khafra

45 ms·

511.
by khafra 15y ago
Since donors, once added to the database, can save anyone in need who's a match, it doesn't affect the outcome whether HNers join to save Gupta or a faceless stranger. I guess what's upsetting you is that a community member makes people car
512.
by khafra 15y ago
I'm trying to think of a dollar amount high enough to let somebody stick a tube into my jugular and extend my circulatory system outside my body for hours. I don't think I can count that high; unless I needed the money for fairly immediate
513.
by khafra 15y ago
The point isn't that they should pander to experts. The point is that they're wrong more often than they're right; seriously, factually, inarguably incorrect.
514.
by khafra 15y ago
I feel the need to point out that it's still deterministic and fully observable; it's just that there's around 50 orders of magnitude more complexity to Go than to Chess; and to humans and the AIs we've created so far that's a qualitative r
515.
by khafra 15y ago
> At least they’re not using a non-linear Bayesian splitting tree didactogram This seems like a weakness to me. I'd love to be able to choose between different algorithm for matching features and set some relevant coefficients. It'd b
516.
by khafra 15y ago
It's been done: http://stupidfilter.org/wiki/
517.
by khafra 15y ago
It's always motivated by expected profit. "Ideological" just means it's not about next quarter's profit, but about creating an environment more conducive to profit in the long term.
518.
by khafra 15y ago
The obvious solution is to terraform Mars so it has a substantial atmosphere.
519.
by khafra 15y ago
They've been rooted.
520.
by khafra 15y ago
I was under the impression that discharge rate scaled fairly reliably with recharge rate. Is that not the case?
521.
by khafra 15y ago
Liquid cooling for batteries wouldn't be worth it in earbuds, but it probably would in an electric car.
522.
by khafra 15y ago
When I save the state of RAM to nonvolatile storage, then later put it back into RAM and continue executing, there are senses in which you could legitimately say it's not the same program. But they're not the senses I care about.
523.
by khafra 15y ago
Tell him that, as a college student with lots of social connections, if he neuters the internet like this you're going to have a whole lot of spare time--and you'll devote all of it to organizing protests against his re-election and/or any
524.
by khafra 15y ago
The husband of the other person behind the study is just upthread: http://news.ycombinator.com/item?id=3218604 He confirms that it's a grossly misrepresentative pop-psychology spin on a real study.
525.
by khafra 15y ago
Your dad's Solar panels weren't PV--they were black panels with water circulating through them. This color would help with that.
526.
by khafra 15y ago
Sounds like a machine learning problem.
527.
by khafra 15y ago
Not publishing positive replications is just as bad a problem as not publishing negative original results. We have things like BigTable and Hadoop now; if 100 laboratories repeat an experiment and publish their results, that just means we c
528.
by khafra 15y ago
> You're missing the point of scientific publication. The entire point of publication is so others can reproduce your work. Speed doesn't work if you need a billion data points taken over 20 years to prove a long term issue. In theory,
529.
by khafra 15y ago
It's not at all controversial--standard social signaling theory. GMU Econ professor Robin Hanson's blog, http://overcomingbias.com , has stuff about it all the time. Here's some more classic citations: http://octavia.zoology.washington.e
530.
by khafra 15y ago
IIRC that got blown by one of the journalists they had been working with.
531.
by khafra 15y ago
For me personally, I don't mind a lot of autoconfig and getting signed up for stuff by default. It demonstrates the value of a new service in a tangible way, If I don't like it, I can un-sign-up, and if that doesn't work I can filter spam.
532.
by khafra 15y ago
Everyone seems to have missed my point, which was that Google has never created a good iOS app. Google Maps is the closest thing to a good iOS app that involves Google. So we have no reason to give them any slack on this; the way we forga
533.
by khafra 15y ago
But where's Google's iOS record of success that should make us overlook this error? The Google Maps app is pretty good...
534.
by khafra 15y ago
I hope your boss said "no."
535.
by khafra 15y ago
The AI Class official twitter account asked people whether they'd be willing to pay $2000 for a Stanford-level CS Master's degree taught in a similar format.
536.
by khafra 15y ago
I agree with the implication that talking about bettering yourself isn't usually the best way to better yourself.
537.
by khafra 15y ago
I'll bet he want to complificate the formula to look something more like "P(H|Q)=P(Q|H)P(H)/P(Q)" where H = "User upvoted a honeypot article" and Q = "User's votes are not a good signal for article quality" And a similar adjustment for flag
538.
by khafra 15y ago
Sounds like a machine learning problem. Showing people the cache of really bad articles would lead to overfitting. Making each user's vote a vector with the value of the honeypot formula determining its strength, and the vote total for any
539.
by khafra 15y ago
How does the total value of privately traded companies compare to publically traded companies, and how many privately traded companies can the Rothschilds plausibly own without it being public knowledge?
540.
by khafra 15y ago
It would be fantastic if the police learned that abusing their power gets them recorded on video, and trying to suppress the video gets the video even more publicity. But it'd be nice if other companies weren't making Google look so unusua

More ›