programmergrind
Region: US
Friday 18 September 2026 01:06:11 GMT
4882
194
11
2
Music
Download
Comments
Jakob Schulz :
sorry but I don't agree with the take, LLMs don't trade code quality for the point of shipping faster, WE trade code quality for the sake of shipping faster. Your agents could review your manual code and give helpful insights, or help us to increase code quality, we decide not to because we want to ship faster. I don't think putting the blame on the LLMs is wrong. We are the problem. We always are.
2026-09-18 01:41:09
14
Futureproof Foundations :
agreed but man I am already midway to earning a certificate in js and py add to that 3 more certificates and I'll get a full-stack one. is that not valuable ? should I ditch the plan ??
2026-09-24 15:59:51
0
cy356 :
honestly it's impossible to review the code at this stage
2026-09-28 15:24:48
0
Shakes932 :
A lot of that can be mitigated with a standardization pass every sprint or so. It doesn’t solve the whole problem, but 90% of bugs happen because something was bolted on in a weird way, that doesn’t match the architecture. It really helps with code readability as well.
2026-09-18 01:36:28
1
Lowspeclabs :
Your presumption is flawed, the models were trained on the average programmers code, as a result most code made by the ai is mid. This is a self perpetuating loop, because people will only get worse as the agents get marigonally better
2026-09-18 20:54:04
2
hombreEmpastillado :
if you set up a good foundation, the agents do a much better job. planning is everything. reading the code doesn’t really take that long either. saving 10 hours or so a week to not read the code is just dumb.
2026-09-18 22:40:41
1
kevnotfound :
ai orchestration
2026-09-18 11:42:32
0
To see more videos from user @programmergrind, please go to the Tikwm
homepage.
