Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Schiphol
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
49 ms
·
571.
▲
by
Schiphol
11y ago
All right, let's go with that :)
572.
▲
by
Schiphol
11y ago
The title is taken verbatim from the link. I agree it's a bit misleading, but I'm finding it hard to come up with a better one. I'm open to suggestions :)
573.
▲
Infinitely fast phase velocity with zero-index metamaterials
(qz.com)
5 points
by
Schiphol
11y ago
|
4 comments
574.
▲
by
Schiphol
11y ago
Thanks for this. Incidentally, the second paper you link to is co-auhored, among others, by Daphne Koller, who teaches [this great course on probabilistic graphical models]( https://www.coursera.org/course/pgm ) and Andr
575.
▲
by
Schiphol
11y ago
5% sounds pretty high to me, considering how low (nonexistent, really) the entry barrier is. Where does that figure come from, if I may ask?
576.
▲
List of the Most Popular MOOCs
(onlinecoursereport.com)
200 points
by
Schiphol
11y ago
|
85 comments
577.
▲
Computational Modelling Courses
(aidanhorner.blogspot.com)
4 points
by
Schiphol
11y ago
|
0 comments
578.
▲
by
Schiphol
11y ago
Link to the original paper: http://arxiv.org/abs/1510.03346 . It hadn't occurred to me that self-driving cars would put manufacturers in the position of actually having to make a decision on the various trolley-pro
579.
▲
The Ultimate Ethical Dilemma: Why Self-Driving Cars Must Be Programmed to Kill
(technologyreview.com)
2 points
by
Schiphol
11y ago
|
1 comments
580.
▲
by
Schiphol
11y ago
The "community of websites" idea is perhaps misleading that way. Websites in this context are, pretty much, analogous to your Twitter or Facebook page.
581.
▲
Does the physics of Marty’s time travel add up?
(newscientist.com)
2 points
by
Schiphol
11y ago
|
1 comments
582.
▲
by
Schiphol
11y ago
In general, it's far from useless to investigate the nature of X before having a clear definition of X. Perhaps the definition will come as a result of the investigation. Perhaps there will never be a clear definition, but paradigm cas
583.
▲
Google’s been quietly recording your voice
(qz.com)
3 points
by
Schiphol
11y ago
|
0 comments
584.
▲
by
Schiphol
11y ago
Then, finally (for now), you have Erdos-Bacon-Sabbath numbers ( http://timeblimp.com/?page_id=195 )
585.
▲
by
Schiphol
11y ago
I found this sentence funny "Representatives of Red Hat and Ansible did not immediately respond to requests for comment". I take it to mean: "we wanted to run the story as quickly as possible; still it would have been nice to
586.
▲
by
Schiphol
11y ago
And a video tour here http://www.sixdegreesoffrancisbacon.com/ Prediction: the next big thing in fan wikis will be a similar 6º of, say, Frodo Baggins.
587.
▲
Six Degrees of Francis Bacon
(cmu.edu)
11 points
by
Schiphol
11y ago
|
9 comments
588.
▲
by
Schiphol
11y ago
And a video here: http://www.sixdegreesoffrancisbacon.com/
589.
▲
Six Degrees of Francis Bacon
(cmu.edu)
1 points
by
Schiphol
11y ago
|
1 comments
590.
▲
by
Schiphol
11y ago
William Rapaport has just won the 2015 Barwise Prize for significant work on philosophy and computing -- http://www.apaonline.org/news/254862/William-Rapaport-is-the...
591.
▲
Philosophy of Computer Science [pdf]
(cse.buffalo.edu)
75 points
by
Schiphol
11y ago
|
10 comments
592.
▲
by
Schiphol
11y ago
The vagaries of HN submissions :)
593.
▲
Kilogram conflict resolved at last
(nature.com)
10 points
by
Schiphol
11y ago
|
3 comments
594.
▲
Strogatz's 'Nonlinear Dynamics and Chaos' Video Lectures (2014)
(youtube.com)
2 points
by
Schiphol
11y ago
|
0 comments
595.
▲
Top 5 ArXiv Deep Learning Papers, Explained
(kdnuggets.com)
1 points
by
Schiphol
11y ago
|
0 comments
596.
▲
Giraffe: Using Deep Reinforcement Learning to Play Chess
(arxiv.org)
17 points
by
Schiphol
11y ago
|
2 comments
597.
▲
Ask HN: Publishing a copyleft book
7 points
by
Schiphol
11y ago
|
2 comments
598.
▲
by
Schiphol
11y ago
Having to come up with scientific lessons on the spot ("this may cause us to rethink..."), for the sake of press releases must be pretty stressful. I'm also not sure that such more-or-less impromptu analyses foster the right
599.
▲
Tips for Reading Math Books
(blog.jpolak.org)
3 points
by
Schiphol
11y ago
|
0 comments
600.
▲
by
Schiphol
11y ago
All right, yes, I do that. I thought the OP was alluding to something more sophisticated :)
More ›