To some extent, there is a tension between "unbiased reporting" and using good journalistic judgment.
For one example you cite -- "lipstick on a pig" -- the "unbiased" thing to do would be to say, "well, the McCain campaign is complaining that it's sexist, so we should probably report their assertion.
The journalistic common sense approach would be to say, "Dude, you're making that into something it's not. What do you have that's worth my time to report?" Instead, we got a couple of days of empty talking heads speculating on whether it was sexist, then a couple more days of jabbering about all the other times politicians (McCain included) had used the same simile, followed closely by a couple of days speculating on whether the story had backfired on McCain.
Had reporters treated it as the non-story that it truly was, that would have given us two or three precious news cycles devoted to... well, anything would have been more relevant than that.
Now, I'm an Obama supporter. But I think the whole Lipstickgate thing actually favored Obama somewhat, making McCain look a bit vicious and desperate. Still, it was a poor use of our collective attention span, and shouldn't have been reported. I'm still a bit peeved at McCain for making the accusation, but it only had an effect because our media is split between "unbiased" stenographers and talking heads who prefer simplistic controversy to substance.
1) If it goes bankrupt quickly, why do you fear it being emulated frequently enough to distort the economy?
2) In existing companies, people are already there "as long as they wish", so that's not an objection. As for "doing whatever they wish", it's only a problem if nobody is making decisions about which activities merit compensation. They're not talking about a company where everyone can join up and can get paid to sweep mines all day.
3) You seem to be under the assumption that the entrepreneur has perfect knowledge into how to satisfy the consumer, and his peons know only how to implement that glorious vision. I would argue that this knowledge is distributed throughout the whole organization. So letting everyone get in on rating the value of contributions may turn that knowledge into action better than letting one manager decide by fiat.
4) Keeping employees happy is critical to an organization's success. This setup seems like an ideal way to discover which tasks employees find least pleasant. Once discovered, you just have to raise the reward for performing those tasks until the tasks start getting done. Far from rejecting the idea of productivity, this seems like the ideal way to raise productivity and morale at the same time.
I don't see it as hopeless, though. The question isn't whether the simulation captures every detail, but whether it captures the relevant details.
I was always under the impression that the electrochemical signals were the most important means of communication. After all, a puff of neurotransmitter is going to have area effect, rather than targeting a single mob^H^H^H neuron. It's also going to be slower to take effect. So beyond affecting the neuron's overall mood, I'm not sure how the neurotransmitters affect the computations going on in the brain.
So it's possible that a rough approximation of neurotransmitter effects could be done simply by reprogramming neurons from the outside. You know, set neuron 237643 to a lower excitation energy, set 123436 to learn faster. I'm not sure how many other settings there are to fiddle with, but we might be able to start out just by changing them in a predictable fashion. Maybe the result will be the same, maybe the differences will manifest through distinct symptoms, like an inability to learn from certain stimuli.
Even if these particular chips show themselves to be too limited, I think we'll learn a lot from comparing their activity to a real-world counterpart.
I don't think the distinction between conscious and unconscious thought is clear enough that a number like 10% could be meaningful. Nor do I think you'd gain anything by bringing unconscious thought under conscious control.
Neurons have memory. The memory comes from the strength of the interconnections, which strengthen and weaken over time.
So these chips aren't going to simply embody the same set of logical operations every time. They're going to learn from past inputs, which is precisely why neurons are worth simulating in the first place.
If you're simulating an actual brain, you're going to need to simulate a world for it to interact with, and give it a body. Otherwise, your mega-Einstein just goes crazy from sensory deprivation.
Some algorithms will necessarily be easier to game than others, and not everyone will be trying to game the system. Further, some attempts at gamesmanship will be relatively easy to detect, and should be punished.
I think the best way to discourage that sort of behavior is for participants to simply recognize that the company will be less successful if they try to game the system. If you try to take money away from the tech writers to pad the developers' pockets, and that leads to poor documentation, it hurts the business.
Of course, that's only important to those in it for the long haul. So only regular contributors should be in the business of affecting compensation.
This will probably lead to less compensation for new members and one-time contributors. But long-timers should recognize that attracting talent -- especially talent that can make good contributions -- should be a priority.
I don't see that this setup is significantly more doomed than your normal startup.
You're absolutely right. Atlas Shrugged does an outstanding job describing Ayn Rand's political philosophy in a clear and memorable way.
Unfortunately, the philosophy itself is crap, which is part of the reason the book is so lousy. A more sane philosophy could have been described using more believable characters.
While I don't think it has as much potential as some libertarian proponents say -- I don't think it could build a $30M bridge, for example -- the assurance contract might be a useful model for some software development.
In a normal startup, the "entrepreneur" is usually the one providing the bulk of the capital. Given that it takes almost no capital to start a software company, I don't see how there is a need for a split between the entrepreneur and her employees.
Once you've got a group of people who decide to start a company -- on a dare, after an all-night reefer party, whatever -- they act like any other company, in that they try to sell their services at a price that will sustain the ongoing operations of the company and pay each employee enough to make it worth doing.
Your comment seems irrelevant, because the proposed company still has to pay people enough to ensure their continued participation, and therefore income has to be sufficient to pay every employee's salary. In other words, the essentials of running a business are still in effect, and what is being proposed is nowhere near as radical as you're claiming.
it would also make profits/losses extinct, thus making it impossible to measure the extent to which consumers are satisfied
You'll need to explain why this statement is correct, because I'm seeing the opposite.
If the guy doing the excellent work isn't getting the compensation he would like under this regimen, he has three choices:
1) Leave.
2) Demand that his contributions be recognized and compensated.
3) Continue doing what he's doing, accepting the imperfections of the system, and maybe even enjoying the fact that the popular guy is buying him the occasional beer.
Come to think of it, those are basically his options in a regular job. Why would this be any less sensible than the metric most companies use to determine salaries: industry average at time of hire?
The lion's share of the contributions wouldn't necessarily go to the largest group. The total going to a given group like accounting, janitorial, developers, etc., might be set by fiat from above, or by::shudder:: committee. But however it gets set, they'd have to compensate for the fact that, in general, people tend to overvalue their own contribution to the whole.
I doubt that would be a practical problem. If there are that many contributors, simply paying them all would be a logistical nightmare, as would coordinating all the contributions. But other things being equal, it would make sense to value 50 contributions by 1 guy above 50 equally valuable contributions by fifty separate individuals. After all, it's far more likely that you'll get a 51st contribution from the one guy than from the fifty one-hit-wonders.
So I think it would make sense for compensation not to kick in on the first contribution.
But if there is enough work for 500 people to do (as opposed to mostly duplicated effort, or 500 mostly one-time contributors who spend only a few days on their contribution), and that many people looking to do it, there's probably some value to the project, and some people are going to at least try to become company salesmen.
If there is unglamorous work that everyone recognizes as needing to be done, you might implement some sort of bounty system. "Update user documentation for release 2.4.2" might be more attractive if workers knew that there was already a guaranteed $50 waiting on completion.
I don't think "getting contributions noticed" would be a big deal. It would be easy enough to have an ongoing list of tasks, due dates, who's working on what, etc., and all contributions would have to be evaluated before being committed. I'm imagining a system where each chunk had some payment associated with it, some percentage bonus for doing it well, and a second bonus that depended on the popularity of the person doing the work.
(pardon my alleged hubris, but my time is worth more than that)
For 98% of the projects out there, this is probably true. For the biggest, most famous projects, this is probably false. But the more relevant fact is, even if you fixed a bug in Mozilla that was worth, say, $100,000 (spread out over tens of millions of end users), there's no way you're going to be able to reclaim even 1% of the value you generate.
So the bugs are fixed by those who enjoy fixing them, not necessarily by those whose time is less valuable than your own. A paid bounty system seems sensible, if not enough bugs are getting fixed. But where does the money come from?
because your uptopian scheme, like the millions that have come before it and the millions that will come after, they DO serve a valid purpose in society: it keeps people like you occupied in mediocrity and obscurity, and away from real businesses that works and real society that works, where you might do real damage
Apparently, Slashdot serves the same purpose for you.
The problem is, under such a system, the work doesn't get done according to "who is least needed on a more valuable task" but "who is most bothered by the mess?" If your best programmer is also the biggest clean freak, it would be a glaring inefficiency.
On the other hand, he'd probably also be the most likely to vote up any commode-cleaning by another person. Hopefully his vote would carry enough weight that the task would get done.
"Of questionable benefit" to whom, exactly? Not to the longer-living individual. Perhaps to those who would come after, but I don't give much thought to the unconceived as individuals.
But it sounds like you're worried about the health of the species. I think your concern is misplaced. Biological evolution is a slow process, especially when compared to the technological adaptation that has characterized our species for the last ten thousand years. I would assume that within the next hundred years,* we will essentially have control over our genetic makeup, making the old evolutionary paradigm moot.
While in most ways I'm in favor of less artificial, industrialized lifestyles, I'm not a neo-Luddite, and extending a healthy life seems like a laudable application of technology.
* Ray Kurzweil would probably say twenty-five, and I'd hesitate to contradict him.
The only solution I can see is to force perfect transparency on the people entrusted with vital data. Also, users ought to have at least some control over what data gets collected, and whether (and to whom) it gets transmitted. The whole "let's never collect any data because somebody might misuse it" paradagm seems like a dead end to me.
I'm guessing that hooking someone up with one of these will be closer to $1000, if we're talking about within the next ten years. If we're talking twenty years on, perhaps $100, or $1000 for a much more elaborate sensing network. Nor do I think that the sensors will require invasive surgery. A lot could probably be learned by sensors put right under the skin.
As for your "we can't do anything about whatever problems we find" argument, it's simply wrong. We know what to do in the event of clogging arteries. We know what to do in the event that we discover early-stage cancer. There are a handful of diseases that we know how to detect, but have no way to treat, but the ones that are the most frequent (heart disease, cancer, diabetes) are ones that we know how to prevent and manage. Early detection of these issues would pay for the equipment, maintenance, and monitoring probably a dozen times over.
Sure, if you step back far enough, everyone is replaceable. But step back a bit further, and our entire species, our entire planet, our entire universe might be replaceable. I feel irreplaceable to myself, and to the few people willing to put up with me. So I don't see the argument as a compelling one.
In closing, do you want to be Bluetooth-enabled or not? C'mon man, everything is better with Bluetooth.
Slashdot Mirror
To some extent, there is a tension between "unbiased reporting" and using good journalistic judgment.
For one example you cite -- "lipstick on a pig" -- the "unbiased" thing to do would be to say, "well, the McCain campaign is complaining that it's sexist, so we should probably report their assertion.
The journalistic common sense approach would be to say, "Dude, you're making that into something it's not. What do you have that's worth my time to report?" Instead, we got a couple of days of empty talking heads speculating on whether it was sexist, then a couple more days of jabbering about all the other times politicians (McCain included) had used the same simile, followed closely by a couple of days speculating on whether the story had backfired on McCain.
Had reporters treated it as the non-story that it truly was, that would have given us two or three precious news cycles devoted to... well, anything would have been more relevant than that.
Now, I'm an Obama supporter. But I think the whole Lipstickgate thing actually favored Obama somewhat, making McCain look a bit vicious and desperate. Still, it was a poor use of our collective attention span, and shouldn't have been reported. I'm still a bit peeved at McCain for making the accusation, but it only had an effect because our media is split between "unbiased" stenographers and talking heads who prefer simplistic controversy to substance.
1) If it goes bankrupt quickly, why do you fear it being emulated frequently enough to distort the economy?
2) In existing companies, people are already there "as long as they wish", so that's not an objection. As for "doing whatever they wish", it's only a problem if nobody is making decisions about which activities merit compensation. They're not talking about a company where everyone can join up and can get paid to sweep mines all day.
3) You seem to be under the assumption that the entrepreneur has perfect knowledge into how to satisfy the consumer, and his peons know only how to implement that glorious vision. I would argue that this knowledge is distributed throughout the whole organization. So letting everyone get in on rating the value of contributions may turn that knowledge into action better than letting one manager decide by fiat.
4) Keeping employees happy is critical to an organization's success. This setup seems like an ideal way to discover which tasks employees find least pleasant. Once discovered, you just have to raise the reward for performing those tasks until the tasks start getting done. Far from rejecting the idea of productivity, this seems like the ideal way to raise productivity and morale at the same time.
I don't see it as hopeless, though. The question isn't whether the simulation captures every detail, but whether it captures the relevant details.
I was always under the impression that the electrochemical signals were the most important means of communication. After all, a puff of neurotransmitter is going to have area effect, rather than targeting a single mob^H^H^H neuron. It's also going to be slower to take effect. So beyond affecting the neuron's overall mood, I'm not sure how the neurotransmitters affect the computations going on in the brain.
So it's possible that a rough approximation of neurotransmitter effects could be done simply by reprogramming neurons from the outside. You know, set neuron 237643 to a lower excitation energy, set 123436 to learn faster. I'm not sure how many other settings there are to fiddle with, but we might be able to start out just by changing them in a predictable fashion. Maybe the result will be the same, maybe the differences will manifest through distinct symptoms, like an inability to learn from certain stimuli.
Even if these particular chips show themselves to be too limited, I think we'll learn a lot from comparing their activity to a real-world counterpart.
I don't think the distinction between conscious and unconscious thought is clear enough that a number like 10% could be meaningful. Nor do I think you'd gain anything by bringing unconscious thought under conscious control.
Still, here's the link you wanted: link.
Neurons have memory. The memory comes from the strength of the interconnections, which strengthen and weaken over time.
So these chips aren't going to simply embody the same set of logical operations every time. They're going to learn from past inputs, which is precisely why neurons are worth simulating in the first place.
I think the book was called Wyrm . The quote was from the main character's mentor.
If you're simulating an actual brain, you're going to need to simulate a world for it to interact with, and give it a body. Otherwise, your mega-Einstein just goes crazy from sensory deprivation.
How did I not see that one coming?
I'm not a proponent of spanking. Your beef is with the guy I responded to.
Thank you for repeating yourself. Now explain how this business model hampers the "profit/loss" model. Thus far, you haven't.
The definitive analysis.
I've heard that The Turner Diaries is also getting a substantial bump in readership. Keep that in mind when drawing your conclusions.
Some algorithms will necessarily be easier to game than others, and not everyone will be trying to game the system. Further, some attempts at gamesmanship will be relatively easy to detect, and should be punished.
I think the best way to discourage that sort of behavior is for participants to simply recognize that the company will be less successful if they try to game the system. If you try to take money away from the tech writers to pad the developers' pockets, and that leads to poor documentation, it hurts the business.
Of course, that's only important to those in it for the long haul. So only regular contributors should be in the business of affecting compensation.
This will probably lead to less compensation for new members and one-time contributors. But long-timers should recognize that attracting talent -- especially talent that can make good contributions -- should be a priority.
I don't see that this setup is significantly more doomed than your normal startup.
You're absolutely right. Atlas Shrugged does an outstanding job describing Ayn Rand's political philosophy in a clear and memorable way.
Unfortunately, the philosophy itself is crap, which is part of the reason the book is so lousy. A more sane philosophy could have been described using more believable characters.
While I don't think it has as much potential as some libertarian proponents say -- I don't think it could build a $30M bridge, for example -- the assurance contract might be a useful model for some software development.
In a normal startup, the "entrepreneur" is usually the one providing the bulk of the capital. Given that it takes almost no capital to start a software company, I don't see how there is a need for a split between the entrepreneur and her employees.
Once you've got a group of people who decide to start a company -- on a dare, after an all-night reefer party, whatever -- they act like any other company, in that they try to sell their services at a price that will sustain the ongoing operations of the company and pay each employee enough to make it worth doing.
Your comment seems irrelevant, because the proposed company still has to pay people enough to ensure their continued participation, and therefore income has to be sufficient to pay every employee's salary. In other words, the essentials of running a business are still in effect, and what is being proposed is nowhere near as radical as you're claiming.
You'll need to explain why this statement is correct, because I'm seeing the opposite.
If the guy doing the excellent work isn't getting the compensation he would like under this regimen, he has three choices:
1) Leave.
2) Demand that his contributions be recognized and compensated.
3) Continue doing what he's doing, accepting the imperfections of the system, and maybe even enjoying the fact that the popular guy is buying him the occasional beer.
Come to think of it, those are basically his options in a regular job. Why would this be any less sensible than the metric most companies use to determine salaries: industry average at time of hire?
The lion's share of the contributions wouldn't necessarily go to the largest group. The total going to a given group like accounting, janitorial, developers, etc., might be set by fiat from above, or by ::shudder:: committee. But however it gets set, they'd have to compensate for the fact that, in general, people tend to overvalue their own contribution to the whole.
I doubt that would be a practical problem. If there are that many contributors, simply paying them all would be a logistical nightmare, as would coordinating all the contributions. But other things being equal, it would make sense to value 50 contributions by 1 guy above 50 equally valuable contributions by fifty separate individuals. After all, it's far more likely that you'll get a 51st contribution from the one guy than from the fifty one-hit-wonders.
So I think it would make sense for compensation not to kick in on the first contribution.
But if there is enough work for 500 people to do (as opposed to mostly duplicated effort, or 500 mostly one-time contributors who spend only a few days on their contribution), and that many people looking to do it, there's probably some value to the project, and some people are going to at least try to become company salesmen.
If there is unglamorous work that everyone recognizes as needing to be done, you might implement some sort of bounty system. "Update user documentation for release 2.4.2" might be more attractive if workers knew that there was already a guaranteed $50 waiting on completion.
I don't think "getting contributions noticed" would be a big deal. It would be easy enough to have an ongoing list of tasks, due dates, who's working on what, etc., and all contributions would have to be evaluated before being committed. I'm imagining a system where each chunk had some payment associated with it, some percentage bonus for doing it well, and a second bonus that depended on the popularity of the person doing the work.
For 98% of the projects out there, this is probably true. For the biggest, most famous projects, this is probably false. But the more relevant fact is, even if you fixed a bug in Mozilla that was worth, say, $100,000 (spread out over tens of millions of end users), there's no way you're going to be able to reclaim even 1% of the value you generate.
So the bugs are fixed by those who enjoy fixing them, not necessarily by those whose time is less valuable than your own. A paid bounty system seems sensible, if not enough bugs are getting fixed. But where does the money come from?
Apparently, Slashdot serves the same purpose for you.
Ass.
The problem is, under such a system, the work doesn't get done according to "who is least needed on a more valuable task" but "who is most bothered by the mess?" If your best programmer is also the biggest clean freak, it would be a glaring inefficiency.
On the other hand, he'd probably also be the most likely to vote up any commode-cleaning by another person. Hopefully his vote would carry enough weight that the task would get done.
"Of questionable benefit" to whom, exactly? Not to the longer-living individual. Perhaps to those who would come after, but I don't give much thought to the unconceived as individuals.
But it sounds like you're worried about the health of the species. I think your concern is misplaced. Biological evolution is a slow process, especially when compared to the technological adaptation that has characterized our species for the last ten thousand years. I would assume that within the next hundred years,* we will essentially have control over our genetic makeup, making the old evolutionary paradigm moot.
While in most ways I'm in favor of less artificial, industrialized lifestyles, I'm not a neo-Luddite, and extending a healthy life seems like a laudable application of technology.
* Ray Kurzweil would probably say twenty-five, and I'd hesitate to contradict him.
First bit of advice: when ranting about how the government is trying to control you, don't ever bring up aliens. It's a bad idea.
Two words for you: Penis cam.
The only solution I can see is to force perfect transparency on the people entrusted with vital data. Also, users ought to have at least some control over what data gets collected, and whether (and to whom) it gets transmitted. The whole "let's never collect any data because somebody might misuse it" paradagm seems like a dead end to me.
I really doubt it will cost $1M per person.
I'm guessing that hooking someone up with one of these will be closer to $1000, if we're talking about within the next ten years. If we're talking twenty years on, perhaps $100, or $1000 for a much more elaborate sensing network. Nor do I think that the sensors will require invasive surgery. A lot could probably be learned by sensors put right under the skin.
As for your "we can't do anything about whatever problems we find" argument, it's simply wrong. We know what to do in the event of clogging arteries. We know what to do in the event that we discover early-stage cancer. There are a handful of diseases that we know how to detect, but have no way to treat, but the ones that are the most frequent (heart disease, cancer, diabetes) are ones that we know how to prevent and manage. Early detection of these issues would pay for the equipment, maintenance, and monitoring probably a dozen times over.
Sure, if you step back far enough, everyone is replaceable. But step back a bit further, and our entire species, our entire planet, our entire universe might be replaceable. I feel irreplaceable to myself, and to the few people willing to put up with me. So I don't see the argument as a compelling one.
In closing, do you want to be Bluetooth-enabled or not? C'mon man, everything is better with Bluetooth.