unsaved159

6 ms·

1.
by unsaved159 5mo ago
Quite the opposite. Review is the bottleneck. LLMs generate more code, hence more cognitive load on my mind to comprehend and review it all. Review isn't optional if it's a production system that other people use. After I see how
2.
by unsaved159 5mo ago
What use cases did you find this approach works for, and what doesn't? Any observations on what topologies work better? I tried doing the same for the cases of maintaining OSS projects. So far, best I could manage is to get the agents
3.
by unsaved159 5mo ago
I am curious what are you using multiple agents for? In my experience, without supervision even the most advanced models degrade quickly.
4.
by unsaved159 5mo ago
Nice idea! Though it's hard to find attention to read the text between jumping the obstacles :) maybe something where text would naturally be a part of the gameplay?
5.
by unsaved159 6mo ago
Nice idea for a diary app. "Can't edit yesterday" is off-putting for me. Such a constraint should not be something a software imposes on you, should be a person's mental policy, if they so wish. I want to have full contr
6.
by unsaved159 6mo ago
Text is a limited medium for communication inherently. Doesn't make too much difference if some words were cleaned. Overthinking hidden meanings of a text message is a good way to misinterpret people. Just ask to meet face-to-face when
7.
by unsaved159 6mo ago
Not really AI problem, more like garbage coworkers.
8.
by unsaved159 6mo ago
Literally never in my life did I receive anything like that website suggests via email or DMs. Curate your social circle is the answer.
9.
by unsaved159 6mo ago
They get a magic wand to turn their words into software, and still complain the wand is not their favourite colour.
10.
by unsaved159 6mo ago
Not clear to me why need this. You can just write a markdown spec without any side projects, then tell an agent to code it.
11.
by unsaved159 7mo ago
Why not context 7?
12.
by unsaved159 9mo ago
Ironically the getting started guide (quite long) is still to be executed by a human, apparently. I'd expect an LLM first approach, such as, "Insert this prompt into Cursor, press Enter and everything will be installed, you'l