consume

why are so many people incapable of understanding std::memory_order?

Instruction reordering on the processor level is counter-intuitive and Sepples is full of morons.

std::vector<(uint64)std::int<long long int>>::algorithms::fast::sort<[](double double_x, double double_y) { return (uint64)double_x > (uint64)double_y; })(int_vector);

>>1,2
It doesn't help that most programmers who need to worry about memory reordering are already accustomed to using memory models other the one defined by C++11. It's not as if the C++ committee invented atomics or memory reordering barriers - everyone who needed them already rolled their own.

I imagine the primary use of the C++11 memory primitives is to replace assembly code in user libraries with high performance requirements. Kernel or applications programmers have little need of them.

"There are many mailing lists and on-line forums where
Joe Appcoder and Joe Sixpack and all the other Joes
are excitedly discussing the imminent arrival of
True AI and Strong AI and the Technological Singularity.
These AI wannabes and Moechte-gerns are brimming each
with their own approach to how AI should be achieved."

"So what's the problem?"

"The problem is these otherwise brilliant people
have not taken the trouble to think out and explore
how basic human intelligence came into existence on Earth.
They have not paid their AI dues. They have not set about
designing a cognitive system based on the obvious aspects
of their own central nervous system. They have the
drive and ambition to create AI, but they are actually
too eager and too impatient to work at the leading edge."

"Nothing wrong with that."

"They should start with the outside of the CNS --
the sensory organs and the motor actuators --
then work their way inwards to the center of
consciousness. For each of the main human or
robotic senses, they should design sensory
communication channels and sensory memory
systems for both storage and retrieval of
sensory data. It will dawn on them that
visual recognition is the most difficult
sense for evolution or technology to create."

"Vision in robots is already far advanced."

"Once the would-be mind-designers have mapped
out five or more sensory inputs and storage
systems, they will eventually tumble to the
main neurotheoretical question: What ties it
all together in a unified system of cognition?"

"Is it a society of mind, like Minsky suggested?"

"It is more like a gridwork of latent systems
ready to spring in to action during the stream
of consciousness. It is the side-by-side existence
of both concerete sensory channels and abstract
conceptual channels, overlaid with linguistic
mechanisms for thinking in natural human language."

"Not all thinking is in words; it can also be in images."

"But language is what distinguishes Homo sapiens from
the apes. Language forces the thinking organism to
express itself with verbal concepts. Therefore the
ineluctable key to True AI is NLP -- natural
language processing."

http://www.amazon.com/dp/B00F8F1FG0

http://bbs.progrider.org/prog/read/1445160164/13

Go to bed, Psy Borg.

and your robotic rendition of Gangnam Style sucked, too.

>>5

But language is what distinguishes Homo sapiens from

the apes. Language forces the thinking organism to
express itself with verbal concepts. Therefore the
ineluctable key to True AI is NLP -- natural
language processing.

Why are these nerds always so obsessed with replicating human intelligence. Start with something simpler, like the intelligence of some lower mammal e.g. a mouse. For all any neuroscientist knows, the human capability for language is just more of the same as far as cognitive algorithms are concerned (with respect to these lower mammals), just given a much larger set of neurons to operate on.

>>7
lol reductionists may ``know'' but they don't understand.
muh neurons

mhmm apes are animals (colloquially) and humans aren't.

humans are plants

>>9
I'm a tomato.

big black banana

>>11
In other words, you're either a myth or you smell like rotten fruit.