Hardly. Is there anybody here who wouldn't put cuddling a teenage Ally ahead of playing with our computer? Of course, the opportunity would never arise. So I guess the unrealistic part is that she was even in his bedroom in the first place.
I must have missed the passage in the bible where it says, "Thou shalt not have female clergy." My issue with female rabbis is that Judaism is such a profoundly patriarchal faith (there's even a prayer where men thank God that they're not women) that it just feels wrong. Which is totally hypocritical of me, since I don't even believe in God these days, but do believe that women should not be arbitrarily excluded from any profession.
But even if we just focus on AI, I don't see us creating — or moding ourselves into — cybernetic gods anytime soon. This has nothing to do with Moore's law, computing power density or whatever. You don't need to cram that extra machinery into your skull, you just have to have a good interface to it. So even current technologies are perfectly adequate.
So why don't we already have cybernetic supermen walking among us? <i>Because we don't understand how natural intelligence works.</i> There's a lot of interesting work, but it's not even close to modeling all the behavior of a simple insect brain, never mind a human.
When we do have that understanding, we might indeed see people with their natural intelligence augmented by artificial intelligence. But I think that will be a lot messier than the "singularity" bozos assume. Intelligence is <i>complicated</i>, and each attempt to augment has to be correspondingly complicated. So rather than a smooth hyperbolic curve, I think we'll see a series of irregular jerks.
All of which assumes that there's still a solid scientific infrastructure to support this work. That, in turn, means that there has to be a functioning human culture with enough extra resources to finance that infrastructure. And <i>that</i> means dealing with a host of issues — resource depletion, environmental degradation, inadequate water supplies, shrinking food production, the population explosion — that are nowhere near a solution. (You'll notice I don't mention the Crisis Du Jour Gl-b-l W-rm-ng, which is actually just one component of the preceding.) These problems are not insoluble, but nobody seems to want to face them. And pretending that we're all going to go to Singularity Heaven seems to be a popular way of not thinking about them.
(Sorry for the raw HTML, but only the Code format seems to be working for long posts.)
I must admit that I've read nothing about "singularity theory" beyond the work of a few SF writers infatuated with the concept. But these include Vernor Vinge, who invented the concept. In these stories, it just isn't AI that jumps off into infinity, its science and technology of every sort. In some of Vinge's stories, this results in the entire human race simply disappearing (except for a few people trapped in time stasis fields), presumably to some godlike state beyond our ken. That's clearly not just about AI.
(Stupid Slashdot doesn't want to let me post more than a paragraph! Will try to continue in a separate post.)
I have never met anyone actually claiming that Moore's "law" was anything more then a semi-coincidental trend that will seize to exist soon. No? This thread started when somebody cited ML as an example of trend towards a "singularity".
The question is, how soon. The discussion is whether the moment at which computers can't increase in capacity anymore lies before or after their ability to be intelligent. In other words, Moore's law is indeed the relevant topic. Relevent for 10 years. Probably less. I don't think we're going to see serious AI in that time. Maybe after we move to spintronics, quantum computing, or (very likely) some technology that hasn't even been imagine yet.
There are tons of interesting side discussion, like what the hell is 'intelligence' anyways? When we talk about "the singularity" that's hardly a side issue. Nor is it one we're going to settle here. But I wouldn't concede that we've made serious progress until (for example) there's software that can always tell when a given word is being used as a verb (time flies like an arrow) and when it's being used as a noun (fruit flies like a banana).
Hey, it's easier than being a lapsed Jew. Religiously, I shouldn't care. Socially, I should be all for it. But deep down inside, it really bothers me that there are now women rabbis!
Sorry, I committed the same sin you did earlier, and stopped reading after that first statement. (It was a pretty sprawling generalization.) So this is the meat of your argument:
The present graphs of technological increase are remarkably linear; there's no sign at all of a slowdown in the doublings, which are approximately yearly. What graphs are you talking about? The one for practical fusion power, which has been "fifty years from now" for decades? The graphs for speed of travel, which plateaued 40 years ago? The graphs for energy reserves, which indicate that more than half of all the fossil fuel that ever existed has been burned in the last couple of centuries? The standard assumption that machines would get smarter and smarter (faster and faster is not the same thing) with human-level AIs commonplace by the end of the 20th?
Not to mention the price of food. Little detail: hungry people don't do good science.
I assume you have some indicators you like better. By all means let's hear them. But if you can find a trend that can't possibly plateau in the next 50 years, I'd be very surprised.
Hey, clerical abuses (no offense, but this mandatory celibacy thing is just not working), Roman Catholicism is pretty un-creepy. Then again, maybe it's because we're all used to it. I imagine that first time Constantine the Great told exactly what Transubstantiation was about, his first reaction was "Euuuu!"
And my point was that just when you think it's going to hit a wall, it changes form and continues despite your pessimism. Always? Sounds pretty religious to me.
Honestly, there is nothing that falls so flat on my ears as doomlike predictions of technology reaching an unbreachable wall. Not going to happen. Never said it would. If you go back and read my post past the first sentence, you'll find I said exactly the opposite.
My only point here is that exponential growth never lasts forever. It's only when your mathematical models are extremely simplistic that your curve vanished up into infinity. Real world models crash when you run through all the resources that sustained growth in the first place.
Technology will continue to improve exponentially (it is right now - see Moore's law) Dude, Moore's "law" will cease to apply in just a few years. It's not a law of nature, it's an observation that chip fab facilities double their resolution every 18 months. But you can only take that so far. Soon, quantum tunneling will make it impossible to print circuits any finer.
Understand, I'm not claiming that progress in creating smaller and cheaper circuits will suddenly come to a halt. There's too much economic demand for that to happen. But development will have to switch to new technologies, and progress almost certainly will be slower.
Every kind of growth has limits. If it didn't bacteria would still be the most conspicuous life form on the planet. Science and technology are no different. They're dependent on natural resources, an educated pool of skilled workers, and a lot of other factors.
Then it is a fair assumption that we will eventually design a superbrain. The superbrain will design a super-duper brain, and the chain reaction (singularity) will be upon us. I can't wait! Again with the silly curve drawing. "Smarter" isn't something you can make just by packing in enough circuits. People don't even know what "smarter" is. Probably they'll figure it out eventually. But it's already clear that "smarter" is more complicated than any technology we'll see in the next century or so.
Schneier is not so much obvlivious as in love with his own ideas, sometimes at the cost of his logical consistency.
Really, signatures are not "proof" of anything, and never have been. Back when many people were illiterate, simply making a mark was an acceptable signature. A signature is just a sign of an agreement that is sustained by collective memory, not the signature itself.
For example, how do we know that John Hancock signed the Declaration of Independence? It's not because his handwriting is hard to forge. (I wanted to say that it's because a lot of people saw him sign it, but that turns out to be a myth.) No, it's part of the collective memory of the time: Hancock was the presiding officer of the Continental Congress, and would have had to sign it; he acknowledged signing it; etc. etc.
Well, I was a satisfied customer of DreamHost for several years. Then they began to have a long series of outages that made them pretty useless.
But lets just say that Planet is different, and this incident doesn't represent their normal competency. The fact is, their power system blew up. Literally! Not a routine event. Maybe, just maybe, it was an accident they couldn't have anticipated. But it's a lot more likely that this represents somebody's incompetence. If not the systemic incompetence that drove me away from DreamHost, then the isolated incompetence you see at even the best-run companies.
Of course, I don't know for sure one way or the other how this thing happened. I will tell you that if that if this "Murphy's Law" nonsense had been spouted by somebody working at Planet, I would have taken it as a sign that they didn't have their act together.
This goes to show that no matter how much planning you do, Murphy's Law still applies. I am so tired of hearing that copout. Does the submitter know for a fact that ThePlanet did everything it could to keep its power system from exploding? I don't have any evidence one way or the other, but if they're anything like other independent data center operators, it's pretty unlikely.
The lesson you should be taking from Murphy's Law is not "Shit Happens". The lesson you should be taking is that you can't assume that an unlikely problem (or one you can con yourself into thinking unlikely) is one you can ignore. It's only after you've prepared for every reasonable contingency that you're allowed to say "Shit Happens".
Not only lacking in technical details, but most of the non-technical details are wrong. A techie at Roku is the wrong person to ask about how Netflix chooses movies for instant watching. And it was obvious the interviewer hadn't read the description of the device on Netflix.com.
The first thing Novell did when they acquired SuSE was change the name to SUSE. Sort of thing that separates the amateurs from the pros. Like when Intergalactic Digital Research switched to a grownup name.
As a scientist raised on a heavy diet of open source software and computational physics, I'll hang my head in shame if our majors start proudly putting Excel down on their resumes. And why is it better for a physicist to be able to put "C++" on their resume? Anybody of average intelligence can stumble their way through an elementary programming course. Doesn't make them an expert scientific programmer. That can only come through applying programming techniques as part of a larger study.
I was once in a moderately prestigious science undergrad program. There were no programming classes as part of the curriculum. And no, this wasn't an oversight on the part of the people who designed the program. Because this was a computer science program.
(Full disclosure: I didn't make it through. Had to admit that in case any of the profs who flunked me out read this.)
That's not uncommon. MIT probably rates as the leading computer science school in the country, maybe the world. I don't know much about their program as a whole, but I do know their key freshman course spends almost no time describing any programming language. Of course, it helps that they do all their actually programs in Scheme, a language whose syntax you can write on the back of an envelope.
Which is not to say that Scheme is a trivial language. Quite the contrary. But I doubt that anybody who makes it through SICP puts "Scheme" on their resume. No, they talk about their newfound knowledge of algorithmics and data structures.
If you want to make your students skilled scientific programmers, don't make them take programming classes. Create physics classes where programming is a key element. Programming is just a tool, and do you have a class for every tool your budding physicists might use?
Slashdot Mirror
Does it matter whether you're black or white?
Hardly. Is there anybody here who wouldn't put cuddling a teenage Ally ahead of playing with our computer? Of course, the opportunity would never arise. So I guess the unrealistic part is that she was even in his bedroom in the first place.
If I was a teenager alone in my room with Ally Sheedy, the computer would get very little attention.
"Clone" is a bit of an overstatement. Kvivio doesn't even begin to have the functionality of Visio.
I must have missed the passage in the bible where it says, "Thou shalt not have female clergy." My issue with female rabbis is that Judaism is such a profoundly patriarchal faith (there's even a prayer where men thank God that they're not women) that it just feels wrong. Which is totally hypocritical of me, since I don't even believe in God these days, but do believe that women should not be arbitrarily excluded from any profession.
(continued)
But even if we just focus on AI, I don't see us creating — or moding ourselves into — cybernetic gods anytime soon. This has nothing to do with Moore's law, computing power density or whatever. You don't need to cram that extra machinery into your skull, you just have to have a good interface to it. So even current technologies are perfectly adequate.
So why don't we already have cybernetic supermen walking among us? <i>Because we don't understand how natural intelligence works.</i> There's a lot of interesting work, but it's not even close to modeling all the behavior of a simple insect brain, never mind a human.
When we do have that understanding, we might indeed see people with their natural intelligence augmented by artificial intelligence. But I think that will be a lot messier than the "singularity" bozos assume. Intelligence is <i>complicated</i>, and each attempt to augment has to be correspondingly complicated. So rather than a smooth hyperbolic curve, I think we'll see a series of irregular jerks.
All of which assumes that there's still a solid scientific infrastructure to support this work. That, in turn, means that there has to be a functioning human culture with enough extra resources to finance that infrastructure. And <i>that</i> means dealing with a host of issues — resource depletion, environmental degradation, inadequate water supplies, shrinking food production, the population explosion — that are nowhere near a solution. (You'll notice I don't mention the Crisis Du Jour Gl-b-l W-rm-ng, which is actually just one component of the preceding.) These problems are not insoluble, but nobody seems to want to face them. And pretending that we're all going to go to Singularity Heaven seems to be a popular way of not thinking about them.
(Sorry for the raw HTML, but only the Code format seems to be working for long posts.)
I must admit that I've read nothing about "singularity theory" beyond the work of a few SF writers infatuated with the concept. But these include Vernor Vinge, who invented the concept. In these stories, it just isn't AI that jumps off into infinity, its science and technology of every sort. In some of Vinge's stories, this results in the entire human race simply disappearing (except for a few people trapped in time stasis fields), presumably to some godlike state beyond our ken. That's clearly not just about AI.
(Stupid Slashdot doesn't want to let me post more than a paragraph! Will try to continue in a separate post.)
Hey, it's easier than being a lapsed Jew. Religiously, I shouldn't care. Socially, I should be all for it. But deep down inside, it really bothers me that there are now women rabbis!
Not to mention the price of food. Little detail: hungry people don't do good science.
I assume you have some indicators you like better. By all means let's hear them. But if you can find a trend that can't possibly plateau in the next 50 years, I'd be very surprised.
Hey, clerical abuses (no offense, but this mandatory celibacy thing is just not working), Roman Catholicism is pretty un-creepy. Then again, maybe it's because we're all used to it. I imagine that first time Constantine the Great told exactly what Transubstantiation was about, his first reaction was "Euuuu!"
My only point here is that exponential growth never lasts forever. It's only when your mathematical models are extremely simplistic that your curve vanished up into infinity. Real world models crash when you run through all the resources that sustained growth in the first place.
The outlook is uncertain.
Understand, I'm not claiming that progress in creating smaller and cheaper circuits will suddenly come to a halt. There's too much economic demand for that to happen. But development will have to switch to new technologies, and progress almost certainly will be slower.
Every kind of growth has limits. If it didn't bacteria would still be the most conspicuous life form on the planet. Science and technology are no different. They're dependent on natural resources, an educated pool of skilled workers, and a lot of other factors. Then it is a fair assumption that we will eventually design a superbrain. The superbrain will design a super-duper brain, and the chain reaction (singularity) will be upon us. I can't wait! Again with the silly curve drawing. "Smarter" isn't something you can make just by packing in enough circuits. People don't even know what "smarter" is. Probably they'll figure it out eventually. But it's already clear that "smarter" is more complicated than any technology we'll see in the next century or so.
Schneier is not so much obvlivious as in love with his own ideas, sometimes at the cost of his logical consistency.
Really, signatures are not "proof" of anything, and never have been. Back when many people were illiterate, simply making a mark was an acceptable signature. A signature is just a sign of an agreement that is sustained by collective memory, not the signature itself.
For example, how do we know that John Hancock signed the Declaration of Independence? It's not because his handwriting is hard to forge. (I wanted to say that it's because a lot of people saw him sign it, but that turns out to be a myth.) No, it's part of the collective memory of the time: Hancock was the presiding officer of the Continental Congress, and would have had to sign it; he acknowledged signing it; etc. etc.
What? Our singular overlords? Our eschatonic overlords.
Is it just me, or does all this poorly-reasoned "singularity" crap have a religious feel to it?
"His biggest disappoint was that he never made the FBI's 10 Most Wanted List. It's all about who you know."
Well, I was a satisfied customer of DreamHost for several years. Then they began to have a long series of outages that made them pretty useless.
But lets just say that Planet is different, and this incident doesn't represent their normal competency. The fact is, their power system blew up. Literally! Not a routine event. Maybe, just maybe, it was an accident they couldn't have anticipated. But it's a lot more likely that this represents somebody's incompetence. If not the systemic incompetence that drove me away from DreamHost, then the isolated incompetence you see at even the best-run companies.
Of course, I don't know for sure one way or the other how this thing happened. I will tell you that if that if this "Murphy's Law" nonsense had been spouted by somebody working at Planet, I would have taken it as a sign that they didn't have their act together.
You can talk about the 19 part.
The lesson you should be taking from Murphy's Law is not "Shit Happens". The lesson you should be taking is that you can't assume that an unlikely problem (or one you can con yourself into thinking unlikely) is one you can ignore. It's only after you've prepared for every reasonable contingency that you're allowed to say "Shit Happens".
Not only lacking in technical details, but most of the non-technical details are wrong. A techie at Roku is the wrong person to ask about how Netflix chooses movies for instant watching. And it was obvious the interviewer hadn't read the description of the device on Netflix.com.
You're either a very good typist, or a very bad one.
The first thing Novell did when they acquired SuSE was change the name to SUSE. Sort of thing that separates the amateurs from the pros. Like when Intergalactic Digital Research switched to a grownup name.
I was once in a moderately prestigious science undergrad program. There were no programming classes as part of the curriculum. And no, this wasn't an oversight on the part of the people who designed the program. Because this was a computer science program.
(Full disclosure: I didn't make it through. Had to admit that in case any of the profs who flunked me out read this.)
That's not uncommon. MIT probably rates as the leading computer science school in the country, maybe the world. I don't know much about their program as a whole, but I do know their key freshman course spends almost no time describing any programming language. Of course, it helps that they do all their actually programs in Scheme, a language whose syntax you can write on the back of an envelope.
Which is not to say that Scheme is a trivial language. Quite the contrary. But I doubt that anybody who makes it through SICP puts "Scheme" on their resume. No, they talk about their newfound knowledge of algorithmics and data structures.
If you want to make your students skilled scientific programmers, don't make them take programming classes. Create physics classes where programming is a key element. Programming is just a tool, and do you have a class for every tool your budding physicists might use?