Dude, Shakespeare was recognized as a genius in his time - his plays had vast popular acclaim and were frequently attended by Queen Elizabeth. Likewise, Goethe was idolized by many (there were even some who considered him some sort of messiah) in his lifetime. Maybe there are some people on the list that weren't recognized until after they were dead (Blake and Nietzsche, though perhaps not in the book, come to mind), but choose a different example.
For 2000 bce to 800 bce, you forgot India. But anyway, your choice of dates is even more arbitrary than his; it looks to me like he went back as far as he could get reliable evidence, and as far forward as he could and still get a consensus on eminence. So his first spin is about as close to neutral as you can get. I guess your real objection is that he didn't try to distort the record to suit your PC notions of cultural equality.
Yes, thermite is easy to make. It also burns way, way hotter than you need to preheat something (so hot as to make containment a major engineering challenge for an application like this). Your suggestion makes as much sense as using thermite to cook spaghetti.
But you can already run all those operating systems at the same time - just get four different computers. If the hardware is two years old, all four of them together will probably cost less than this monster. Of course, communication between computers/CPUs/OSes will still be an issue - but oddly enough the article didn't give any details on Intel's solution to that problem either...
Yeah - though I suspect he was actually talking about a band saw (for his first accident) since as far as I can tell a table saw would not have been useful for cutting those legs. I think if he'd actually managed to cut himself on a table saw he would have needed more than one stitch. Anyway he sounds like an accident-prone guy - but perhaps he will learn.
Boris (who once drilled a hole through his finger)
Well, let's see: 1) Less sources of nuclear contamination
OK, the fuel may or may not be as bad as in a conventional fission reactor: I agree that I'd rather have a tritium leak than a plutonium one. But there will still be a neutron flux that will ultimately result in nuclear waste that needs to be disposed of. And unless you have an idea as to what the final configuration of a working fusion reactor is going to look like, I think it's premature to declare contamination less of an issue.
2) Theoretically more energy
Not terribly relevant since in theory both technologies can provide vast quantities of energy. In practice of course our fusion technology has yet to produce any positive energy returns.
3) A safe byproduct
Yes, the fact that the result of the fusion reaction is nice, innocuous helium is a big selling point, and goes a long way towards explaining all the interest in this technology. Too bad you also get neutrons (which in turn are absorbed to create various isotopes of the surrounding elements, aka radioactive waste). You might end up with less waste per unit of energy, but in the end you still have to solve the problem of radioactive waste disposal.
Advances in the evolution of all kinds of technology will continue to progress at an exponential pace;... Hand-waving nonsense. What does it mean for technology to advance at 'an exponential rate'? Is technology some sort of scalar quantity? And yes, I know about Moore's law, and I am aware that memory density is also following something like an exponential curve. Neither of those get you to nano-scale self-replication simply by getting bigger, though. If replication is so easy, why hasn't a replicator been built at the macro scale? What do you mean "if it can be made to work?" Well, I think it probably can be - I basically agree with you here. But being a skeptic and an engineer by nature I am always doubtful until implementation is actually successful.
...Fuel cells, large numbers of networked embedded computers, nanotech, and the hydrogen economy have all been around before
Nanotech has been around? What? The idea yes, actual implementation no. I am pessimistic about nanotech in the near term, but really, if it can be made to work you aren't going to need a lot of other new, new ideas.
Or we need to make some of the important older new ideas, like fusion power, work.
What does fusion power get you that a breeder reactor doesn't? You still have nuclear waste issues as a result of neutron emission. Even without (fission) breeder reactors, the thorium in the earth's crust would allow us to meet our energy needs through nuclear fission for some millenial span into the future - if we could deal with the environmental issues.
Re:Super ultra elite developers
on
The Bionic Office
·
· Score: 2, Insightful
Well, interestingly, I do think that there's something to the '10x as productive' idea. When I was at CMU there were a few freaks^H^H^H^H^H talented programmers who could throw down page after page of C code as fast as they could type, and I don't think it was worse, quality-wise, than what more mundane folks were able to achieve. But then, a lot of that speed came from implementing a vision that existed full-blown in their heads; in real life there's a lot of overhead in reading requirements, writing design docs and test framework, communicating, etc., which is not something that gets done ten times as fast.
Super ultra elite developers
on
The Bionic Office
·
· Score: 5, Insightful
One thing I've noticed, of which this article is a very good example, is how most everyone who hires software developers claims to be hiring (or at least looking for) the very best of the best.
'We have an elite team'.
'On a scale of one to ten, all our developers are at least a nine.'
'We hire only the top two percent.'
And of course in this article Joel kicks it up a notch by claiming to be after the 99.9th percentile. Makes you whether the industry is vastly deluded as to the actual abilities of those they hire...
And yet, I really liked Cryptonomicon and thought that all of the Dune books except the first one sucked. No arguing with taste, clearly. Unfortunately my memory of the works is not sharp enough to go on at length as to why I liked Cryptonomicon but not the later Dune books, but I doubt my reasons would be all that enlightening anyway... some like Stephenson and some don't...
The notion that "much of this wealth was a result of being in the right place at the right time" is patently absurd (although I guess it might make one feel better about not being rich). Luck may have a factor in presenting rare opportunities...
Precisely. If Bill Gates had been born four years earlier or later, or if Paul Allen had not met Gates, etc., do you think that he would still be worth tens of billions? I will gladly concede that his abilities would have made him a multimillionaire regardless of his circumstances, and that only one person in a hundred thousand could have taken his place - but that still means that most of his wealth could easily have ended up in someone else's hands, if certain random events had proceeded differently. I should make it clear that I am not arguing that the system is unfair, nor hoping for some sort of redistribution. My beef is with the idea that, simply because the capitalist system resulted in certain people having billions, we can tautologically state that their labor has/had that much value. Given the random factors mentioned above, such a claim makes no more sense than Aquinas' theories of intrinsic value.
For those individuals who made their wealth instead of inheriting it, by definition the wealth they've amassed is the "worth" of their labor... While the original poster was wrong in claiming they had stolen the money, this is also bogus. Chance has a hand in things - much of this wealth was a result of being in the right place at the right time. If I won the World Series of Poker (cash prize of nearly $2 million) would you say that my labors were worth that much? That would be a hard position to defend. Similarly, Gates, Ellison and Allen got their money fairly, after a fashion - they did not steal it - but claiming that they contributed that much *worth* to everyone else, rather than simply outgaming them, is not tenable.
The case might not have much merit, but there's not enough detail for us to decide that - and no apparent reason to dismiss it as 'bizarre'. Don't you remember that Apple sued M$ over the 'look and feel' of Windows? And if I wrote a program that exactly duplicated the functionality of Warcraft III (even if all the code was my own) do you think I wouldn't get sued by Blizzard? Everything depends on what the patents and copyrights cover.
Ron Paul was elected to the House as a Republican, but is really a libertarian. In fact he ran for the presidency on the Libertarian ticket in 1988 (I think). So it's no surprise that he's often at odds with the mainstream of the GOP.
These are cool. You can build any *shape* you want. Too bad you're limited to one (or a few) specific materials chosen more for their useability in this process than for other useful properties. What do you do when you need a copper winding for a motor? Iron core for a transformer? Hardened steel for a bearing race? Basically, you can use these to make toys, mockups, and maybe most of the parts for certain items. But don't expect them to replace real manufacturing anytime soon.
There's no definite answer to your question. You must judge the circumstances and make the call. Much as we'd like to do everything properly, quick and dirty is often first-to-market - and I've used plenty of products that had significant bugs and yet were adequate for my purpose.
OK, so there are ways to improve quality. Big deal. You can easily spend too many resources trying to improve quality beyond where it's useful. Japanese software shops are famous for trying to make things nearly perfect before they ship - one of the big reasons why they're not competitive worldwide. One of the key insights that Motorola had in deciding to implement six-sigma was that by achieving this level of manufacturing reliability they would be able to get rid of many of their downstream testing costs. But it doesn't automatically follow that everyone will benefit from trying to do that, or even from paying the costs to move from two-sigma to three-sigma quality. LCD manufacturers are a good example of an industry where the reject rate is huge (I'd say that for big screens they're at about *one* sigma quality) but they are still successful because they have enough downstream testing to compensate.
I read the book. I liked it. But: - we have yet to build a magnetic rail launcher of anything like the capabilities described in the book here on Earth, let alone the moon, and - we still have Mutual Assured Destruction. You may not like the doctrine, but in the end it means that even if the Chinese will soon have a novel way to annihilate us from their moon base, the balance of power remains the same.
Did you read the article? They *did* account for this possibility, by doing a second experiment in which they trained a group of people (a second subset of their volunteers played Tetris instead). And the test scores went up...
Yes - of course the claim that it was harmless to humans comes from the submitter, who seems to have misread the original article. There it only claims that collateral damage could be minimized (basically that it would be less destructive to humans than a 1000-kilo bomb...)
Quite the Baker fanboy, it seems. They did pioneer the use of N-grams, but other researchers were responsible for the use of HMMs. And they've been busy the past 15 years commercializing their work, not doing groundbreaking research. For what it's worth, L&H's internal evaluation of the different speech recognition technologies it had acquired gave the nod to a more modern recognition engine built by ISI in Pittsburgh - this technology was reacquired by the principal researchers subsequent to L&H's collapse, and they reincorporated as Multimodal Technologies. Pretending that Dragon's technology is somehow critical to this effort, when there are now half a dozen alternatives that are equally good or better available, makes no sense.
Slashdot Mirror
The Bell Curve (the book not the concept) has pretty much been completely discredited.
Just keep saying that often enough, and maybe people will believe it. Care to provide a link to this complete discrediting that you imagine?
Dude, Shakespeare was recognized as a genius in his time - his plays had vast popular acclaim and were frequently attended by Queen Elizabeth. Likewise, Goethe was idolized by many (there were even some who considered him some sort of messiah) in his lifetime. Maybe there are some people on the list that weren't recognized until after they were dead (Blake and Nietzsche, though perhaps not in the book, come to mind), but choose a different example.
For 2000 bce to 800 bce, you forgot India. But anyway, your choice of dates is even more arbitrary than his; it looks to me like he went back as far as he could get reliable evidence, and as far forward as he could and still get a consensus on eminence. So his first spin is about as close to neutral as you can get. I guess your real objection is that he didn't try to distort the record to suit your PC notions of cultural equality.
Don't teach your grandma to chew gum.
Yes, thermite is easy to make. It also burns way, way hotter than you need to preheat something (so hot as to make containment a major engineering challenge for an application like this). Your suggestion makes as much sense as using thermite to cook spaghetti.
But you can already run all those operating systems at the same time - just get four different computers. If the hardware is two years old, all four of them together will probably cost less than this monster. Of course, communication between computers/CPUs/OSes will still be an issue - but oddly enough the article didn't give any details on Intel's solution to that problem either...
Yeah - though I suspect he was actually talking about a band saw (for his first accident) since as far as I can tell a table saw would not have been useful for cutting those legs. I think if he'd actually managed to cut himself on a table saw he would have needed more than one stitch. Anyway he sounds like an accident-prone guy - but perhaps he will learn.
Boris (who once drilled a hole through his finger)
Well, let's see:
1) Less sources of nuclear contamination
OK, the fuel may or may not be as bad as in a conventional fission reactor: I agree that I'd rather have a tritium leak than a plutonium one. But there will still be a neutron flux that will ultimately result in nuclear waste that needs to be disposed of. And unless you have an idea as to what the final configuration of a working fusion reactor is going to look like, I think it's premature to declare contamination less of an issue.
2) Theoretically more energy
Not terribly relevant since in theory both technologies can provide vast quantities of energy. In practice of course our fusion technology has yet to produce any positive energy returns.
3) A safe byproduct
Yes, the fact that the result of the fusion reaction is nice, innocuous helium is a big selling point, and goes a long way towards explaining all the interest in this technology. Too bad you also get neutrons (which in turn are absorbed to create various isotopes of the surrounding elements, aka radioactive waste). You might end up with less waste per unit of energy, but in the end you still have to solve the problem of radioactive waste disposal.
Advances in the evolution of all kinds of technology will continue to progress at an exponential pace; ...
Hand-waving nonsense. What does it mean for technology to advance at 'an exponential rate'? Is technology some sort of scalar quantity? And yes, I know about Moore's law, and I am aware that memory density is also following something like an exponential curve. Neither of those get you to nano-scale self-replication simply by getting bigger, though. If replication is so easy, why hasn't a replicator been built at the macro scale?
What do you mean "if it can be made to work?"
Well, I think it probably can be - I basically agree with you here. But being a skeptic and an engineer by nature I am always doubtful until implementation is actually successful.
...Fuel cells, large numbers of networked embedded computers, nanotech, and the hydrogen economy have all been around before
Nanotech has been around? What? The idea yes, actual implementation no. I am pessimistic about nanotech in the near term, but really, if it can be made to work you aren't going to need a lot of other new, new ideas.
Or we need to make some of the important older new ideas, like fusion power, work.
What does fusion power get you that a breeder reactor doesn't? You still have nuclear waste issues as a result of neutron emission. Even without (fission) breeder reactors, the thorium in the earth's crust would allow us to meet our energy needs through nuclear fission for some millenial span into the future - if we could deal with the environmental issues.
Well, interestingly, I do think that there's something to the '10x as productive' idea. When I was at CMU there were a few freaks^H^H^H^H^H talented programmers who could throw down page after page of C code as fast as they could type, and I don't think it was worse, quality-wise, than what more mundane folks were able to achieve. But then, a lot of that speed came from implementing a vision that existed full-blown in their heads; in real life there's a lot of overhead in reading requirements, writing design docs and test framework, communicating, etc., which is not something that gets done ten times as fast.
One thing I've noticed, of which this article is a very good example, is how most everyone who hires software developers claims to be hiring (or at least looking for) the very best of the best.
'We have an elite team'.
'On a scale of one to ten, all our developers are at least a nine.'
'We hire only the top two percent.'
And of course in this article Joel kicks it up a notch by claiming to be after the 99.9th percentile. Makes you whether the industry is vastly deluded as to the actual abilities of those they hire...
And yet, I really liked Cryptonomicon and thought that all of the Dune books except the first one sucked. No arguing with taste, clearly. Unfortunately my memory of the works is not sharp enough to go on at length as to why I liked Cryptonomicon but not the later Dune books, but I doubt my reasons would be all that enlightening anyway... some like Stephenson and some don't...
The notion that "much of this wealth was a result of being in the right place at the right time" is patently absurd (although I guess it might make one feel better about not being rich). Luck may have a factor in presenting rare opportunities...
Precisely. If Bill Gates had been born four years earlier or later, or if Paul Allen had not met Gates, etc., do you think that he would still be worth tens of billions? I will gladly concede that his abilities would have made him a multimillionaire regardless of his circumstances, and that only one person in a hundred thousand could have taken his place - but that still means that most of his wealth could easily have ended up in someone else's hands, if certain random events had proceeded differently.
I should make it clear that I am not arguing that the system is unfair, nor hoping for some sort of redistribution. My beef is with the idea that, simply because the capitalist system resulted in certain people having billions, we can tautologically state that their labor has/had that much value. Given the random factors mentioned above, such a claim makes no more sense than Aquinas' theories of intrinsic value.
For those individuals who made their wealth instead of inheriting it, by definition the wealth they've amassed is the "worth" of their labor...
While the original poster was wrong in claiming they had stolen the money, this is also bogus. Chance has a hand in things - much of this wealth was a result of being in the right place at the right time. If I won the World Series of Poker (cash prize of nearly $2 million) would you say that my labors were worth that much? That would be a hard position to defend. Similarly, Gates, Ellison and Allen got their money fairly, after a fashion - they did not steal it - but claiming that they contributed that much *worth* to everyone else, rather than simply outgaming them, is not tenable.
The case might not have much merit, but there's not enough detail for us to decide that - and no apparent reason to dismiss it as 'bizarre'. Don't you remember that Apple sued M$ over the 'look and feel' of Windows? And if I wrote a program that exactly duplicated the functionality of Warcraft III (even if all the code was my own) do you think I wouldn't get sued by Blizzard? Everything depends on what the patents and copyrights cover.
Ron Paul was elected to the House as a Republican, but is really a libertarian. In fact he ran for the presidency on the Libertarian ticket in 1988 (I think). So it's no surprise that he's often at odds with the mainstream of the GOP.
These are cool. You can build any *shape* you want. Too bad you're limited to one (or a few) specific materials chosen more for their useability in this process than for other useful properties. What do you do when you need a copper winding for a motor? Iron core for a transformer? Hardened steel for a bearing race?
Basically, you can use these to make toys, mockups, and maybe most of the parts for certain items. But don't expect them to replace real manufacturing anytime soon.
At least according to the link in the technical dossiers...
There's no definite answer to your question. You must judge the circumstances and make the call. Much as we'd like to do everything properly, quick and dirty is often first-to-market - and I've used plenty of products that had significant bugs and yet were adequate for my purpose.
OK, so there are ways to improve quality. Big deal. You can easily spend too many resources trying to improve quality beyond where it's useful. Japanese software shops are famous for trying to make things nearly perfect before they ship - one of the big reasons why they're not competitive worldwide. One of the key insights that Motorola had in deciding to implement six-sigma was that by achieving this level of manufacturing reliability they would be able to get rid of many of their downstream testing costs. But it doesn't automatically follow that everyone will benefit from trying to do that, or even from paying the costs to move from two-sigma to three-sigma quality. LCD manufacturers are a good example of an industry where the reject rate is huge (I'd say that for big screens they're at about *one* sigma quality) but they are still successful because they have enough downstream testing to compensate.
I read the book. I liked it. But:
- we have yet to build a magnetic rail launcher of anything like the capabilities described in the book here on Earth, let alone the moon, and
- we still have Mutual Assured Destruction. You may not like the doctrine, but in the end it means that even if the Chinese will soon have a novel way to annihilate us from their moon base, the balance of power remains the same.
Did you read the article? They *did* account for this possibility, by doing a second experiment in which they trained a group of people (a second subset of their volunteers played Tetris instead). And the test scores went up...
Yes - of course the claim that it was harmless to humans comes from the submitter, who seems to have misread the original article. There it only claims that collateral damage could be minimized (basically that it would be less destructive to humans than a 1000-kilo bomb...)
Quite the Baker fanboy, it seems. They did pioneer the use of N-grams, but other researchers were responsible for the use of HMMs. And they've been busy the past 15 years commercializing their work, not doing groundbreaking research.
For what it's worth, L&H's internal evaluation of the different speech recognition technologies it had acquired gave the nod to a more modern recognition engine built by ISI in Pittsburgh - this technology was reacquired by the principal researchers subsequent to L&H's collapse, and they reincorporated as Multimodal Technologies. Pretending that Dragon's technology is somehow critical to this effort, when there are now half a dozen alternatives that are equally good or better available, makes no sense.