Don't be nice, be honest: would you use my app or do you think I'm wasting my time?
1. Fill a form with your google and insta links
2. Print & display a QR code in your store/business
3. Ask people to scan to review/follow instead of manually looking your business up
@alexandr_wang tough product decision - I appreciate FOCUS and optimization on some areas of coding, e.g. not 3D
but I get that "visual benchmarking" is useful and fun for humans. A future 'focused' model, e.g. no 3D, could maybe have a name which helps manage expectations?
@narigato315066@alexandr_wang it's there but it's buggy, at least on Muse Code 1.4.6 - go to /settings and set the Reasoning effort. I got to enable 'max' that way
@The_MaxWell95@AIatMeta Q: how do you know it's doing that routing? All I know is on Muse Code 1.4.6 the 'contributor' tag went away from the bottom status like AND from the model picker. It's now in /settings -> Contributor mode
@Basileus_X@jeffholden I rarely come to twitter. And scrolling through a mile of silly hype to find this first CORRECTLY cynical comment reminds me why
@ThatRetiredDude@Y2cc4@cursor_ai@xai@OpenAI so far so good w Kimi K2.5 - it's more token efficient vs grok-code in the past few weeks so cost is not exploding
tho I'm on MEGA standard stacks (React running on node) so no wonder almost any model does the trick
@Y2cc4@ThatRetiredDude@cursor_ai@xai@OpenAI FWIW after literally MONTHS using grok-code-fast-1 almost exclusively in Cursor, I'm testing Kimi K2.5 - more expensive but maybe will be more efficient w Tokens and come up the same or even better
@Y2cc4@ThatRetiredDude@cursor_ai@xai@OpenAI man, you're doing the lord's work by reaching out and posting about it, AND also trying out previous versions of Cursor - which I was going to waste time trying today. Thanks man. Let's hope there's a fix soon, it does seem related to tool calls
@ThatRetiredDude It's been getting worse and worse over the past few hours, fully unclear to me whether it's @cursor_ai or @grok that is at fault - I think Grok itself released an update to the model but seems like it was one of those "dumb it down" updates
@Y2cc4@ThatRetiredDude@cursor_ai@xai@OpenAI It's been dismal for the past ~48 hours for me too. I'm talking literal insanity like getting stuck on endless loops or saying "I'm done - here's what I changed" and making 0 changes to the codebase
@raechellambert I quit creative writing when I realized computer code and more rigid/structured thinking came to me more easily
eventually I realized: I'm an avid reader and enjoy it tremendously, but writing fiction didn't make me happy