dymk

5 ms·

61.
by dymk 2mo ago
Understanding how to write good prompts (prompt engineering) is still very much a relevant skill if you want to effectively use LLMs. Harnesses aren't magic.
62.
by dymk 3mo ago
The weights are the output artifact, the training corpus and system are the source.
63.
by dymk 3mo ago
If that’s the case I still don’t get it. 640k was enough in 1981, same as how 150 (or 9k) tok/sec would suffice for 2026 The comment seems like the nerd equivalent of 6 7
64.
by dymk 3mo ago
This post describes ideas found in existing agent orchestrator systems since they've existed. You'll find the same "top level agent that breaks a problem down into spec / plan / implement / verify" pattern
65.
by dymk 3mo ago
I am really confused about the point you're trying to make. 150tok/s is slow, but so is 9000tok/s? Or they're both fast? Or 150tok/s should be enough?
66.
by dymk 3mo ago
Oh, "you're holding it wrong". I can break basic user interaction with one line of CSS, does that mean I should?
67.
by dymk 3mo ago
This is why you don't override basic interactions like scrolling. Good luck ensuring that your setup matches everybody else.
68.
by dymk 3mo ago
It's completely broken if you use it with a mouse that has a scrollwheel.
69.
by dymk 3mo ago
Copying a block of multi-line text out of the agent doesn't work over ssh, for instance.
70.
by dymk 3mo ago
Eyebrows will raise at Anthropic / OpenAI, perhaps. Why are they charging 1000x more for the exact same business? The 'distillation is theft!' shtick won't last long. In fact, unlike Anthropic/OpenAI, these Chinese
71.
by dymk 3mo ago
No, you need to apply the sandbox to the codex process itself. Codex does not do this.
72.
by dymk 3mo ago
It’s in ads
73.
by dymk 3mo ago
People were absolutely upset with Github for pausing new signups. The token-based billing was a separate thing.
74.
by dymk 3mo ago
I remember people being so upset when GitHub did exactly this. It was the right move then and now!
75.
by dymk 3mo ago
The parts outside of the “Claude wrote this” guards seem plausibly written by a human, it’s just that 70% of the article is in the “Claude wrote this” guard.
76.
by dymk 3mo ago
Maybe, but remember what happened last time The Board tried to demote sama?
77.
by dymk 3mo ago
They're standard because everybody uses them. There are no big hiring CMS offers which don't integrate AI. I'm not saying this is good, but I am saying that the companies hiring have less of a choice in all this than you&#x27
78.
by dymk 3mo ago
"They" being the family; Robinhood couldn't settle with the guy who was wronged because he, ya know, killed himself first
79.
by dymk 3mo ago
…for emotional damage?
80.
by dymk 3mo ago
It’s all nostalgia mining
81.
by dymk 3mo ago
The pressure is “you can’t sell this here”
82.
by dymk 3mo ago
It’s a nice thought, but it’s probably not true as the AI becomes integrated into standard hiring tools
83.
by dymk 3mo ago
No it doesn't
84.
by dymk 3mo ago
source?
85.
by dymk 3mo ago
I'm in the PNW and I pay 11c/kWh (well, I would, if I didn't have solar). Seattle is 13c, King County averages 16c. Where are you paying 5c/kWh? That's exceptionally cheap.
86.
by dymk 3mo ago
Only once per compaction
87.
by dymk 3mo ago
Add "... unless the changes are trivial, docs-only, or typo fixes" to the "always verify with tests" instruction and see how that does
88.
by dymk 3mo ago
The article says that one possible reason the vaccine protects against dementia is incidental protection from other diseases. Getting chickenpox as a kid might protect you against shingles, but maybe not against those other diseases.
89.
by dymk 3mo ago
Why do you think that?
90.
by dymk 3mo ago
I'm a Ruby enthusiast - Sorbet is one of the best things since sliced bread to happen to the ecosystem. matz is pushing hard on static typing as part of the standard Ruby ecosystem as well.

More ›