Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
smallnamespace
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
31.
▲
by
smallnamespace
2y ago
> using max(0, exp(x)-1) instead of exp(x) Won't this cause the gradient to vanish on the left half, causing problems with training?
32.
▲
by
smallnamespace
2y ago
The international code is also defining the key distinguishing factor of insurance: it makes the insured whole against a risk that they actually have . There are ways to bet on things where you don’t have that underlying risk: gambling, de
33.
▲
by
smallnamespace
2y ago
> What's wrong with just letting the premium take care of it? Offering a deal that nobody honest would take is a waste of time for everyone involved.
34.
▲
by
smallnamespace
2y ago
> why should it matter what others are willing to pay for it? Because the actual value of the item determines your incentive to commit fraud. If you insure a $10 pillow for $10, when you damage your pillow, you personally will definitely
35.
▲
by
smallnamespace
2y ago
> Credit reporting and collections. This has a loss rate of around 70-80% across the collections industry, which is an extremely strong disincentive to go this route since it's just highly inefficient. The high cost of the collectio
36.
▲
by
smallnamespace
2y ago
You are saying planning never works without even the ability to point to any specific cases. Why do you swallow your own ideology so uncritically? Does Soviet-style central planning work? It didn't seem to work well in the few societie
37.
▲
by
smallnamespace
2y ago
> That the favoured industry (or company) is doing well isn't necessarily a sign that the policy is overall good for the country's economy. You are answering specifics with generalities. If Taiwan didn't support and nurtur
38.
▲
by
smallnamespace
2y ago
High finance and crypto are two places where a high concentration of technical skills intersect with using those same skills to game our financial and legal systems.
39.
▲
by
smallnamespace
2y ago
> Would you be willing to set a deadline (not fuzzy dates) when my job is going to be taken by an LLM and bet $5k on that? This is just a political question and of course so long as humans are involved in politics they can just decide to
40.
▲
by
smallnamespace
2y ago
> asking is it worth the actual effort If prompting ability varies then this is not some objective question, it depends on each person. For me I've found more or less every interaction with an LLM to be useful. The only reason I
41.
▲
by
smallnamespace
2y ago
But if I save money in the US I can decide to spend it anywhere else in the world later, including places that are much cheaper than Toulouse. If I make money in Toulouse then I had better really love Europe, since I will have limited res
42.
▲
by
smallnamespace
2y ago
I disagree with the main thrust of the article, that the Visitor pattern is primarily a primitive way to do pattern matching. For one, the Visitor pattern (as written in the article) fails to fulfill a core feature of Rust pattern matching,
43.
▲
by
smallnamespace
2y ago
> Nobody talks like this, it's super awkward Maybe people who communicate this way are his tribe. Birds of a feather, etc. Seen in this light, these questions would be a highly effective tool.
44.
▲
by
smallnamespace
2y ago
Yes, looking it into it more, Lit-html is doing a few things to make it all work: 1. It leans on support for the <template> element from browsers to hold fragments of HTML 2. For most use cases, it has a more restricted programming mo
45.
▲
by
smallnamespace
2y ago
A a high level there's not much difference between template literals and JSX, they are both syntax-sugary ways to represent trees of essentially function calls. > efficiently update just the changed portions Since actually applying
46.
▲
by
smallnamespace
2y ago
How is that done without a vdom?
47.
▲
by
smallnamespace
2y ago
> I never understand why people don't just test this? One reason might be that you get the benefit of feeling superior to LLMs without any cost or risk of finding yourself mistaken.
48.
▲
by
smallnamespace
2y ago
It doesn't have anything to do with projections. Let's say you live in a non-flat space. You come up with the idea of a "circle" with the usual definition: the set of points on the same plane equidistant from a central p
49.
▲
by
smallnamespace
2y ago
Have you considered short paying him again through the same entity? After all, that's one use case of limited liability ;)
50.
▲
by
smallnamespace
2y ago
Drivers undergo a vetting process by Microsoft (at least, the ones that are signed and offered as updates) = quality control through their control of the channel. This has been the case at least since the late 90s if I remember correctly.
51.
▲
by
smallnamespace
2y ago
The grokking papers show that after sufficient training models can transition into a regime where both training and test error gets arbitrarily small. Yes, this is out of reach of how we train most models today. But it demonstrates how even
52.
▲
by
smallnamespace
2y ago
Unfortunately it's also hard to buy (find) people who don't know how to sell.
53.
▲
by
smallnamespace
2y ago
> to match the observed motions of the celestial bodies The point is that even with current observational data there's no reasonable distribution of dark matter that correctly explains all evidence that we have. Your intuition that
54.
▲
by
smallnamespace
2y ago
No, physicists do not think that space is like a tinily subdivided grid. Lots of physics would break if that were the case, including QM. We have a lot of hints that the amount of information in any given space might be bounded, but yet t
55.
▲
by
smallnamespace
2y ago
Yes, but that doesn't explain how OpenAI able to walk and chew bubble gum at the same time, unless you think there is some extra spark that Google is lacking.
56.
▲
by
smallnamespace
2y ago
Oh I fully agree the situation is precarious. It's just fun seeing people independently start waking up to that fact. I mean, you could probably see the latter being a cause of the former (if people generally realized the world was no
57.
▲
by
smallnamespace
2y ago
These observations while good are probably old news (about a decade old) by now. The US has viewed tech as strategic and a realm of competition with China since at least the late Obama admin (so 2015 or so). Semicon as a key strategic goal
58.
▲
by
smallnamespace
2y ago
VCs will go along with or sometimes even encourage founders to take a little bit out, but employees rarely don’t have the same level of bargaining power.
59.
▲
by
smallnamespace
2y ago
Anyone can make stuff up. I'll happily make up 19 different things and tell it to you with a straight face, all before lunch. It seems like the whole point of elevating a field of activity and dignifying it with a fancy term like jour
60.
▲
by
smallnamespace
2y ago
> Remember when everyone was expected to have an opinion on Syrian children, within 24 hours? Honestly, no, I think I may be in a different filter bubble than yours.
More ›