Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rndn
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
46 ms
·
421.
▲
Predictions made by Ray Kurzweil (1999)
(en.wikipedia.org)
9 points
by
rndn
12y ago
|
0 comments
422.
▲
by
rndn
12y ago
Isn’t that an infinite regress problem?
423.
▲
by
rndn
12y ago
It’s easy to prove "we don’t know that" since maybe it is possible to introspect our universe. For example, suppose someone interacts with some quantum system and it unexpectedly outputs a binary string. It turns out that this str
424.
▲
by
rndn
12y ago
That’s the idea of an infinite number of onion layers [1]. What I find intriguing is that if you assume that everything is based on cause and effect, then (I think) you necessarily run into an infinite regress problem (the cause of a cause
425.
▲
by
rndn
12y ago
https://news.ycombinator.com/item?id=8811252
426.
▲
by
rndn
12y ago
In other words it's a material which has a very low reflectance across a large range of frequencies (i.e. a black material)?
427.
▲
by
rndn
12y ago
Awesome idea and execution. I’m wondering whether it would work better with an edge detection algorithm instead of using the triangle edges.
428.
▲
by
rndn
12y ago
While we’re on it, here is a really good neck exercise: https://www.youtube.com/watch?v=LT_dFRnmdGs
429.
▲
by
rndn
12y ago
Three more resources: http://web.eecs.umich.edu/~cscott/rsi.html http://www.rsipain.com/ http://www.amazon.com/Repetitive-Strain-Injury-Computer-User...
430.
▲
by
rndn
12y ago
It seems to me the more different input devices, the better, because one can quickly switch the input device when it becomes too repetitive. I’m currently experimenting with a vertical mouse (Anchor), a handheld trackball (IMAGE), a handhel
431.
▲
by
rndn
12y ago
The wonderful and terrifying implications of computers that can learn https://www.ted.com/talks/jeremy_howard_the_wonderful_and_te...
432.
▲
by
rndn
12y ago
Why do you think the revolution will come from startups rather than from tech giants such as Google and Amazon?
433.
▲
tiny.el
(oremacs.com)
4 points
by
rndn
12y ago
|
0 comments
434.
▲
by
rndn
12y ago
Previous submission: https://news.ycombinator.com/item?id=8643774 Response: https://news.ycombinator.com/item?id=8689912
435.
▲
The FCC ignored hundreds of thousands of net neutrality comments
(reddit.com)
5 points
by
rndn
12y ago
|
0 comments
436.
▲
by
rndn
12y ago
> There is no "best" way to develop technology, and there is certainly no predetermined, given course that technological development must inherently follow. On the other hand, many inventions and discoveries have been made simu
437.
▲
by
rndn
12y ago
Of course, but naming only Apple is misleading.
438.
▲
by
rndn
12y ago
* as any other company that outsources manufacturing.