sfink

26 ms·

331.
by sfink 1y ago
This makes intuitive sense, to the extent that I'm surprised the number 250 is so high -- surely there are things LLMs are supposed to know about that have only a handful of instances in the training data? (Note that if the study found
332.
by sfink 1y ago
I'm not going to argue the point, I'm no climate scientist. But Wikipedia https://en.wikipedia.org/wiki/Contrail disagrees with you as well, so you'd need to argue with their sources too. "In genera
333.
by sfink 1y ago
The article claims that the warming effect from bouncing heat back down is overall larger than the cooling effect from bouncing heat back up. If you disagree with this assertion, you'll need to say why, not just call someone a liar. Th
334.
by sfink 1y ago
Yes, because those don't capture any lexicals, so they're hoisted as usual. My solution would only stop hoisting closures that capture lexicals, with the idea that capturing a lexical means that the closure's identity is now
335.
by sfink 1y ago
Ah, fair, I didn't actually pay attention to which example you were referring to. That example is specifically about `var` being terrible, not `let/const`. I was really using your comment as a jumping off point for my rant. I woul
336.
by sfink 1y ago
Actually, let/const do the opposite of adding flexibility. Simple example: if you have a REPL, it has to cheat (as in, violate the rules of the language) in order to do something sensible for let/const. Once you do `let x;`, you c
337.
by sfink 1y ago
That's not how lexical scope works anywhere but in JavaScript. Or rather, it's the interaction between "normal" lexical scope and hoisting. In a "normal" lexically scoped language, if you tried: function
338.
by sfink 1y ago
Because I am seeing how this is playing out in classrooms. Tons of kids are requesting accommodations. Some need those accommodations, some don't, and the ones who do often don't need all of the accommodations they're getting
339.
by sfink 1y ago
I agree completely with this comment, though most often I see it for ADHD. It's a level of nuance that people don't seem to be able to handle, though. People want to be on either the "just suck it up, losers" side or the
340.
by sfink 1y ago
Oh wow, I was just feeling grateful that I don't have misophonia, but then you had to mention chewing noises. I had that for decades. It wasn't all the time, but it didn't have to be loud or open-mouth or anything. It would j
341.
by sfink 1y ago
If a company professes to be supportive of neurodivergence, it means either (1) they're supportive of neurodivergence; or (2) they are hostile to neurodivergence and have gotten into trouble for it, so have strong motivation to claim t
342.
by sfink 1y ago
A dag represents a partial ordering. (Strictly speaking you could have redundant edges in a dag which wouldn't change the partial ordering, so dag != partial order.) upriver/downriver is the partial order relationship between no
343.
by sfink 1y ago
Maybe, though I'll note that this matches how I use regular knives with cheese & similar things: with a dull unscalloped knife I have to carefully go straight down to avoid tearing. If my knife is sharp (or temporarily oiled) and t
344.
by sfink 1y ago
Yes. For example, you could always say "give me a jpeg image file that is encoded as the bytes 255, 216, 255, 224, 0, 16, 74, ...". But that's just pointing out that the input to your "LLM" function includes the pro
345.
by sfink 1y ago
I like to think of myself as having modest desires, but this story does a good job of making me wonder: am I the fisherman or the wife? I don't want a palace, but I wouldn't want a filthy shack either. Which of my desires, for mys
346.
by sfink 1y ago
Given my own experience with depression, it makes some amount of intuitive sense. For me at least, "sadness" is not wrong, but other than that it doesn't describe the experience very well at all. When I'm really down, I
347.
by sfink 1y ago
> I do like evolutionary explanations for a lot of human behavior, but this one feels a little too pat to me. I agree. > Patience, "biding our time", "hunkering down", etc. are actual emotional states we can experi
348.
by sfink 1y ago
First, contracts often come with usage restrictions. Second, this article is incredibly dismissive and whiny about anyone ever taking safety seriously, for pretty much any definition of "safety". I mean, it even points out that An
349.
by sfink 1y ago
Sure, that happens, but that's also pretty clear to the job seeker. Don't try to BS your way past that one. More commonly, that list of requirements comes from the recruiter quizzing the developers on what they need, and they thro
350.
by sfink 1y ago
Strict honesty here has always been a losing proposition. The "requirements" section of a job posting has almost never been accurate. It's more of an image they're painting. An honest applicant is one who reads the whole
351.
by sfink 1y ago
I wouldn't label this as "support for Israel"/"against Israel". One can support Israel without supporting Israel's current approach. Many within Israel are not happy with Netanyahu's methods, and pres
352.
by sfink 1y ago
It's a callout to the concept of "dramatic tension". See also all the talk of stories, movies, etc. https://fiveable.me/key-terms/introduction-creative-writing/... is as good a source as any. To me,
353.
by sfink 1y ago
> The premise is bullshit... there were LOTS of competing options when React first came out... Good thing that wasn't the premise, then. The article is specifically looking at reasons for React's success other than its techni
354.
by sfink 1y ago
By that logic, Larry Ellison would have a soul. You've disproven panpsychism! Congratulations!
355.
by sfink 1y ago
This covers part of the problem, the part where your tests are enough to indicate whether the changes are good enough to keep. In that scenario, relying only on fast-forward merges is good enough. One trickier problem is when you don't
356.
by sfink 1y ago
Why not put the loop on the case?
357.
by sfink 1y ago
Nope. I don't know why just checking the tags during speculation wouldn't stop Spectre V1, at least for cross-type accesses? I mean, it's not that simple because your program won't crash if speculation has mismatched tag
358.
by sfink 1y ago
To be fair, they didn't claim it to be a meaningful attack deterrent. They said "success...in increasing exploitation complexity". Sure, the whole sentence is a bit of a weird mess. Paraphrased: it made exploits more complex,
359.
by sfink 1y ago
I understand this might seem unlikely given that it's working fine as 64-bit, but that crash dump makes me want to suggest running a memory tester. It has "Possible bit flips max confidence" of 25%. Ignore the exact percentag
360.
by sfink 1y ago
No. Let's say our initial boost got it up to 0.1c. After 20 years, it's gone 2 light years. If we make our space-based laser aperture really big, let's say 1km, then the light reaching our probe is something like a 25000km wi

More ›