I get my news from Slashdot. My wife tells me I have a very republican-leaning family -- if I was on facebook, I'd probably be exposed to a larger breadth of opinion. Confirmation bias? I try to look elsewhere for news... but ultimately, I keep coming back to the places that present the stories that I care about.
I'd bet quite a lot that you're wrong on that count. The source was a high-ranking official. Mattzie is not a source of information about intelligence matters. He was very briefly a reporter. And he outed Hayden as badmouting Obama behind a cover of anomymity. So, Hayden might be the source of this. But definitely not Mattzie.
I, for one, would rather ride a horse at 70mph+. If you can't provide that, I'll be keeping my manual pickup with manual locks, manual windows, manual brakes, manual airbags, and manual manual, thankyouverymuch.
I mean, who the fuck are they trying to fool with "Right off the bat you can imagine autonomous driving easily topping your average intoxicated drivers' ability behind the wheel."? Really? Oh boy, it's safer than a drunk driver! Fuck yeah, sign me up!! Real vote of confidence there.
Seriously: no, I cannot* imagine that the autonomous cars are actually better than the average intoxicated driver, according to modern limits. Your average intoxicated driver gets home safe and doesn't put anybody at risk. Given what I know about the difficulty in training a computer to behave intelligently in unplanned situations, I cannot imagine that autonomous cars are significantly better than your average intoxicated driver.
* Actually, I can imagine a lot of things. I can imagine winning a pulitzer prize for this slashdot post. I've got an amazing imagination, possibly surpassing the imagination of an intoxicated driver.
I didn't say 'trust nothing'. Trust in a large and diverse community is preferable to trusting a single entity, be it a person, government agency, or corporation. If your entire hardware and software stack is open and verified by a large number of people representing a diversity of motives, then and only does it derserve any trust. "Trust no one" does not rule out trust in humanity. Crowds can either be much smarter, or much stupider, than their constituents -- use them wisely, and trust no one.
But from whom do you learn the math? A teacher? A textbook? Unless you derive it all yourself from base axioms, you do have to trust someone at some point.
A proper math education starts from basic axioms. A teacher should merely guide a math student through derivation of that mathematical knowledge which is taught, all from base axioms. In my undergrad, math majors were required to take a course on the axiomatic foundation of math (set theory). The classes that depended on that built the foundations of algebra, analysis, etc. upon those same axioms. The subsequent built upon those results, etc.
One problem with crypto is that we've never seen a hardness result of any of this shit. Until somebody proves P!=NP and builds an NP-complete cryptosystem, I won't trust any of it. The math is inherently untrustable without a proof of trustworthyness. The other problem is in implementation. All the math in the world won't save your data from a shit implementation that leaves you open to side-channel attack, etc.
Society does not break down with a lack of trust, society proceeds, as always, with constant implicit and explicit cost-benefit analysis.
You have to trust someone, somewhere along the line.
No. No you don't. You should always remain skeptical, consider motivation as well as message, and read between the lines. Think for yourself and check facts.
It is better to trust the math community (a large body of people of varying motives, hence less corruptible by a single source) than Schneier. It is even better to learn the math yourself. It's even better to kick down the doors of the NSA and demand answers. But that last one will probably get you shot or worse.
Yeah... I watched a talk by a D-Wave guy. This is a summary of his talk: "So, you have an NP-Complete problem. We have a quantum solver that works on a large graph with a special structure. If you can find a homomorphism from your problem into our graph structure, and you can figure out how slowly to evolve the adiabatic process, then we can solve your problem!"
Okay, that's great. But finding that graph homomorphism? Probably NP-Complete itself. Figuring out how slowly to evolve the system? I have no idea, but the guy said "it's hard", which means the physicists don't have any idea. Maybe also NP-Complete or worse? Who knows. Tell ya one thing, D-Wave sure doesn't.
See, Wikipedia started like that, too. But very soon, our Internet Congress would be populated by corporate fucktards, and not long afterwards, anybody with a clue would be outnumbered and banned from the internet. Also, any mechanism that you create to ban folks from the internet will be used for governments for censorship. Great intention you've got, but that road leads straight to hell.
I have little idea what works for supercomputers and highly parallized data analysis (I've never used one).
Oh ok. So when you said Sage is okay for "small-midsize projects" and recommended Mathematica for large projects? You've only used Mathematica for small projects, and have no idea what a large-scale project is. And your pricing info is woefully outdated if you're recommending this to a new user.
Documentation is a skill. If you think it's not hard, you're either ignorant to your own failings, or those downstream of you are very lucky. Lots of people try hard and spend a long time on documentation and still fail to make it good. It *is* hard for most people to think of the many ways documentation needs to be used, the many mindsets of its users, and balance those two into concise, readable documentation.
I've used Sage on a supercomputer, chugging through hundreds of gigs of data. Do you know what you're talking about, or are you just recommending the shiny thing that you paid lots of money for?
I use Sage. When Python isn't fast enough, I can essentially write in C with Cython. It's gloriously easy. Have some trivially parallelizable data mining? Just use the @parallel decorator. Sage comes with a slew of fast mathematical packages, so your toolbox is massive, and you can hook it all in to your Cython code with minimal overhead.
Said novice computer user would probably be like 'who is this Eve person and why does this strange website want my credit card info?!?' and trash the email. The more likely targets are the semiliterate, who are familiar with the service. They'll fall for the "pay $3 and I'll send you $100!" bullshit, get greedy, and give their credit card to a non-genuine site without noticing. Or... straightup fall for the "pay $3 and I'll send you $100" bullshit and give their credit card to the completely genuine and complicit Square, and never report it 'cause they're embarrassed about getting duped.
RTFS. You only send the amount. Who you're paying and how much is insecure, but when you get an email from Square, they send you a link where you put in your bank info.
Of course, this will spawn a cottage industry of phishers, but you shouldn't worry about phishing -- you're already paranoid. Worry about your grandma getting phished.
Nope. This is nothing like TSP. The problem for this is "given a single new user, find the bus whose current route plan would be disturbed least by adding that user". Optimal point-to-point path finding in (nearly-)plane graphs is really quite easy, and these path deformations shouldn't be hard either. Besides... you only need reasonable approximations, not a global optimum.
You don't understand atheism. Atheists believe that there is no god. That takes faith, acknowledged or not. Belief zero? That's skepticism. But there is even extreme skepticism. Extreme skeptics don't believe that the world exists.
The problem is in the language. People hear "don't believe" to mean "believe the opposite". The very notion of undecideability just doesn't occur to people. I try to not believe things that I haven't seen proof of. Sometimes I fail in that.
Atheists who claim to be rational and faithless are either idiots or ignorant. If they don't realize they're agnostic they're simply ignorant. But hey, good news: ignorance can be fixed.
GP was quoting TFA, which was quoting a historian who knew more than what you gleaned from Wikipedia in 30 seconds. Check your own ignorance, friend. Nobody's questioning the existence of those children, merely the true identity of the father.
Forget about making busywork for people. Make robots. Let humans make art, learn, explore, teach, heal... those things humans are good at. Let robots do repetitive tasks. Focus on making people happy and healthy. Leave boredom to the robots. The creation of robots is undergoing a serious democratization, and these utopian ideals have a chance -- if only billionaires profit from the robot uprising, the world will be a terrible place. If we all profit, we can do what we love.
Indeed. I've oft considered what would happen if a private police force were to arrest public police officers for "accidental" shootings... the shitstorm would be incredible.
Before you're going to build a computer that works like a human brain you're going to have to figure out how the human brain actually works.
Oh bullshit. I mean, you might be right and everything, but the certainty with which you make that statement is wholly unsupported. As long as we aim for a design where a human understands every little piece and every line of code, then you're certainly correct. If we take a hands-off approach and focus instead on analog components and neural networks, we might have surprises in store.
Slashdot Mirror
What part of "robot overlords" do you not understand?
I get my news from Slashdot. My wife tells me I have a very republican-leaning family -- if I was on facebook, I'd probably be exposed to a larger breadth of opinion. Confirmation bias? I try to look elsewhere for news... but ultimately, I keep coming back to the places that present the stories that I care about.
I'd bet quite a lot that you're wrong on that count. The source was a high-ranking official. Mattzie is not a source of information about intelligence matters. He was very briefly a reporter. And he outed Hayden as badmouting Obama behind a cover of anomymity. So, Hayden might be the source of this. But definitely not Mattzie.
I, for one, would rather ride a horse at 70mph+. If you can't provide that, I'll be keeping my manual pickup with manual locks, manual windows, manual brakes, manual airbags, and manual manual, thankyouverymuch.
I mean, who the fuck are they trying to fool with "Right off the bat you can imagine autonomous driving easily topping your average intoxicated drivers' ability behind the wheel."? Really? Oh boy, it's safer than a drunk driver! Fuck yeah, sign me up!! Real vote of confidence there.
Seriously: no, I cannot* imagine that the autonomous cars are actually better than the average intoxicated driver, according to modern limits. Your average intoxicated driver gets home safe and doesn't put anybody at risk. Given what I know about the difficulty in training a computer to behave intelligently in unplanned situations, I cannot imagine that autonomous cars are significantly better than your average intoxicated driver.
* Actually, I can imagine a lot of things. I can imagine winning a pulitzer prize for this slashdot post. I've got an amazing imagination, possibly surpassing the imagination of an intoxicated driver.
I didn't say 'trust nothing'. Trust in a large and diverse community is preferable to trusting a single entity, be it a person, government agency, or corporation. If your entire hardware and software stack is open and verified by a large number of people representing a diversity of motives, then and only does it derserve any trust. "Trust no one" does not rule out trust in humanity. Crowds can either be much smarter, or much stupider, than their constituents -- use them wisely, and trust no one.
But from whom do you learn the math? A teacher? A textbook? Unless you derive it all yourself from base axioms, you do have to trust someone at some point.
A proper math education starts from basic axioms. A teacher should merely guide a math student through derivation of that mathematical knowledge which is taught, all from base axioms. In my undergrad, math majors were required to take a course on the axiomatic foundation of math (set theory). The classes that depended on that built the foundations of algebra, analysis, etc. upon those same axioms. The subsequent built upon those results, etc.
One problem with crypto is that we've never seen a hardness result of any of this shit. Until somebody proves P!=NP and builds an NP-complete cryptosystem, I won't trust any of it. The math is inherently untrustable without a proof of trustworthyness. The other problem is in implementation. All the math in the world won't save your data from a shit implementation that leaves you open to side-channel attack, etc.
Society does not break down with a lack of trust, society proceeds, as always, with constant implicit and explicit cost-benefit analysis.
You have to trust someone, somewhere along the line.
No. No you don't. You should always remain skeptical, consider motivation as well as message, and read between the lines. Think for yourself and check facts.
It is better to trust the math community (a large body of people of varying motives, hence less corruptible by a single source) than Schneier. It is even better to learn the math yourself. It's even better to kick down the doors of the NSA and demand answers. But that last one will probably get you shot or worse.
Sorry madam or sir, you cannot like anything today. As a perpetual grump, this suits me fine. Can't say I like it, though.
Yeah... I watched a talk by a D-Wave guy. This is a summary of his talk: "So, you have an NP-Complete problem. We have a quantum solver that works on a large graph with a special structure. If you can find a homomorphism from your problem into our graph structure, and you can figure out how slowly to evolve the adiabatic process, then we can solve your problem!"
Okay, that's great. But finding that graph homomorphism? Probably NP-Complete itself. Figuring out how slowly to evolve the system? I have no idea, but the guy said "it's hard", which means the physicists don't have any idea. Maybe also NP-Complete or worse? Who knows. Tell ya one thing, D-Wave sure doesn't.
Context is important, friend. RTFP and STFU yourself.
See, Wikipedia started like that, too. But very soon, our Internet Congress would be populated by corporate fucktards, and not long afterwards, anybody with a clue would be outnumbered and banned from the internet. Also, any mechanism that you create to ban folks from the internet will be used for governments for censorship. Great intention you've got, but that road leads straight to hell.
I have little idea what works for supercomputers and highly parallized data analysis (I've never used one).
Oh ok. So when you said Sage is okay for "small-midsize projects" and recommended Mathematica for large projects? You've only used Mathematica for small projects, and have no idea what a large-scale project is. And your pricing info is woefully outdated if you're recommending this to a new user.
Documentation is a skill. If you think it's not hard, you're either ignorant to your own failings, or those downstream of you are very lucky. Lots of people try hard and spend a long time on documentation and still fail to make it good. It *is* hard for most people to think of the many ways documentation needs to be used, the many mindsets of its users, and balance those two into concise, readable documentation.
Srsly. I type my SQL in through a keyboard. My handwriting and OCR do not play nicely together.
obligatory
I've used Sage on a supercomputer, chugging through hundreds of gigs of data. Do you know what you're talking about, or are you just recommending the shiny thing that you paid lots of money for?
I use Sage. When Python isn't fast enough, I can essentially write in C with Cython. It's gloriously easy. Have some trivially parallelizable data mining? Just use the @parallel decorator. Sage comes with a slew of fast mathematical packages, so your toolbox is massive, and you can hook it all in to your Cython code with minimal overhead.
Said novice computer user would probably be like 'who is this Eve person and why does this strange website want my credit card info?!?' and trash the email. The more likely targets are the semiliterate, who are familiar with the service. They'll fall for the "pay $3 and I'll send you $100!" bullshit, get greedy, and give their credit card to a non-genuine site without noticing. Or... straightup fall for the "pay $3 and I'll send you $100" bullshit and give their credit card to the completely genuine and complicit Square, and never report it 'cause they're embarrassed about getting duped.
RTFS. You only send the amount. Who you're paying and how much is insecure, but when you get an email from Square, they send you a link where you put in your bank info.
Of course, this will spawn a cottage industry of phishers, but you shouldn't worry about phishing -- you're already paranoid. Worry about your grandma getting phished.
Nope. This is nothing like TSP. The problem for this is "given a single new user, find the bus whose current route plan would be disturbed least by adding that user". Optimal point-to-point path finding in (nearly-)plane graphs is really quite easy, and these path deformations shouldn't be hard either. Besides... you only need reasonable approximations, not a global optimum.
You don't understand atheism. Atheists believe that there is no god. That takes faith, acknowledged or not. Belief zero? That's skepticism. But there is even extreme skepticism. Extreme skeptics don't believe that the world exists.
The problem is in the language. People hear "don't believe" to mean "believe the opposite". The very notion of undecideability just doesn't occur to people. I try to not believe things that I haven't seen proof of. Sometimes I fail in that.
Atheists who claim to be rational and faithless are either idiots or ignorant. If they don't realize they're agnostic they're simply ignorant. But hey, good news: ignorance can be fixed.
GP was quoting TFA, which was quoting a historian who knew more than what you gleaned from Wikipedia in 30 seconds. Check your own ignorance, friend. Nobody's questioning the existence of those children, merely the true identity of the father.
Forget about making busywork for people. Make robots. Let humans make art, learn, explore, teach, heal... those things humans are good at. Let robots do repetitive tasks. Focus on making people happy and healthy. Leave boredom to the robots. The creation of robots is undergoing a serious democratization, and these utopian ideals have a chance -- if only billionaires profit from the robot uprising, the world will be a terrible place. If we all profit, we can do what we love.
Right, I think electrical-analog has huge potential to achieve high intelligence. Turing machine brain? Awful idea.
Also... what is "true thought"? Have you or I ever had one?
Indeed. I've oft considered what would happen if a private police force were to arrest public police officers for "accidental" shootings... the shitstorm would be incredible.
Before you're going to build a computer that works like a human brain you're going to have to figure out how the human brain actually works.
Oh bullshit. I mean, you might be right and everything, but the certainty with which you make that statement is wholly unsupported. As long as we aim for a design where a human understands every little piece and every line of code, then you're certainly correct. If we take a hands-off approach and focus instead on analog components and neural networks, we might have surprises in store.