You need to ensure the whole system is optimised for LLM. Getting human written code and trying to run like this will be a failure since you'll be missing essential guidelines particularly on the domain-specific design fundamentals that are acquired in the beginning in a human feedback loop before the automation really kicks in
My workflow these days is basically ultracode + PRs. I do control the features that go but barely need to check the code due to the amount of checks: lint, TS, Local AI visual Model for visual check for screen frame comparison (Mac M5 48GB mem), Pixel/Screenshot regression test for basic UI, code coverage/c8 to avoid trash code without tests, etc. etc.
Essentially the tests are just for the AI to remember what they did (as in the description + break if they change smth unintentionally).
I remember really enjoying the first season and then feeling like the series went entirely off the rails after that... but still a really good first season IMO, especially by 2006 standards (there were some great TV series out by then, but few and far between).
In any case, absolutely tragic for her to have died at only 36. Life is vanishingly short.
Heroes was a series where the writers strike that happened afters season 1 (or 2?) really made a difference. Ok, I should have formulated that better but I am no writer either.
RIP. Too young.
I loved it at the time, it was one of those shows everyone would be talking about the next day. Having to wait a week to find out what happened next was a delightful frustration.
Sadly, did not hold up on a recent rewatch with my wife.
This is a cached version from a Zhihu answer from China Aerospace Science and Technology Corporation that was deleted a few minutes ago, so that's why I'm linking the cached version. Right click -> Translate to English on Chrome should work.
Disclaimer: You need to login to be able to save the imports and it's a 14 days trial. But you can use the code "HN" on checkout for a year with 50% off if you're keen.
reply