badfortrains

Civil forfeiture. From wikipedia: http://en.wikipedia.org/wiki/Civil_forfeiture#United_States "In civil cases, the owner need not be judged guilty of any crime; it is possible for the Government to prevail by proving that someone other than the owner used the property to commit a crime. In contrast, criminal forfeiture is usually carried out in a sentence following a conviction and is a punitive act against the offender."

Or if you're up for a long read: http://www.newyorker.com/reporting/2013/08/12/130812fa_fact_...

More is correct according to the US energy information administration [http://www.eia.gov/todayinenergy/detail.cfm?id=9991]. Only about 10% more though according to their graph

Fun seeing what people do with their PIs. Really appreciate the time people put into documenting their projects.

I've been running a custom mp3 media server[1] built around a similar stack (node,socket.io,express) and I was surprised how capable the pi is for this kind of thing. The only real slowdown I saw in my project was inserting/reading thousands of records from a sqlite3 db (displaying all tracks for example) and even that was solved pretty easily with some caching. My project is all running headless though, with presentation handled on the client (phone/desktop browser); I'd be curious how the chromium render speeds are for this. In my testing any kind of web browsing on the pi was pretty painful.

[1]: https://github.com/badfortrains/wupnp my much more poorly documented raspberry pi project

It seems that the F-word/nipples moratorium is pretty much over. While it's true that broadcast tv still can't show it, cable/satellite tv is pretty ubiquitous and free to bath us in as many nipples and F-words as they please.

As for 3-strikes and "zero tolerance" that's a whole other can of worms. Looking at the disparity in US crime statistics vs Scandinavian countries, it's pretty understandable that cultural attitudes toward legal punishment are so different. 3-strikes and "zero tolerance" are wrong headed and ineffective, in my view, but you can at least empathize with the motivation (as opposed to the whole nipple thing)

There is no guarantee against this, which is part of what keeps web advertising prices low. Fraud is essentially priced into the product. The title is misleading, as it assumes all cost is bared by the advertisers. In truth part of the cost is implicitly payed for by legitimate website owners/the ad network who must price impressions lower to offset this kind of fraud

The thing that NNs have in their favor that other "20 year old techniques" lack is their ability to model any mathematical equation. There is no fundamental limit to the complexity of systems NNs can model (as there is with other AI techniques).

The problem with NNs is the difficulty of training them. Back propagation with random initial weights is simple, but it can easily converge on suboptimal local maximum if the learning rate is too aggressive. On the other hand, a slow learning rate requires an exponential increase in training time and data. Back propagation as a method was never really broken, it simply wasn't efficient enough to be effective in most situations. Deep belief techniques seem to remedy these inefficiencies in a significant way, while remaining a generalized solution.

Essentially deep belief networks seem to optimize NNs to the point where new problems are now approachable, and greatly improve the performance of current NN solvable problems. The complaint that "the core of the recipe itself is not very new", seems irrelevant in light of the results.