kmoser

4 ms·

1.
by kmoser 5d ago
This comes just in time. My app has been using Amazon Mechanical Turk for left-pad queries, and that service is shutting down in less than two weeks.
2.
by kmoser 7d ago
> a 10 years old kid and give him a PC with Microsoft Paint You'd run afoul of child labor protestors. I'd just prompt the AI to generate a poster in the style of a 10-year-old with Microsoft Paint.
3.
by kmoser 8d ago
My understanding is that wear-leveling on SSDs leaves forensic traces which allows hidden volumes to be detected.
4.
by kmoser 8d ago
That's the theoretical worst that should happen. The theoretical worst that could happen is far worse than that.
5.
by kmoser 8d ago
Sometimes people get searched randomly, and occasionally it correlates to their skin color.
6.
by kmoser 11d ago
By design, even the same LLM, when asked the same question multiple times, will almost never hallucinate in the same way. By that (flawed) logic, you could have the same LLM check itself.
7.
by kmoser 15d ago
Do you mean the part where it says they use signals to determine your age?
8.
by kmoser 16d ago
Are you saying you've never had an experience even somewhat like this while working with Claude, or any LLM for that matter?
9.
by kmoser 17d ago
It's reductio ad absurdum , satirizing the Claude experience.
10.
by kmoser 18d ago
Distrust of the government comes in different levels: you can be just distrustful enough to be skeptical yet accepting, and so distrustful that you actively try to vote the bums out. I don't think the polls distinguish between them.
11.
by kmoser 19d ago
Tufte would love this interface; it packs a ton of info, but is easy to use and understand.
12.
by kmoser 20d ago
> But when a hyperscaler does it, they just get inflated stock prices. This reminds me of the early days of AOL when they received so much traffic that their (dial-up) lines became inaccessible. In a rational world, their stock would hav
13.
by kmoser 20d ago
This article is quite well done, especially from a programmer's perspective, but falls flat (no pun intended) for me because of my terrible ear for pitch. > Here are five notes. Every one is double the frequency of the one before it
14.
by kmoser 22d ago
Serious question: how do you prove that the Lean interpreter itself (not to mention the toolchain built around it) is error-free? Isn't this turtles all the way down to some degree?
15.
by kmoser 22d ago
It looks more like awards for events, e.g. "Most Creative and Engaging Event" and "Best Event Series (over 1800 attendees)," none of which seem to have anything to do with the quality of any awards given at those events.
16.
by kmoser 22d ago
> Gee, you need a way to rank the players to decide. Do you also need to randomize a neural net before training it to play Tic-tac-toe? https://simple.wikipedia.org/wiki/Hacker_koan
17.
by kmoser 22d ago
To some degree, yes, except that Anderson's sets and costumes are highly stylized to the point where they come across as contrived and artificial, which is at odds with the naturalistic deadpan aesthetic.
18.
by kmoser 22d ago
From the article: "During a dinner conversation at a gaming convention in 2012, Eric Harshbarger was asked by a board game designer if he could come up with dice that would determine who goes first — without the possibility of a tie. T
19.
by kmoser 26d ago
It's also discussed in this article: https://time.com/3802766/r-crumbs-snapshots-source-material-...
20.
by kmoser 26d ago
Consistency of abbreviation style is important, too. It drives me crazy that the Google Maps API uses lat/lng rather than lat/lon.
21.
by kmoser 27d ago
Ok, but who do you think fixes the bugs? There must be countless devs out there who are slaves to fixing Jira tickets all day, so I still don't understand OP's assertion that full-time bug fixer isn't a thing.
22.
by kmoser 27d ago
To some degree, yes. Not to the point that I can't or won't do it, but I would certainly prefer books to be printed much wider. I can mitigate some of the issue by keeping my index finger along the left margin as I read, so I have
23.
by kmoser 27d ago
I had this same discussion on another HN thread a while back. I prefer lines to run as long as possible; I find it difficult/distracting to move my eyes to the beginning of the next line every few seconds. I'm sure I'm in the
24.
by kmoser 27d ago
> At times, I've tried to see if I can get a job somewhere where I just fix quality issues all day. This has never panned out, due to some combination of this not being a very high priority and it also not being a normal role that c
25.
by kmoser 29d ago
Yep, that was it. This is not the first time a website failed to perform as advertised because of this. To be clear, I'm blaming the websites or their creators; this is not a bug, but rather a feature working exactly as intended. But I
26.
by kmoser 1mo ago
Are these transitions supposed to be animated, i.e. the video changes size over a few seconds? Because it just jumps instantly to the new size for me on both Firefox and Chrome.
27.
by kmoser 1mo ago
> In regard to “cyber”, there is, IMO, no valid reason whatsoever to train a model to autonomously create exploit chains. A valid reason would be to find those exploit chains so you can fix them. Of course, the model should be sandboxed
28.
by kmoser 1mo ago
OP's scenario specified only "AI Agent" and did not describe how it was trained or what its parameters were supposed to be. You're assuming that the tool was designed such that it couldn't break the law, and therefo
29.
by kmoser 1mo ago
Said another way: "Your engineers were so preoccupied with whether or not they could , they didn’t stop to think if they would ."
30.
by kmoser 1mo ago
"Who gets prosecuted?" depends on the size of the perpetrator and victim (lone individual or employee of large corporation), egregiousness of the violation, and either financial appetite of the victim to bring a civil lawsuit or t

More ›