Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ftxbro
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
25 ms
·
241.
▲
by
ftxbro
3y ago
If you can't leave USA, you got to just get your bag and understand that you need a lot, more than what people think is normal. You can pay for private medical care it's just expensive. Like it's too expensive for people who
242.
▲
by
ftxbro
3y ago
Here's a 2016 Wired article about Doug Lenat, he was the guy who made Eurisko and CYC https://www.wired.com/2016/03/doug-lenat-artificial-intellig...
243.
▲
by
ftxbro
3y ago
> But he didn't say that. he didn't say what?
244.
▲
by
ftxbro
3y ago
He says the definition of consciousness isn't a mystery but he feels the need to make his own definition. Doesn't that mean its definition was kind of mysterious? Also this is just an ad for some kind of woo woo tiara that they ar
245.
▲
by
ftxbro
3y ago
This is an internet comment section, if you are talking about things being exponentially hard and part of your comment is asking "why this is not merely O(28^2) harder (or do you need the electrons too?)" then of course people are
246.
▲
by
ftxbro
3y ago
> For instance: https://ieeexplore.ieee.org/document/5763440 > Our GPU implementation shows a speedup up to 97.3 times over Matlab Implementation and 2.89 times over the Intel Math Kernel Library implementation o
247.
▲
by
ftxbro
3y ago
> I don't think there is such a thing as a silly question but consider this: it's not just 'running the code', you need to actually understand the problem and short of a career in physics you're looking at '
248.
▲
by
ftxbro
3y ago
Yeah you would think, which is why I was wondering "is there a code that i can run to check the stability" but apparently that's a silly thing to ask for some reason.
249.
▲
by
ftxbro
3y ago
it sound like they faked it without making it
250.
▲
by
ftxbro
3y ago
that's cool but you can also just use gpt directly and also not be so worried that the website wants to front run your domain name or cause other mischief
251.
▲
by
ftxbro
3y ago
I meant more like the code they are using these days for that same problem of ab initio oxygen isotopes or whatever it would be called. I know they didn't use GPU for things until recently, like maybe even now they haven't convert
252.
▲
by
ftxbro
3y ago
It's because it reduces the candidate pool too much, and even jobs where you will definitely use monads like some devs at Jane Street they will accept candidates who don't even know functional programming yet.
253.
▲
by
ftxbro
3y ago
Is there a code that i can run to check the stability, if i have a lot of time to spare and a fast computer with a lot of memory? Do the codes use GPU for their Lanczos things yet?
254.
▲
by
ftxbro
3y ago
They found them by blasting protons at each other and seeing what happened when they collided like this https://en.wikipedia.org/wiki/Quark#/media/File:Charmed-dia-... and then using detective skills to figur
255.
▲
by
ftxbro
3y ago
The simple shell model said 'stable' but we can calculate with more detailed models than that one.
256.
▲
by
ftxbro
3y ago
This isn't any more interesting than saying the same thing about smuggling secret messages in controlled 'dice rolls' or any other generation where you know the distribution. It doesn't have any important connection to A
257.
▲
by
ftxbro
3y ago
it's paywalled
258.
▲
by
ftxbro
3y ago
Can we not simulate even one atom of oxygen well enough to determine if it's stable or not. What if they used a big computer like the ones they keep building at national labs or the ones they used for training GPT.
259.
▲
by
ftxbro
3y ago
I heard that you can grind leetcode to compare yourself to other programmers. My understanding is that it's for self contained puzzles like in interviews or contests, so it's obviously not representative of everything that makes a
260.
▲
by
ftxbro
3y ago
maybe they will go straight into quantum computing and not bother with archaic liquid tin laser printers
261.
▲
by
ftxbro
3y ago
i feel like any people-first view will run into capitalism which has a capital-first view
262.
▲
by
ftxbro
3y ago
a lot of 'productionised' things keep getting enshittified and ruined so i don't blame anyone for remaking a wheel
263.
▲
by
ftxbro
3y ago
wait isn't 7nm pretty good, the hacker news story about a new thing by intel used 7nm too https://news.ycombinator.com/item?id=37315802 probably you could do some stuff with enough 7nm chips > a lot of tech breakthr
264.
▲
by
ftxbro
3y ago
package.json says MIT license but license.txt says GPLv3
265.
▲
by
ftxbro
3y ago
> Every single student in the morning section of my class scored higher on the final exam than [GPT], which only managed a C-minus with a score of 72.5. My class average was in the mid-80s. In my afternoon section, and facing a different
266.
▲
by
ftxbro
3y ago
imagine opening a bookstore in the year of our lord 2023
267.
▲
by
ftxbro
3y ago
yes but on the other hand it reduces churn
268.
▲
by
ftxbro
3y ago
> He's written "over 800 scientific papers". Is that... a good thing in his field? as some others have said, that's not the flex he thinks it is
269.
▲
by
ftxbro
3y ago
> this unfamiliar abundance pattern may reflect an extraterrestrial technological origin These spherules are from extraterrestrial alien technology!
270.
▲
by
ftxbro
3y ago
you're right i'm going to go into the pacific ocean and look at those spherules myself and then go visit another star and see for myself that it's like the ones from outside our solar system
More ›