albntomat0

40 ms·

391.
by albntomat0 8y ago
I was unclear in my previous comment. Humans are fine with recognizing gibberish answers, but I'm skeptical of the ability to differentiate great writing from simple decent.
392.
by albntomat0 8y ago
Adversarial ML where the attack has to fit in the mind of the test taker and iteration is slow is an interesting problem. I can't say that I trust the original humans grading thousands of SAT essays prior to any ML assistance though.
393.
by albntomat0 8y ago
This seems easy to game, especially if you have any insight into the model behind the grading. I doubt that those running it will release any specifics, but with a large industry[0] behind test prep, general guidance on how to beat the syst
394.
by albntomat0 8y ago
How did it go though? I've seen a large number of politically charged statements about it, but little actual fact (I've done some research, and would appreciate a competent citation)
395.
by albntomat0 8y ago
Alternatively, this allows better defense, rather than everyone and their grandmother having to run their own setup. It is a single point, but also allows defenders to focus their efforts.
396.
by albntomat0 8y ago
Different, but related issue (at least philosophically, to me): US colleges currently are 57:43 women to men [0]. Does this warrant a response, and if so, how should it be implemented? Male quotas? Gender as a "plus" in admission
397.
by albntomat0 8y ago
> Why is such an evaluation is even necessary? What is its purpose? In general, selecting for good personality traits make sense. In the extreme, only selecting for test scores and grades ignores all of the characteristics that make a pe
398.
by albntomat0 8y ago
My understanding of it is that they are multiple samples of the same underlying distribution (the applicant's subjective characteristics). Multiple references who know/have interacted with the applicant attest to his/her pers
399.
by albntomat0 8y ago
That's not what the article said. The key point is that Asian applications scored well in "whole person" metrics when evaluated face to face, from teachers, high school counselors, and alumni interviews, but much worse when g
400.
by albntomat0 8y ago
At what point should the military refuse the orders of civilian leadership? A tactical level order to shoot civilians (or similar, non-cliche situation) should be disobeyed. However, is it the military's responsible to disobey the deci
401.
by albntomat0 8y ago
Ahh ok, I misread your message. My mistake. I've spent a good amount of time looking into the issue (for a number of reasons), and it seems to me that nuclear policy, on the level of an individual US citizen, is highly complex, especia
402.
by albntomat0 8y ago
Debatable. There's a strong argument that nuclear weapons have prevented direct conflicts between the major powers (mutual assured destruction). Of course, there are a number of situations in which nuclear weapons could end up being us
403.
by albntomat0 9y ago
If they did, it would be in their best interests to prove it, via some statement from a reputable bank or such, that users would trust holding $500+ million.
404.
by albntomat0 9y ago
Out of a country with a population of 320+ million, with 750,000+ "sworn officers" (able to make arrests) [0], 1188 is not "absurdly common", especially as that count is an overestimate of the objectionable cases (by how
405.
by albntomat0 9y ago
Here are some statistics, to help put things in context. There have been 976 people shot and killed by police this year. "Three out of five of the people shot and killed by police were armed with a gun, while fewer than 1 in 10 were un
406.
CMU Led High School Computer Security Competition Returns(2017.picoctf.com)
2 points by albntomat0 10y ago|0 comments
407.
Drone Delivery of Medical Supplies(nytimes.com)
1 points by albntomat0 11y ago|0 comments
408.
by albntomat0 11y ago
Is there anything currently in Go like in chess where computer assisted humans play best (centaurs)? I did a brief internet search, but nothing specific came up.