retsibsi

32 ms·

151.
by retsibsi 9mo ago
This is fantastic! I feel like it's right at the sweet spot where "comically overengineered fun project" and "actually a great idea" overlap.
152.
by retsibsi 9mo ago
There's a story (true AFAIK, but it's the kind of thing I can easily imagine having been debunked) of Pepsi winning in direct taste-test comparisons against Coca Cola, but only when the test was done with small quantities. Apparen
153.
by retsibsi 9mo ago
I don't know if the parent comment has been edited, but in its current form I read it much differently from you! It seems like fair criticism without any added snark or contempt. I don't want hostility or gratuitous negativity, bu
154.
by retsibsi 9mo ago
Partly, the issue isn't "they would have to learn about Linux, and that's bad", it's "they would have to learn about Linux, and they wouldn't want do that, and so they would get frustrated and quite likely
155.
by retsibsi 9mo ago
The guy who made the video is (was?) a pilot, so I think he's as good a primary source as we would often deem acceptable for something like this.
156.
by retsibsi 9mo ago
> So, in other words, you want the cargo, but without the tiring "pay for it" aspect? I'm not trying to be difficult, but I'm not quite sure what this means! In literal terms, I want the ability to read tweets, see th
157.
by retsibsi 9mo ago
When the information is "what exactly did person X tweet", then yeah. I'm in favour of mostly avoiding X, and I make a point of not spending significant time on it or actively participating. But sometimes, I want to follow a
158.
by retsibsi 9mo ago
I think you're either over- or underthinking this. I don't want to have an X account, but I do sometimes want to follow a link to a specific tweet and be able to view the surrounding context. So sites like xcancel and nitter are u
159.
by retsibsi 9mo ago
There was definitely a lot of stupid stuff happening. IMO the clearest accurate way to put it is that it was overhyped for the short term (hence the crazy high valuations for obvious bullshit), and underhyped for the long term (in the sense
160.
by retsibsi 9mo ago
Agreed! I don't think it undermines your original point, and IMO the linked site could do some good by giving people a better intuitive sense of just how low the odds are.
161.
by retsibsi 9mo ago
This is interesting but IMO it's very likely to be chosen more often than average. If you choose a random number, then for each other player your chance of picking the same numbers as them is the same as your chance of winning: in the
162.
by retsibsi 9mo ago
> It has nothing to do with "raytracers are well-represented in the training set" though. I find it so strange when people get overly specific in an attempt to sound savvy. You should be able to easily think of like five other
163.
by retsibsi 10mo ago
The AI summary is good enough often enough that it's tempting to rely on it. But, at least as of a week or two ago when I finally decided to block it, it still gets things badly wrong. (Sometimes seemingly inexplicably, but often becau
164.
by retsibsi 10mo ago
Are you sure he actually has that address? I get lots of emails mistakenly sent to me, some via a dotted version of my address, but I'm pretty sure those people (or the ones trying to contact them) have just misremembered or typoed the
165.
by retsibsi 10mo ago
> Then just use 0.75 or 0.5x speed? I think this often sounds unsettling (like the reader is drunk or otherwise impaired), and anyway the listener doesn't need more time to recognise each individual word -- they want time to take in
166.
by retsibsi 10mo ago
For what it's worth, I found (the start of the first book of) the Expanse to be this in a bad way, but the Martian to be this in a good way. I can definitely see why some people would find The Martian annoying too, but it feels more li
167.
by retsibsi 10mo ago
> the more votes and comments it has, the higher the score Unless this has changed, I don't think comments contribute positively to the score. Apparently they (used to?) contribute negatively if there are more comments than upvote
168.
by retsibsi 10mo ago
I don't think I missed the point; my point is that LLMs do something more complex and far more effective than memorise->regurgitate, and so the original analogy doesn't shed any light. This actor has read billions of plays and
169.
by retsibsi 10mo ago
> Imagine an actor who is playing a character speaking a language that they actor does not speak. Due to a lack of time, the actor decides against actually learning the language and instead opts to just memorise and train how to speak th
170.
by retsibsi 10mo ago
What's your specific concern here? I certainly wouldn't want to, e.g., give young kids unmonitored use of an LLM, or replace their books with AI-generated text, or stop directly engaging with their games and stories and outsource
171.
by retsibsi 10mo ago
It's my impression that the shift from 'basically normal government' to 'authoritarian nightmare', when it happens, tends to happen quite abruptly rather than via the ratchet effect/frog-boiling. And there seem
172.
by retsibsi 10mo ago
Let's suppose the cause really is as simple as "parents can't be bothered to parent". By default, this will continue to be the case. And realistically we're not going to fix it by telling bad parents to please start
173.
by retsibsi 10mo ago
> And, from now on, if you ever criticize the government you will lose your bank account and your job. We could have a real conversation about tradeoffs (and maybe this one isn't worth it!) but not if you just assume/pretend th
174.
by retsibsi 10mo ago
Does it happen much with non-Claude models? If someone genders ChatGPT, it makes me worry that they're taking the character it's playing too literally. But if someone genders Claude, that seems pretty normal, given that it has a m
175.
by retsibsi 10mo ago
> That model does not exist. It does (unless the previous comment was edited? Currently it says Opus 4.1): https://www.anthropic.com/news/claude-opus-4-1 . You can see it in the 'more models' list on the ma
176.
by retsibsi 10mo ago
I love Warsow but am I right that it's very hard to find an opponent these days? I just checked https://arena.sh/wa/ and there are 4 non-empty servers, but most (maybe all) of the players seem to be bots.
177.
by retsibsi 10mo ago
They are wrong, though. Locks also stop people who would happily commit an opportunistic theft but who lack the necessary tools or skills, people who would trespass if they could retain some plausible deniability ("oops, I didn't
178.
by retsibsi 10mo ago
Not a unicycist (or a physicist), but I wonder if it would throw off your balance by changing the centre of mass?
179.
by retsibsi 10mo ago
In the end the analogy doesn't really work, because 'eternal September' referred to what used to be a regular, temporary thing (an influx of noobs disrupting the online culture, before eventually leaving or assimilating) beco
180.
by retsibsi 11mo ago
Not this time: > The global leaderboard was one of the largest sources of stress for me, for the infrastructure, and for many users. People took things too seriously, going way outside the spirit of the contest; some people even resorted

More ›