Armed Israeli settlers shooting at Palestinians.

TheMotto · 4 days

Ignoring the plight of the downtrodden is exactly what america/uk did while millions of jews were tortured to death in germany - do you agree it was a good choice then as well? Would you have stopped reading the paper.

At this point the israel/palestine problem comes down to the idiocy of american jews (who despite being the most academically succesful group in america cannot realize that even one death of a human life is not worth the little patch of land that is Isreal/palestine). War/violence is almost always stupid.

It's one thing for 3rd worlders or dumbass religious zealots to advocate conflict, but it's a damn shame when an american jew advocates it.

marshallp · 5 days

Exactly - I'm a offering a good deal.

(mainly because I'm working on a "startup" and don't want to go into an office every day).

EquanimousMind · 5 days

what about if you're a single mom with kids. Or a guy that doesn't have the option of getting a "sitting" job like call center, but the options are ditch digging or toilet cleaning.

It's the poor, like you, who have screw themselves over with "i'm a working man" talk. The upper middle classes/wealthy get all sorts of government regulations to help them out (doctors/lawyer regulations, bailouts, public-private partnerships).

marshallp · 18 days

Toronto, Canada

marshallp · 20 days

pm me

runninground · 1 month

Some more links on evil robots http://curatedvideos.blogspot.ca/2012/04/movies-and-shows-with-evil-robots.html

burnabee13 · 1 month

So a politician from one of the most corrupt countries on earth is a better choice than an innovator?

The world bank and imf need innovation, silicon valley style. What type of person has Bill Gates appointed in his "development bank" - a bunch of politically astute economists or innovators - that's the choice the economist should be making.

It seems like more and more the economist magazine is out of touch with the era we live in.

pwnies · 1 month

bronzbroh · 1 month

They really messed up with the cash prize. If they made it $100 million, or even $1 billion (which is nothing when you consider darpa's budget of 3 billion a year), you would see real worldwide interest. You might even see hedge fund/private equity investors jump in with outsourced development to india/china.

This project should actually even be 10-100 billion dollar prize. The amount of wealth created would be in the trillions (elimination of basically all manual labor). Imagine how rich everyone would be if everyone got access to a hundred servants, who are not paid a salary (manufacture a trillion of these humanoids). Poverty is eliminated and massive speedups in biomedical experimentation can occur (cure all diseases).

bronzbroh · 1 month

You could create specialized robots for each task but why bother? Everything is already setup for "humanoid" operation. With a working humanoid platform engineers can focus on software (which is cheaper to develop) than creating a hardware platform everytime.

Plus, as with everything, porn can be huge driver of progress - and you can only do that with humanoids.

strayadvice · 1 month

I see what you're saying.

(I suppose anyone who successfully used it on the stock market wouldn't be publishing literature on that!)

Would autoencoding still be preferable if I have a million features and millions of examples, all labelled. (I've been looking at generating graphics and training it with vw for object detection, starting with simpler shapes in cluttered environments to more complex objects - a hierarchical system.)

Is there an autoencoding software that works as fast as vw.

strayadvice · 1 month

Is there any literature on that.

I would think running something like vowpal wabbit (which can handle as much data as you can get) would be better than selecting data and running expensive autoencoding to select an even smaller subset.

strayadvice · 1 month

autoencoders are probably the wrong tool for stock market prediction.

Autoencoding is a way of getting to use lots of unlabeled data (which is cheaper to obtain).

With the stock market pretty much all your data is already labeled.

Also, when it comes to stock market prediction I would caution against using more complicated algorithms, you'll end up finding lots of strategies that work in the "past" but don't work in the future (you'll overtrain a lot) and this will be costly if you're using real money.

gfintegrity · 1 month

Kim would be in a league by himself. The vast majority of the benefit of big projects goes into the pocket of mega construction corps and the dictators. The people least trustable at the world bank are people at the top like ngozi who get there by doing whatever the mega corps and dictators tell them. Why this isn't obvious to everyone is beyond me.

lmc2179 · 1 month

Cloud is too expensive for small players. Use fast algorithms like vowpal wabbit.

DrRichardCranium · 1 month

Read Brad Templeton's writings about robocars. With self-driving cars most people won't be buying cars anymore, they'll be taxis.

Anyway, there's a lot bullshit in the media, maybe there is a limit or not.

moscheles · 1 month

The robot is willow garage's, they are selling them for 240k to academics and over 400k otherwise. They also develop the ROS which is fast becoming the standard robot platform.

DrRichardCranium · 1 month

The point is, Lithium batteries are used because wet batteries are too heavy, that's the only reason lithium batteries are used. If only need a small amount of energy (a short few mile trip) than you can just use some wet batteries - therefore no shortages of minerals.

marshallp

TROPHY CASE

you'll need to login or register to do that

create a new account

login