Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
habitue
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
17 ms
·
211.
▲
by
habitue
3y ago
I'm really glad people are iterating on the abstraction for these systems. This is exactly the place where creativity is really useful
212.
▲
by
habitue
3y ago
nobody knows anything for a fact. But we have our priors. Our priors should reflect no radio transmissions from aliens, no non-blurry pictures of UFOs, etc etc
213.
▲
by
habitue
3y ago
The OP might be talking about a family office, but they might also be talking about how the family decides how much each member gets. That's a separate issue from how the money is invested and grows
214.
▲
by
habitue
3y ago
This seems like splitting hairs? They totally understand how modern neural networks work, they spend all day with them.
215.
▲
by
habitue
3y ago
Conversely, being too close to the problem can cause practitioners to miss the forest for the trees. I think in this case it's a false premise though. There are plenty of real practitioners like Joshua Bengio and Geoffrey Hinton who ha
216.
▲
by
habitue
3y ago
fair
217.
▲
by
habitue
3y ago
all the bad parts of sugar are from the sugar part of it, not some magic poison added by "refining"
218.
▲
by
habitue
3y ago
Tbf, unrefined sugar is just as bad as refined sugar
219.
▲
by
habitue
3y ago
Yeah, there are absolutely people who drink this much aspartame. Probably we should think of its danger level like Advil: normal sparse amounts are fine. But if you are doing it constantly you're going to have issues.
220.
▲
by
habitue
3y ago
Alternately: file formats and saving to disk are just a very primitive database where the cache is the size of the entire database (you load the entire thing into memory). Also you don't have good disaster recovery (see "I forgot
221.
▲
by
habitue
3y ago
Actually unpopular opinion: Existential risk from AI. I think people think "AI is overhyped" etc, but there is a difference between there being too many low value AI startups right now (a bubble) and whether a few top labs will pr
222.
▲
by
habitue
3y ago
Wow, they cloned twitter so well I didn't even notice it wasn't a Twitter link
223.
▲
by
habitue
3y ago
Not saying everything in this article is wrong, but I actually think the opposite is more true. Google + Wikipedia + chatgpt + youtube are imperfect sources of knowledge, but they are vastly better than the no-knowledge people had in the da
224.
▲
by
habitue
3y ago
You can do further training on them, just like you can patch a binary blob. There are some surgeries you can do to the weights, and there are analyses you can do to poke at them and try to understand them, but ultimately they weren't c
225.
▲
by
habitue
3y ago
One thing I don't see discussed enough is that, ok let's say the weights are unencumbered, and the source is under an OSI license: the point of open source licenses and free software was to expose the *human understandable* meanin
226.
▲
by
habitue
3y ago
> This is why we should not roll our own anything. This seems like a major overcorrection. It's equivalent to saying "you can't know anything" "no one is qualified to do anything" You really have to consider
227.
▲
by
habitue
3y ago
Personally, I use the Samsung fold 4 and it's really nice. I don't want to go back to a smaller screen and was happy to see google copying it since it makes it more likely these phones will still exist (and maybe be competitively
228.
▲
by
habitue
3y ago
loops are typed gotos, so are if statements, so, really are functions. Goto isn't bad, it's integral to programming. It's just too powerful, so we tame it in various ways. Algebraic effects are one of those ways.
229.
▲
by
habitue
3y ago
> -Everything- is negotiable at the end of the day. Right, and saying "We aren't going to negotiate" is a negotiating tactic
230.
▲
by
habitue
3y ago
Think of a terminal app like html: a cross platform UI markup language but with massive limitations. (The markup is terminal escape codes and various extensions). If you want speed, memory efficiency, and a good UI[0] you have to go with a
231.
▲
by
habitue
3y ago
This seems like a pretty even-handed discussion of the issues
232.
▲
by
habitue
3y ago
Yeah, I'm not an expert here, but I believe it might be the reporting rules. This is the law that governs it, though it delegates a lot to the SEC and the particular requirements could be in the rules that the SEC puts out itself http
233.
▲
by
habitue
3y ago
The issue is that if most coins are securities, then that means in order to operate, Coinbase has to become a registered securities exchange in order for it to do what it does. The regulations for securities exchanges are onerous, but also
234.
▲
by
habitue
3y ago
Yeah, it really seems like this article is describing the default. It's good to accept and understand that writing stuff into notes doesn't ever imply you'll read it again. That doesn't mean that if you make the effort t
235.
▲
by
habitue
3y ago
From time to time this concept of "boring development" etc comes up. And to a point I get it: we've all been burned by people switching around tech to the latest thing which ends up being buggy, poorly supported, and the bene
236.
▲
by
habitue
3y ago
It was in college, had a couple of courses that involves writing some things in assembly. There's not a lot to assembly, but when you write it you realize how function calls involve a bunch of arbitrary choices, and the stack is part o
237.
▲
by
habitue
3y ago
I kinda disagree, I learned assembly after I learned C and I felt like I still appreciated it
238.
▲
by
habitue
3y ago
Just a recommendation, but learn Rust first. Once you have a good intuition about mutability xor aliasing (which is what the borrow checker enforces), you'll actually write better C and C++ code should you need to. Personally, I think
239.
▲
by
habitue
3y ago
> It's possible to simultaneously believe Adderall is a performance-enhancing drug and be just fine with people taking it. I agree, and this was largely my point: if you think it's ok for adults to take drugs just for fun, then
240.
▲
by
habitue
3y ago
The brain is complicated. We don't understand it with the same level of detail that we understand computers. We can't refactor it, or reorganize it or fix bugs in it. All we can do is these large, scattershot system-level shifts t
More ›