←back to thread

Open-source Zig book

(www.zigbook.net)
692 points rudedogg | 3 comments | 16 Nov 25 19:44 UTC | HN request time: 0.623s | source
Show context
eibrahim ◴[16 Nov 25 23:53 UTC] No.45949622[source]▶
So many comments about the AI generation part. Why does it matter? If it’s good and accurate and helpful why do you care? That’s like saying you used a calculator to calculate your equations so I can’t trust you.

I am just impressed by the quality and details and approach of it all.

Nicely done (PS: I know nothing about systems programming and I have been writing code for 25 years)

replies(13): >>45949682 #>>45949740 #>>45949749 #>>45949796 #>>45950174 #>>45950334 #>>45950401 #>>45950481 #>>45951133 #>>45951367 #>>45951767 #>>45953838 #>>46094809 #
gassi ◴[17 Nov 25 00:14 UTC] No.45949749[source]▶
> Why does it matter?

Because AI gets things wrong, often, in ways that can be very difficult to catch. By their very nature LLMs write text that sounds plausible enough to bypass manual review (see https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-s...), so some find it best to avoid using it at all when writing documentation.

replies(2): >>45949854 #>>45950125 #
DrNosferatu ◴[17 Nov 25 00:35 UTC] No.45949854[source]▶
Humans get things wrong too.

Quality prose usually only becomes that after many reviews.

replies(5): >>45949983 #>>45950389 #>>45950602 #>>45950870 #>>45951344 #
1. righthand ◴[17 Nov 25 02:34 UTC] No.45950389[source]▶
That’s fine. Write it out yourself and then ask an AI how it could be improved with a diff. Now you’ve given it double human review (once in creation then again reviewing the diff) and single AI review.
replies(1): >>45951284 #
2. maxbond ◴[17 Nov 25 06:30 UTC] No.45951284[source]▶
That's one review with several steps and some AI assistance. Checking your work twice is not equivalent to it having it reviewed by two people, part of reviewing your work (or the work of others) is checking multiple times and taking advantage of whatever tools are at your disposal.
replies(1): >>45955934 #
3. righthand ◴[17 Nov 25 17:42 UTC] No.45955934[source]▶
The point was to bookend your human review around automated. Not stamp out a blueprint.