Because it's almost literally impossible for someone to actually put in all of the work required to publish hundreds of papers during their career. A paper might typically take six months of gruelling, full-time work. Instead of actually doing the work, what a lot of scientists do is they bring in a lot of students and act as project supervisors, as it says in the article: "Many of these prolific scientists are likely the heads of laboratories or research groups; they bring in funding, supervise research, and add their names to the numerous papers that result." In other words, they drop in for maybe half an hour every two weeks or so to get an 'update' (without really understanding anything), throw around some bs pieces of 'advice' (which everyone ignores) and then leave.
You bootstrap it by not trying to transfer all of this to space at once. Start with just a simple plan that takes small asteroids and brings them (or chunks of them) over to Earth orbit for processing. Stuff that's hard to build (like computers) are usually lightweight. Send them up from Earth in bulk.
But all of this is beside the point. That asteroid mining is difficult I completely concede. But how would sticking humans into the equations fix anything at all? Any gain in repair ability would be at the expense of a huge amount of additional complexity and risk in keeping the humans alive and functioning.
It's worth pointing out that all existing practical proposals for Mars colonization that I've seen involve sending hard-to-manufacture supplies (basically anything other than structural materials) to the colonies for at least a century afterwards. If that's what you're going to do then why not just cut humans out of the equation.
You hit the nail on the head. The problem isn't AI. The problem is that scaling it up to industrial levels is hard. And the absolute WORST way of solving that problem is making manned mining ships, raising the complexity, cost, and risk by 100x.
Why does it have to? Build them out of cheap, expendable, easily-replaceable parts. Which you can do, because the entire premise of asteroid mining is that it will make the cost of building and deploying space equipment dead cheap (otherwise why do it in the first place?)
We don't? We already have AI that can autonomously land on other bodies and extract material. In fact we've had it for 4 decades. See: almost any planetary lander/rover ever. It seems the barrier to mining is more up-front cost and on-site materials processing than AI.
I'm a techno-optimist but I agree with you. The rate things are going, it isn't going to make much sense to have people living in space colonies. I can't think of any good reason to do it other than the coolness factor. Unfortunately a lot of people are emotionally invested in this idea and will fight logic tooth and nail to promote their fantasies.
But that's just begging the question of why they didn't use liquid-fuelled boosters instead. As I said, the crazy and conflicting design requirements (for example, the 'need' to launch satellites into polar orbit, the unnecessary focus on re-usability, and the decision to have lower up-front cost at the expense of greater future operating costs) contributed to these flawed decisions.
I've been actively working in the field for the past few years and I don't think he's incredibly off the mark. Google, for instance, has some pretty advanced tech in production and lots more in development. The 'new AI' (statistical machine learning and large-scale, distributed data mining) is getting pretty advanced and scary.
Symbolic manipulation as a route to AI was a period of collective delusion in computer science. Lots of people wasted their talents going down this route. In the 80's this approach was all but dead and AI researchers finally sobered up. They started actually learning about the human brain and incorporating the lessons into their designs. It's sad that so much time was wasted on that approach, but the good news is that the new approaches people are using now are based on actual science and grounded in reality. The intelligence in search, natural language, object and facial recognition, and self-driving cars (that ShanghaiBill pointed out) is due to these new approaches.
AI spent its youth confused and rebellious. That was when you were in your graduate studies. Now it's far more matured. I encourage you to read up on new machine intelligence approaches and the literature in this area. You won't be disappointed.
The problem isn't reuse of old technology. The problem is the selection of old technology you reuse, and how you go about re-using it.
Start with the solid rocket boosters as an example. There's very good reasons why most space launch platforms don't use solid rockets. Cost, efficiency, and inherent lack of safety (you can't turn off an SRB once it's been lit) are just a few. So why did the shuttle use them? Because it was kind of forced upon them by the crazy and contradictory design decisions they had to comply with. The end result was that 70% of the takeoff thrust was actually provided by the two solid boosters, with only 30% coming from the three high-tech hydrogen rockets.
With the SLS, NASA had the opportunity to fix the warts in the shuttle program. Instead what we have now is the maximum-pork option.
It's probably not the case that she wrote the paper cackling to herself madly and proclaiming "Those suckers will never find out!"
It's probably the case that through self-delusion and carelessness she managed to partially convince herself that the results were true, and this, coupled with pressure to produce results, caused her to take a few shortcuts to get it published. What she did was wrong, and her career is over. It's not something a rational mind would have done. But scientists are just human and sometimes prone to making irrational decisions. The great thing is that we have the scientific method to weed out the good from the bad.
The clean-up was less due to the severe amount of radioactivity and more due to the fact that he was careless and got it everywhere.
The total amount of radioactive material was small and the actual dose of radiation he was exposed to was probably minimal. Although the exact dose isn't known because he never completely revealed his experiments and he never underwent testing.
One thing I find interesting is that he was arrested again in 2007 on charges related to stealing smoke detectors for their Americium, 13 years after his boy scout experiments.
Yup, that's exactly it. Another thing about those computers is that they are often more 'geek-oriented' overall. I got a vostro 1320 laptop with freedos, for instance, and it has a backside panel for easy access to the fan and heatsink assembly. Cleaning the computer is literally a matter of just removing a pair of screws. Compare to some other laptops where you virtually have to disassemble down to the bare motherboard just to get the fan clean. It was also about $90 cheaper than the windows version, even though the specs were almost exactly the same. It's nice that companies still make computers like this. I wonder when it will end.
As the article says, that's irrelevant because it's a tribe affiliation. When it comes to these things, logic and intelligence are completely overwhelmed and suppressed. That's how our brains work. Ever notice how all so-called "freethought" and "rationalist" groups soon turn into hilariously ironic examples of conformity and groupthink?
At this point, there's actually no need for century-old data (although it does add to the increasing pool of evidence). Weather satellites have been up in space for long enough now to establish warming beyond any reasonable doubt.
I actually do know a few people who've obtained comfortable office jobs without degrees, but it was very hard for them and they had to go through personal connections. YMMV.
About the 'remote classroom', It's been my experience that these are mostly projects to reduce costs for the guys at the top without really changing the system. The real advances have come from initiatives from professors themselves, but their hands are tied. A professor can give you the best education but he can't give you a degree.
Ultimately, it's going to take cultural change. Slowly, people will realize that an online education can be just as good as a brick-and-mortar one, depending. But as I said, the guys at the top will do ANYTHING to prevent this, even if it means devious subversion and spreading FUD (which many people will fall for, hook line and sinker).
Re:Perl 6ers just can't get shit done.
on
Perl Is Undead
·
· Score: 1
> that is, replace your loop conditionals with direct jumps to the next instruction handler
This would accomplish nothing as GCC already highly optimizes loop conditionals with jumps.
> And if you think using LLVM to write an interpreter is easier, than you're just... I don't even know what... totally out of touch and have no clue how this stuff works.
I didn't suggest that the perl6 team use LLVM to write an interpreter. I'm just saying that examples exist of people successfully implementing interpreters in this way (using a high-level frontend and compilation to bytecode for performance) for fairly complex languages, without major development hell like perl6 experienced. The resulting code is often simpler than C (really -- have you ever looked at the source code for CPython?), even considering the added complexity of generating bytecode. Thus your conclusion (that the reason they failed was because they didn't use C) doesn't follow from your premises. It's more likely that they were just incompetent and bad at language design and implementation. Perhaps LLVM is a bad choice for perl6, but again, that boils down to language design.
> The VM isn't the problem. It's all the bells and whistles. The typing and object system.
I couldn't agree more that feature creep is the enemy when it comes to language implementation. Just look at C++.
Re:Perl 6ers just can't get shit done.
on
Perl Is Undead
·
· Score: 3, Insightful
You're implying that C-based interpreters are quicker and easier to develop, but that's patently not the case. Just look at the recent huge surge of LLVM-based JIT languages that have been developed extremely rapidly and in high-level languages. C is probably a poor choice now that good alternatives really exist.
Maybe the problem isn't the choice of language. Maybe the problem is that the designers were incompetent.
You still need a college degree to get a white-collar job, and the fact is that most people want white-collar jobs. I 100% agree with you that it's an evil system designed to exploit the vulnerable, but it's not just an empty sales pitch like you're describing. The system is rigged in such a way that a lot of students HAVE to be burdened with massive debt to succeed in life. We have ZERO need for huge buildings with lecture halls in them in the age of the internet, but those people at the top are going to do anything it takes to keep this cash cow flowing.
Obviously, it's flat-out stupid to spend this debt getting a liberal arts degree or anything that doesn't have good job prospects.
It's easy to level hate on administrators, and it's true that there are far more administrative staff than needed on most campuses. But a lot of universities have recently drastically slashed down on administrative staff (including my own), and what happened? Surprise, surprise, tuition fees are still high, professors are still scrambling after grants, and instructors are still poor. Let's not lose sight of the real culprits here.
> you'd have a colony full of presumably quite intelligent, capable, and emotionally stable people (we did screen them for the hardships of colony life, right?)
They're still human and could act unpredictably under pressure. And even if they were angels, there's no guarantee that their children would be the same.
> And Earth is unlikely to have a reason to get Mars involved - unlike orbital and possibly lunar colonies which might be able to play a tactical role in terrestrial battles,
If there's no reason for Mars to get involved then there's no reason to build a Mars colony. The whole point of interplanetary colonization is economic expansion. At some point there is going to be trade between Earth and Mars. And I mean trade of all kinds, not just raw materials. For instance, trade of intellectual property developed by Martian colonists.
Plus, you're going to have a lot of people with cultural ties to the various nations on Earth. What if a war between the USA and China breaks out - how will American-descended colonists get along with Chinese-descended colonists?
But all of this is ignoring the elephant in the room - human politics is based on brinksmanship, and always has been. It's our nature. And yes, it even exists among 'smart' groups (in fact, it's more pronounced in 'smart' groups). We survived in spite of this because of sheer numbers and resilience. But things won't be the same on Mars. You can't have brinksmanship in a fragile colony.
Slashdot Mirror
Because it's almost literally impossible for someone to actually put in all of the work required to publish hundreds of papers during their career. A paper might typically take six months of gruelling, full-time work. Instead of actually doing the work, what a lot of scientists do is they bring in a lot of students and act as project supervisors, as it says in the article: "Many of these prolific scientists are likely the heads of laboratories or research groups; they bring in funding, supervise research, and add their names to the numerous papers that result." In other words, they drop in for maybe half an hour every two weeks or so to get an 'update' (without really understanding anything), throw around some bs pieces of 'advice' (which everyone ignores) and then leave.
You bootstrap it by not trying to transfer all of this to space at once. Start with just a simple plan that takes small asteroids and brings them (or chunks of them) over to Earth orbit for processing. Stuff that's hard to build (like computers) are usually lightweight. Send them up from Earth in bulk.
But all of this is beside the point. That asteroid mining is difficult I completely concede. But how would sticking humans into the equations fix anything at all? Any gain in repair ability would be at the expense of a huge amount of additional complexity and risk in keeping the humans alive and functioning.
It's worth pointing out that all existing practical proposals for Mars colonization that I've seen involve sending hard-to-manufacture supplies (basically anything other than structural materials) to the colonies for at least a century afterwards. If that's what you're going to do then why not just cut humans out of the equation.
You hit the nail on the head. The problem isn't AI. The problem is that scaling it up to industrial levels is hard. And the absolute WORST way of solving that problem is making manned mining ships, raising the complexity, cost, and risk by 100x.
Why does it have to? Build them out of cheap, expendable, easily-replaceable parts. Which you can do, because the entire premise of asteroid mining is that it will make the cost of building and deploying space equipment dead cheap (otherwise why do it in the first place?)
Sure, but the problem is that once the coolness factor wears off, they stop doing it. It's not sustainable.
We don't? We already have AI that can autonomously land on other bodies and extract material. In fact we've had it for 4 decades. See: almost any planetary lander/rover ever. It seems the barrier to mining is more up-front cost and on-site materials processing than AI.
I'm a techno-optimist but I agree with you. The rate things are going, it isn't going to make much sense to have people living in space colonies. I can't think of any good reason to do it other than the coolness factor. Unfortunately a lot of people are emotionally invested in this idea and will fight logic tooth and nail to promote their fantasies.
But that's just begging the question of why they didn't use liquid-fuelled boosters instead. As I said, the crazy and conflicting design requirements (for example, the 'need' to launch satellites into polar orbit, the unnecessary focus on re-usability, and the decision to have lower up-front cost at the expense of greater future operating costs) contributed to these flawed decisions.
I've been actively working in the field for the past few years and I don't think he's incredibly off the mark. Google, for instance, has some pretty advanced tech in production and lots more in development. The 'new AI' (statistical machine learning and large-scale, distributed data mining) is getting pretty advanced and scary.
Symbolic manipulation as a route to AI was a period of collective delusion in computer science. Lots of people wasted their talents going down this route. In the 80's this approach was all but dead and AI researchers finally sobered up. They started actually learning about the human brain and incorporating the lessons into their designs. It's sad that so much time was wasted on that approach, but the good news is that the new approaches people are using now are based on actual science and grounded in reality. The intelligence in search, natural language, object and facial recognition, and self-driving cars (that ShanghaiBill pointed out) is due to these new approaches.
AI spent its youth confused and rebellious. That was when you were in your graduate studies. Now it's far more matured. I encourage you to read up on new machine intelligence approaches and the literature in this area. You won't be disappointed.
The problem isn't reuse of old technology. The problem is the selection of old technology you reuse, and how you go about re-using it.
Start with the solid rocket boosters as an example. There's very good reasons why most space launch platforms don't use solid rockets. Cost, efficiency, and inherent lack of safety (you can't turn off an SRB once it's been lit) are just a few. So why did the shuttle use them? Because it was kind of forced upon them by the crazy and contradictory design decisions they had to comply with. The end result was that 70% of the takeoff thrust was actually provided by the two solid boosters, with only 30% coming from the three high-tech hydrogen rockets.
With the SLS, NASA had the opportunity to fix the warts in the shuttle program. Instead what we have now is the maximum-pork option.
If you look at the case, there's evidence she did slight but deliberate manipulation and misrepresentation of results. That's what I meant by wrong.
It's probably not the case that she wrote the paper cackling to herself madly and proclaiming "Those suckers will never find out!"
It's probably the case that through self-delusion and carelessness she managed to partially convince herself that the results were true, and this, coupled with pressure to produce results, caused her to take a few shortcuts to get it published. What she did was wrong, and her career is over. It's not something a rational mind would have done. But scientists are just human and sometimes prone to making irrational decisions. The great thing is that we have the scientific method to weed out the good from the bad.
The clean-up was less due to the severe amount of radioactivity and more due to the fact that he was careless and got it everywhere.
The total amount of radioactive material was small and the actual dose of radiation he was exposed to was probably minimal. Although the exact dose isn't known because he never completely revealed his experiments and he never underwent testing.
One thing I find interesting is that he was arrested again in 2007 on charges related to stealing smoke detectors for their Americium, 13 years after his boy scout experiments.
As far as google is concerned, you're going to be using voice commands anyway.
Yup, that's exactly it. Another thing about those computers is that they are often more 'geek-oriented' overall. I got a vostro 1320 laptop with freedos, for instance, and it has a backside panel for easy access to the fan and heatsink assembly. Cleaning the computer is literally a matter of just removing a pair of screws. Compare to some other laptops where you virtually have to disassemble down to the bare motherboard just to get the fan clean. It was also about $90 cheaper than the windows version, even though the specs were almost exactly the same. It's nice that companies still make computers like this. I wonder when it will end.
As the article says, that's irrelevant because it's a tribe affiliation. When it comes to these things, logic and intelligence are completely overwhelmed and suppressed. That's how our brains work. Ever notice how all so-called "freethought" and "rationalist" groups soon turn into hilariously ironic examples of conformity and groupthink?
At this point, there's actually no need for century-old data (although it does add to the increasing pool of evidence). Weather satellites have been up in space for long enough now to establish warming beyond any reasonable doubt.
I actually do know a few people who've obtained comfortable office jobs without degrees, but it was very hard for them and they had to go through personal connections. YMMV.
About the 'remote classroom', It's been my experience that these are mostly projects to reduce costs for the guys at the top without really changing the system. The real advances have come from initiatives from professors themselves, but their hands are tied. A professor can give you the best education but he can't give you a degree.
Ultimately, it's going to take cultural change. Slowly, people will realize that an online education can be just as good as a brick-and-mortar one, depending. But as I said, the guys at the top will do ANYTHING to prevent this, even if it means devious subversion and spreading FUD (which many people will fall for, hook line and sinker).
> that is, replace your loop conditionals with direct jumps to the next instruction handler
This would accomplish nothing as GCC already highly optimizes loop conditionals with jumps.
> And if you think using LLVM to write an interpreter is easier, than you're just... I don't even know what... totally out of touch and have no clue how this stuff works.
I didn't suggest that the perl6 team use LLVM to write an interpreter. I'm just saying that examples exist of people successfully implementing interpreters in this way (using a high-level frontend and compilation to bytecode for performance) for fairly complex languages, without major development hell like perl6 experienced. The resulting code is often simpler than C (really -- have you ever looked at the source code for CPython?), even considering the added complexity of generating bytecode. Thus your conclusion (that the reason they failed was because they didn't use C) doesn't follow from your premises. It's more likely that they were just incompetent and bad at language design and implementation. Perhaps LLVM is a bad choice for perl6, but again, that boils down to language design.
> The VM isn't the problem. It's all the bells and whistles. The typing and object system.
I couldn't agree more that feature creep is the enemy when it comes to language implementation. Just look at C++.
You're implying that C-based interpreters are quicker and easier to develop, but that's patently not the case. Just look at the recent huge surge of LLVM-based JIT languages that have been developed extremely rapidly and in high-level languages. C is probably a poor choice now that good alternatives really exist.
Maybe the problem isn't the choice of language. Maybe the problem is that the designers were incompetent.
The world is richer than ever, on average --- it's just that all that wealth is now in the hands of a few people.
You still need a college degree to get a white-collar job, and the fact is that most people want white-collar jobs. I 100% agree with you that it's an evil system designed to exploit the vulnerable, but it's not just an empty sales pitch like you're describing. The system is rigged in such a way that a lot of students HAVE to be burdened with massive debt to succeed in life. We have ZERO need for huge buildings with lecture halls in them in the age of the internet, but those people at the top are going to do anything it takes to keep this cash cow flowing.
Obviously, it's flat-out stupid to spend this debt getting a liberal arts degree or anything that doesn't have good job prospects.
It's easy to level hate on administrators, and it's true that there are far more administrative staff than needed on most campuses. But a lot of universities have recently drastically slashed down on administrative staff (including my own), and what happened? Surprise, surprise, tuition fees are still high, professors are still scrambling after grants, and instructors are still poor. Let's not lose sight of the real culprits here.
> you'd have a colony full of presumably quite intelligent, capable, and emotionally stable people (we did screen them for the hardships of colony life, right?)
They're still human and could act unpredictably under pressure. And even if they were angels, there's no guarantee that their children would be the same.
> And Earth is unlikely to have a reason to get Mars involved - unlike orbital and possibly lunar colonies which might be able to play a tactical role in terrestrial battles,
If there's no reason for Mars to get involved then there's no reason to build a Mars colony. The whole point of interplanetary colonization is economic expansion. At some point there is going to be trade between Earth and Mars. And I mean trade of all kinds, not just raw materials. For instance, trade of intellectual property developed by Martian colonists.
Plus, you're going to have a lot of people with cultural ties to the various nations on Earth. What if a war between the USA and China breaks out - how will American-descended colonists get along with Chinese-descended colonists?
But all of this is ignoring the elephant in the room - human politics is based on brinksmanship, and always has been. It's our nature. And yes, it even exists among 'smart' groups (in fact, it's more pronounced in 'smart' groups). We survived in spite of this because of sheer numbers and resilience. But things won't be the same on Mars. You can't have brinksmanship in a fragile colony.