asgraham

8 ms·

31.
by asgraham 1y ago
I don’t know about your first point: at some point the three-year difference may not be worth the premium, as local models reach “good enough.” But the second point seems even less likely to be true: why will Claude code and Gemini cli alw
32.
by asgraham 1y ago
Fair enough. Cancer causation is just so weird, I'm going off the heuristic that stuff in your lungs that your lungs aren't prepared for is probably not great, especially with chronic exposure. And your lungs are really only prepa
33.
by asgraham 1y ago
I can’t find any complete numbers (most prevalent factors after smoking would be environmental, and therefore underreported, especially when as hard to detect as radon), but national health agencies tend to put the Radon section second, aft
34.
by asgraham 1y ago
Not to agree with GP, but I think it’s more accurate to say they’re saying “if validation is quick (to code), who cares how long a solution takes an AI because computation is cheap.” They’re not really making any claims about how quickly th
35.
by asgraham 1y ago
> It's actually hard to find a math tutor to proof your vector math program in your late 30s. They exist, if you know where to look and are willing to pay (source: me, or generally and probably more affordably wyzant.com)
36.
by asgraham 1y ago
What a wild thing to comment. a) “professional” writ large often encompasses academic research, even though it’s also often specifically used in contrast. Language is weird like that. Former and current academics like to have a term for all
37.
by asgraham 2y ago
re: the arxiv link Why is it that microarray true positive p-values follow a beta distribution? Following the citations led to a lot of empirical confirmation but I couldn't find any discussion of why . More to the point of this rebut
38.
by asgraham 2y ago
I knew it wasn't for inspectors of physical buildings (though that took a minute to accept; I thought it might be a joke), but I could not figure out what kind of buildings it would be. Was it some lisp thing I hadn't heard of
39.
by asgraham 3y ago
As a chronic premature optimizer my first reaction was, "Is this even possible in vanilla python???" Obviously it's possible , but can you train an LLM before the heat death of the universe? A perceptron, sure, of course. A
40.
by asgraham 3y ago
For me it makes it harder to read. I think it's because my skimming muscles are so tuned to both bold and first-sentence text that when presented such a stark opportunity, I naturally try to jump from one bold title to the next. But t
41.
by asgraham 4y ago
I'm absolutely behind both the message and call to action of this article; the evidence is irrefutable; this comment should not be interpreted as calling into question the headline at all: the underlying data looks solid. However. Why
42.
by asgraham 4y ago
It's hard to say what the hardest constraint will be, at this point. Imaging and scanning are definitely hard obstacles; right now even computational power is a hard obstacle. There are 100 trillion synapses in the brain, none of which
43.
by asgraham 4y ago
"...a sorta-popular JavaScript package. I say “sorta” because it’s used by lots of people, but it’s not pervasive. It had 105 million downloads in 2022." Is there any standard of popularity where 100 million+ downloads in a year i
44.
by asgraham 4y ago
They can't possibly be working with genomes spanning the past 250k years--- the oldest known human remains are only estimated to be ~230k years old, and I doubt they have parent/child trios nicely spanning the intervening few hund
45.
by asgraham 4y ago
The author of The Wandering Inn [1], a web serial, has been trying to write less the past few months (the past year?). They peaked around 80k words/week, and have been aiming closer to 30-40k words/week with more breaks. They&#x27
46.
by asgraham 4y ago
Sure, this is ideal given unlimited time and money, but hiring dedicated devs for tooling is going to be expensive. And once they fix "the problem," they're still on payroll. Yay they'll keep fixing problems and improv
47.
by asgraham 5y ago
"A feeling of safety comes from within." --- This is an actively counterproductive statement. Just look at the impetus behind "defund the police": the real police violence causes people of color to feel unsafe. On the fl
48.
by asgraham 5y ago
I'm torn about this sort of technology. On the one hand, the best way to drive adoption is external recording. On the other hand, the data are sooooo coarse-grained. And on the third hand, we don't have anything resembling the the
49.
by asgraham 5y ago
> 'strengthening' neurons with a fiber like substance around it reduces the electrical activity required to achieve the same neural-result. I think you're talking about myelination, which is a result that's been aroun
50.
by asgraham 5y ago
Fair point-- I don't know how the data defined a "sale." However, the numbers are almost identical for print books, just a little lower. But same order of magnitude.
51.
by asgraham 5y ago
Isn't it saying that ten titles each sold between 500,000 and one million copies? So collectively they sold 5-10 million copies? (I really enjoyed the article, by the way, once it was past claims about this data-- the points about fanf
52.
by asgraham 5y ago
These are two completely different statistics: 1) last year, new releases sold poorly, perhaps unexpectedly (NYT article's claim); 2) last year, a relatively small portion of all currently released titles sold a lot of books, and mos
53.
by asgraham 8y ago
I would highly recommend reading this recent post in the Julia forums [1]. It's from a "relatively new" Julia user explaining why Julia is actually a very good general purpose language, e.g. for string manipulation of scraped
54.
by asgraham 8y ago
Oh absolutely! I totally understand you're not saying that's what's happening here, but just to clarify for other readers, I'd say in this case there are a few reasons we don't need quite that level of suspicion: 1.
55.
by asgraham 8y ago
It's called multiple testing. You're not guaranteed to get a misleading result, but it would be expected. The definition of the 5% significance level used in p-value testing is that given 100 junk datasets (distributed similarly),
56.
by asgraham 8y ago
Similarly, quantification is separate from being rational. In your simple example "sum these assigned numbers and pick the highest option," how did you assign those numbers? In most real life scenarios, qualities are difficult to
57.
by asgraham 8y ago
Does workspace grid [1] work on 18.04? Not sure how to get wrapping, though. [1] https://extensions.gnome.org/extension/484/workspace-grid/
58.
by asgraham 8y ago
I began making the change in January and was fully morning-person by mid-February. I'm thoroughly enjoying it, and would highly, highly recommend it. The key for me was having both a reason to wake up, and a reason to go to sleep. To h
59.
by asgraham 9y ago
A quick recap of the basics (multiple testing appears at the X): Precisely, what a p-value means is: "Assuming the null hypothesis is true, the probability that I would get this test statistic from the null distribution is p," whe
60.
by asgraham 10y ago
Two points: First, it's important to keep in mind the difference between artificial "neurons" and real neurons. Real neurons, with their complicated dendritic arbors, are much more complicated than anything you'll see in

More ›