HNHacker News
TopNewBestAskShowJobs

dark_silicon

3 karma · joined September 15, 2017

submissionscomments
dark_silicon··on Training crows to pick up cigarette butts for food
Just to play devil's advocate here: you do have the alternative of not smoking.

(I dislike smoke, but if you're not blowing it around in a public place I don't particularly mind your decision to smoke)

dark_silicon··on Training crows to pick up cigarette butts for food
The failings of the broken windows policies hinge on their faulty premise: that the sight of minor crimes going unenforced leads to people brashly escalating to higher crimes.

I don't believe this premise, yet I wonder sometimes if keeping the peace in regards to smaller offenses would lead to fewer incidences of anger turning to road rage or violent crime. If people do not feel safe or that they have civil outlets for resolution of grievances, perhaps they're more violent? I'm curious what the more social science oriented on this site might know about this.

In any case, there's still yet another reason to enforce laws against small crimes: littering is a terrible thing in itself!

dark_silicon··on Three Paths in the Tech Industry: Founder, Executive, or Employee
"fancy apartment". Cute idea.
dark_silicon··on Consciousness Goes Deeper Than We Think
I'd encourage anyone to examine their conscience, read about the viability of babies born prematurely, and consider that even post-birth a baby is not self-sufficient. We don't have good criteria for what makes something alive or conscious, yet many draw an arbitrary line at birth.

I'm not looking to continue this discussion, but I'm glad you're thinking about this.

dark_silicon··on PassGAN: A Deep Learning Approach for Password Guessing
Is there any sort of metaheuristic a good password cracker can follow for juggling between rounds of dictionary iteration, character substitution, brute force suffixing, etc?

In other words, how do you most effectively combine different strategies?

dark_silicon··on PassGAN: A Deep Learning Approach for Password Guessing
This makes me wonder: is it possible to iterate through the generated outputs of a GAN rather than generate random outputs?