I really struggle to get past AI writing. Aside from feeling disrespected, something about that style is just so tiring to read… the marketing style of setup and reveal, repetition, only works in small doses
I was just about to comment the same. It's incredibly rude to just dump AI slop on people and expect them to parse it and suffer through the grating cadence that AI writes in.
Interesting! I'm in a similar boat, supporting windows and macOS out of one swift codebase. No MP4 issues, but it's been interesting seeing how's much code can and cannot be shared, and how much AI needs to be cajoled into writing code that's shared vs reimplementing it twice.
Really? I’ve had no trouble at all directing an agent harness to simply maintain multiple UIs (in my case, Android, macOS + iOS, and a TUI, the latter mostly to ease debugging).
I really struggle to get past AI writing. Aside from feeling disrespected, something about that style is just so tiring to read… the marketing style of setup and reveal, repetition, only works in small doses
I was just about to comment the same. It's incredibly rude to just dump AI slop on people and expect them to parse it and suffer through the grating cadence that AI writes in.
I already have to read the output from a fleet of agents every day for my job. Sometimes I just want to read what a human wrote.
I have never wanted to read ai outside of me asking it something specific. If someone writes a blog post using it I consider it a sin.
Interesting! I'm in a similar boat, supporting windows and macOS out of one swift codebase. No MP4 issues, but it's been interesting seeing how's much code can and cannot be shared, and how much AI needs to be cajoled into writing code that's shared vs reimplementing it twice.
Really? I’ve had no trouble at all directing an agent harness to simply maintain multiple UIs (in my case, Android, macOS + iOS, and a TUI, the latter mostly to ease debugging).