found it great. However if given an ASCII diagram, you cannot edit it with the same ease as creating a new one (e.g. reflowing text or resizing boxes).
I really like the idea of having the mermaid source and the ASCII diagram together, so you could use the source to change the diagram if needed. But I feel that would feel cluttered to have both in a plain text file or comment, where ASCII diagrams shine.
I imagine there are lots of people like me who have a subscription to be aware of the product and do some very light work, but the "real" users who rely on the tool might be badly affected by this.