Are we seeing such a change, is the headline just lazy?
Sadly, I think we are seeing such a change. Which I find really, really annoying, since crypto(graphy) has been my livelihood for 20 years and it'll probably take me a decade to redirect my mental pointer, but the word "crypto" has gone mainstream, and in the mainstream it means "cryptocurrency".
I don't want my dead fingers to be more useful to the cops than my living fingers. That's a bad-mojo sort of incentive brewing right there.
They can make you put your living finger on the scanner, so your dead fingers would be equally useful, not more useful.
As for theories that they might shoot you because they don't have probable cause to arrest you, it's much easier to manufacture probable cause than it is to justify shooting. Even if their probable cause gets tossed later, they can still use whatever evidence they got from the device, just not against you. And of course if they kill you they can't use the evidence against you, because you're dead. So there's really no advantage in killing you.
That is a debate worth having. One concern I have with the notion of not limiting liability is that it would tend to further concentrate investment in the largest corporations. Investing in Apple, for example, would be quite safe because the possibility that they could generate liabilities in excess of their assets is very low, because their assets are so large. Small and medium-sized corporations, on the other hand would be a much riskier proposition, compared to the status quo where you always know that your risk is limited to the amount you invest. It would be worse even than the risks incurred by short positions, because you can almost always use a stop limit order to limit your total risk with those.
This means that small and medium-sized corporations would have to become much more risk-averse themselves, so they could demonstrate that risk-aversion to potential investors. That has some potential benefits, I suppose, but they seem very weak from a macroeconomic perspective. I don't see that creditors are being significantly harmed by bad debt from small corporations. OTOH, it could be very bad for economic growth, because the ability and willingness of small companies to take risk is a significant enabler of growth. If that weren't the case, we should expect to see all of the growth coming from the megacorps, given their economies of scale, easy capital and synergy potentials. But we don't, and I think risk-aversion is one of the primary reasons.
On balance, I suspect that the societal cost/benefit analysis comes down on the side of liability limitation.
There's also the possibility of an appliance, like a NAS, with this functionality built in.
Economies of scale and network effects will drive centralization and push those sorts of solutions out.
Many people like to argue that there's a pendulum in computing, swinging between centralization and decentralization, but I think they're wrong. The natural state of computing resources is centralized. We saw one anomalous swing away from centralization, when the cost of computing dropped dramatically faster than the cost of communication. As soon as communication costs dropped far enough, though, we moved quickly back towards centralization of computing infrastructure, now called "the cloud" rather than "the mainframe". Sure, our pocket and laptop computers are a million times more powerful than our desk-filling PCs were, but we use them mostly as portals to cloud systems which are a trillion times more powerful. Of course, there is still value in having some powerful nodes at the edges -- my desktop machine has 56 cores, 128 GiB RAM and 3 TB of NVMe -- but that is and will be the exception that proves the rule.
Mostly I believe that you don't have many or perhaps any personal contacts in the google these days.
You're certainly welcome to believe what you like. Clearly you've decided to believe that I'm brainwashed or otherwise fooled (or simply lying about my beliefs -- though I do appreciate that you've opted to apply Hanlon's Razor) and I'm not going to change your mind on that point. Which means I'm unlikely to change your mind on any other point, so I'm wasting my time here. But it's a lazy Sunday afternoon and I'm doing this to procrastinate finishing my taxes, so...
The real issue is the abuse of our personal information, and it doesn't really matter if Facebook or the google are doing it directly, or they are playing shell game with the abuse by selling or loaning or giving the data to some other entity for the abusive uses. Actually, in terms of protecting their reputations, I think corporate cancers often prefer such tactics.
A couple of points:
First, your use of the phrase "corporate cancer" indicates a deep and systematic bias that can't help but affect your thinking in all sorts of ways. Corporations have their down sides, clearly, but it's also the case that without them we would have much less of basically everything that enables us to live longer, healthier, happier lives. Doing things on a large scale requires organization and pooling of resources, and of all the forms of organization and resource pooling we have, only corporations and governments can operate at sufficient scale. Governments are good at some things, but very bad at many others, in particular they're very bad at accurately responding to the daily needs of the people (c.f. every attempt at central planning, ever).
It would really help you to get a better perspective on what is good and bad about corporations. They are not an unalloyed good, but they are a net positive. A large net positive.
Second, it does matter whether Facebook or Google are selling data or selling services around the data. Most obviously, it matters because if the data is sold then it will be re-sold, and re-sold, and re-sold, which means there is absolutely no way to control its use and it will eventually be put to every nefarious use possible. As long as it stays locked up in one place, there is the potential to implement controls on it, whether that's through government regulation or through soft power, used to make the owners of the data provide controls in order to avoid the threat of regulation. Or, in the case of Google, through significant internal pressure to do the right thing -- which is the source of the Takeout, deletion and opt-out controls that Google offers.
Of course, there are also significant societal risks in having so much data in one place. One risk is that if Google or Facebook suffered a serious data breach, we're back in the "out of control" situation. Another is that such large piles of data are an attractive target for the other huge concentration of social power, government, and government is also prone to abuse.
Monetize can include renting access to the data; renting analytic access to the data without allowing actual access to the underlying data
I don't see the harm, assuming it's aggregated and anonymized. If the analytic system provides data about individuals, then it can be extracted and we're back into the out-of-control situation.
plus plenty of other nefarious activities
Like?
If your starting point is that this data is a monetisable asset, you're going to end up doing bad things with it.
Also, you're moving the goalposts here. Your original claim was that there was no difference between monetizing and selling. I pointed out there's a large difference. Now you're essentially saying that monetizing is also bad. Which is fine, you can try to make that argument (though it does reek of a slippery slope fallacy), but it's a different argument than you started with.
If step one of your plan to replace Facebook is everyone running family servers, your plan is doomed from the start. Most families don't have anyone capable of doing that, and hardly any families have anyone capable of doing it well -- keeping the machine running, updated, and properly secured.
There's a remote chance that it could work if there were a competitive network of service providers who ran the servers. For example, if ISPs did it, the way they all used to run email servers. It might also be somewhat possible if cloud providers operated and maintained the servers. In both cases, though, I think it would just lead the cloud providers to exploit economies of scale by putting up one big infrastructure for all of their users, and to compete by offering features that others don't have... then network effects would kick in and one of them would become dominant and you'd just have a new Facebook.
I think the bottom line is that widely-used services that are subject to intense network effects are natural monopolies. And natural monopolies require regulation.
You're not thinking it through. If we were paying the google to protect our privacy, then the google would have an actual incentive to do it, but the money is coming from corporate cancers that want to have it all. The only concern of the corporate cancers is creating the image that our privacy is being protected.
No, you're not thinking it through. Were Google to sell the data to "the corporate cancers", the latter would have it and would not need to buy it again; Google could only sell new data as they receive it. If Google keeps the data, though, they can monetize its use indefinitely. Plus, the fact is that the data plus Google's analytical abilities are actually of greater value to Google's customers than they could extract from the data themselves. They mostly know this.
In addition, people at Google really aren't assholes (well, some are, like in any population, but fewer than in most) and profit not only isn't their only motive, it's not even the primary motive. Not under the current leadership, anyway. Also, I can tell you that if Google were to start selling data, you wouldn't have to guess or extrapolate. Lots of Google employees would be shouting it from the rooftops. And quitting. Probably in that order.
Monetize vs sell: a distinction without a difference here.
It's a huge difference, assuming the service doesn't provide APIs that allow what Cambridge Analytica did. As long as all of the data is locked up in one place there is the possibility of control, and the likelihood that it will be used only in non-harmful ways. Once it leaks, absolutely anything can happen.
It feels different when it's topsy-turvey, with you turvey instead of topsy, doesn't it?
It feels exactly the same level of ludicrous both ways to me. Actually, it doesn't "feel" any way; I try to think with my brain.
If the purpose of your post was simply to illustrate that profiling is stupid, then we agree. But that's not how it came across. It came across as a serious proposal and you are a sufficiently assholish and irrational leftist that I found that completely believable.
So they got some heavy duty PR, invented a word "jaywalking", pushed this into the press, and lobbied intensely that it was all really the fault of the pedestrians.
Not really, at least with respect to the origin of the term. And it seems it was more automobile clubs than the auto industry that pushed it. https://en.wikipedia.org/wiki/...
The bomber was white, Christian, home-schooled, anti-LGBT and conservative. This fits the profile of almost all domestic terrorists in the US.
That profile also fits millions of non-terrorists. With a sufficiently loose definition of anti-LGBT, it fits a large minority of people where I live. If you remove "home-schooled" as well (and with the aforementioned loose definition) it fits almost everyone.
Your statement is akin to saying that the FBI should be watching all Arab Muslims who go to Mosque daily.
Profiling is questionable in any case, but it's just stupid when the 99.999% of the people who fit the profile aren't a problem.
I think you've proved my point rather succinctly actually.
No, he refuted your point quite succinctly. I'm perfectly happy to call a white guy a terrorist. Timothy McVeigh and Dylan Roof come immediately to mind. But the fact is that terrorism is politically-motivated by definition, and in the absence of a political motive Conditt's actions weren't terrorism. As far as we know, he was just a serial killer.
And you're holding a weapon (a gun). So does that mean that when you draw yours to shoot the rock-thrower, I'm justified in drawing mine to shoot you? (And, note that while mine is an itty bitty.380, or sometimes a 9mm, not a.357 Magnum, it'll do the job.)
Intent matters here, a lot. Someone throwing a rock at an unoccupied car is not threatening anyone with serious injury or death. Someone throwing the same rock at a person probably is. In no state in the US is deadly force legally justified for preventing vandalism. In only one state (Texas) is it justified even for preventing theft, and then only if there's no other option, the property would not be recoverable and if it's nighttime (and that Texas law is a travesty and should be repealed).
Dude, people like you give responsible gun owners a bad name. Please stop it.
Oh, and by the way, I should point out that in the situation described I absolutely would be legally justified in putting a bullet into your head to stop you from shooting the rock thrower. Deadly force is justified in every state in the union if it's necessary to prevent the unlawful use of deadly force against another. Think about that. Hard.
And, no, it wouldn't cover this case, unless it was clear that the intent of the rock-throwing was to steal the vehicle. The law allows use of deadly force when necessary to prevent "arson, burglary, robbery, aggravated robbery, theft". Note that this only applies during the nighttime, and there are some other caveats as well.
I know just enough to know that one should use the right caliber for the right job and that the tactical advantage of having a gun is lost as soon as someone is aware that you do in fact have a gun.
Right, because the only way to utilize a gun is as a deterrent; it's not as if they shoot real bullets or anything.
Actually, the value of a gun as a deterrent is dependent upon the potential aggressor knowing that you have it. If they don't know about it, it can't be a deterrent, right? (Unless enough people carry concealed that it becomes a general deterrent).
The flip side is that if a potential aggressor knows about your gun, and is determined to attack anyway, then they can try to plan to somehow neutralize you or your gun before you can bring it into play. If the aggressor has a gun, their obvious solution is to shoot you first, assuming they're willing to take the big step up in legal risk of actually using deadly force. If they're an active shooter they'll have no problem with that. If they're a mugger... probably not.
This is a commonly-cited reason for preferring open carry over concealed carry. The fact is that most potential assailants really don't want to tangle with someone with a gun and if they see that their potential target is armed, they'll choose a different target. An obvious exception is those whose goal is to steal the gun, but that's an easily-countered problem.
So, open carry is a strategic advantage, but possibly a tactical disadvantage. I say "possibly" because the lawful carrier is almost always in a reactive mode. The aggressor initiates, and if that means presenting a gun or a knife someone who is carrying concealed probably finds themselves in a position where they can't draw without being shot or stabbed unless they can first create a distraction, or unless they can draw as part of an action expected by the assailant. For example, if a mugger asks for your wallet and you can retrieve a pocket gun instead. A wallet can also make a good distraction. Drop it or toss it and take a step back.
(Though, honestly, if the mugger will be happy with the wallet give them the damned wallet. Whatever is in it is not worth a life and even if you're the sort of asshole who doesn't care about that, whatever is in it is less than what it will cost you in time, inconvenience and probably cash to deal with the fallout of having shot and possibly killed someone.)
They absolutely can, if that is their job as a professional test driver, and they have someone else monitoring the systems.
I don't think this is true. Being a "professional" doesn't change human nature, especially if the human is being asked to remain attentive but inactive for long stretches of time. But there are solutions.
A common practice early in Google's testing was to have three people in the car. A safety driver behind the wheel, an engineer in back to monitor the system and a random Google employee in the front passenger seat who was getting a "fun ride", to experience the coolness of a self-driving car. But... that random employee wasn't just cargo. The engineers realized that the safety drivede a tendency to get complacent, even with adequate breaks. So the "extra" employee provided an additional set of eyes for whom the experience was novel and interesting and who could be expected not only to remain focused on the driving environment but also to stimulate the attention of the safety driver, who acted as "tour guide", pointing out the things the vehicle was reacting to.
I got such a "fun ride" a few years ago. It was rather impressive. What convinced me of the promise of the self-driving tech was when the car noticed and reacted to a cyclist that no human could have seen. I had already realized, intellectually, that the perfect attentiveness of a computer plus the superhuman abilities of LIDAR + RADAR promised dramatically safer "drivers", but seeing it in action drove the point home.
It totally is. Others in the industry (e.g. Google/Waymo) have done it. They use a combination of starting first in safer conditions (closed environments; safer, slower roads; multiple, attentive safety drivers on short shifts) and massive use of simulation. The thing about self-driving systems is that you can take a real data feed and replay it as many times as you want, and you can also alter it or generate fake data to provide a realistic situation that is more challenging.
There's a reason that Google has been working on this problem for years. You have take a methodical, incremental approach to be safe. Uber's trying to take shortcuts, "move fast and break things", and that's not the right approach in a situation where you're operating in the middle of population centers.
Sorry, guys, we're not going to the moon. Gotta get these rocket thingies 100% safe first in controlled conditions.
You'll note that they didn't test or launch the rockets from downtown Tempe. NASA was always very careful to ensure that the only people at risk were those who volunteered for it.
H1Bs are currently handed out by a lottery. Instead, hand them out based on the salary of the employee, highest-paid first.
Really need an H1B employee? Gonna have to pay more.
This will heavily bias H1B towards the east and west coast.
Lottery unfortunately is the most fair way to do it.
You could apply regional adjustments. Relative cost of living is well-known and there are standard adjustment tables that big corporations use when they move employees around. Just apply those to normalize the salaries, then sort.
Solution:
Government or public university work. Job security and 4-5 wks of vaca are features, not bugs.
It should also be pointed out that "tech" is a very broad swath of job categories. Tech support, operations support and software development are all very different kinds of jobs, and very different job markets.
Indeed but the question here is did the autonomous vehicle even drive as safely as a careful and considerate driver? It wouldn't surprise me if uber cars were not as well programmed as google cars, a reflection of the companies programming those vehicles.
I agree. I was responding to the OP's hypothetical in which it was found that the vehicle was not at fault, meaning it did at least as well as a human could have done, not claiming that was what happened. And I definitely agree that Google has proceeded cautiously, methodically and responsibly, while it wouldn't surprise me at all if Uber has cut a lot of corners.
then we've learned that self-driving does not equal infallible
What will we learn next? That water is wet?
No one with a brain has ever claimed that self-driving cars are or ever could be infallible. Nor do they need to be. They only need to be better than the average human, which is a low bar. And, then, over time they'll get better and better as the algorithms and sensors are refined. In a few decades the NTSB will be combing through car wrecks with something akin to the same scrutiny they apply now to plane crashes because the wrecks will be so rare.
Slashdot Mirror
Are we seeing such a change, is the headline just lazy?
Sadly, I think we are seeing such a change. Which I find really, really annoying, since crypto(graphy) has been my livelihood for 20 years and it'll probably take me a decade to redirect my mental pointer, but the word "crypto" has gone mainstream, and in the mainstream it means "cryptocurrency".
I don't want my dead fingers to be more useful to the cops than my living fingers. That's a bad-mojo sort of incentive brewing right there.
They can make you put your living finger on the scanner, so your dead fingers would be equally useful, not more useful.
As for theories that they might shoot you because they don't have probable cause to arrest you, it's much easier to manufacture probable cause than it is to justify shooting. Even if their probable cause gets tossed later, they can still use whatever evidence they got from the device, just not against you. And of course if they kill you they can't use the evidence against you, because you're dead. So there's really no advantage in killing you.
That is a debate worth having. One concern I have with the notion of not limiting liability is that it would tend to further concentrate investment in the largest corporations. Investing in Apple, for example, would be quite safe because the possibility that they could generate liabilities in excess of their assets is very low, because their assets are so large. Small and medium-sized corporations, on the other hand would be a much riskier proposition, compared to the status quo where you always know that your risk is limited to the amount you invest. It would be worse even than the risks incurred by short positions, because you can almost always use a stop limit order to limit your total risk with those.
This means that small and medium-sized corporations would have to become much more risk-averse themselves, so they could demonstrate that risk-aversion to potential investors. That has some potential benefits, I suppose, but they seem very weak from a macroeconomic perspective. I don't see that creditors are being significantly harmed by bad debt from small corporations. OTOH, it could be very bad for economic growth, because the ability and willingness of small companies to take risk is a significant enabler of growth. If that weren't the case, we should expect to see all of the growth coming from the megacorps, given their economies of scale, easy capital and synergy potentials. But we don't, and I think risk-aversion is one of the primary reasons.
On balance, I suspect that the societal cost/benefit analysis comes down on the side of liability limitation.
Business Opportunity! Ding Ding. Offer a device with wifi/rj 45 that does this with minimal to no set up and it's doable.
Won't matter if someone does. Economies of scale will still favor big central providers. And network effects will do the rest.
There's also the possibility of an appliance, like a NAS, with this functionality built in.
Economies of scale and network effects will drive centralization and push those sorts of solutions out.
Many people like to argue that there's a pendulum in computing, swinging between centralization and decentralization, but I think they're wrong. The natural state of computing resources is centralized. We saw one anomalous swing away from centralization, when the cost of computing dropped dramatically faster than the cost of communication. As soon as communication costs dropped far enough, though, we moved quickly back towards centralization of computing infrastructure, now called "the cloud" rather than "the mainframe". Sure, our pocket and laptop computers are a million times more powerful than our desk-filling PCs were, but we use them mostly as portals to cloud systems which are a trillion times more powerful. Of course, there is still value in having some powerful nodes at the edges -- my desktop machine has 56 cores, 128 GiB RAM and 3 TB of NVMe -- but that is and will be the exception that proves the rule.
Mostly I believe that you don't have many or perhaps any personal contacts in the google these days.
You're certainly welcome to believe what you like. Clearly you've decided to believe that I'm brainwashed or otherwise fooled (or simply lying about my beliefs -- though I do appreciate that you've opted to apply Hanlon's Razor) and I'm not going to change your mind on that point. Which means I'm unlikely to change your mind on any other point, so I'm wasting my time here. But it's a lazy Sunday afternoon and I'm doing this to procrastinate finishing my taxes, so...
The real issue is the abuse of our personal information, and it doesn't really matter if Facebook or the google are doing it directly, or they are playing shell game with the abuse by selling or loaning or giving the data to some other entity for the abusive uses. Actually, in terms of protecting their reputations, I think corporate cancers often prefer such tactics.
A couple of points:
First, your use of the phrase "corporate cancer" indicates a deep and systematic bias that can't help but affect your thinking in all sorts of ways. Corporations have their down sides, clearly, but it's also the case that without them we would have much less of basically everything that enables us to live longer, healthier, happier lives. Doing things on a large scale requires organization and pooling of resources, and of all the forms of organization and resource pooling we have, only corporations and governments can operate at sufficient scale. Governments are good at some things, but very bad at many others, in particular they're very bad at accurately responding to the daily needs of the people (c.f. every attempt at central planning, ever).
It would really help you to get a better perspective on what is good and bad about corporations. They are not an unalloyed good, but they are a net positive. A large net positive.
Second, it does matter whether Facebook or Google are selling data or selling services around the data. Most obviously, it matters because if the data is sold then it will be re-sold, and re-sold, and re-sold, which means there is absolutely no way to control its use and it will eventually be put to every nefarious use possible. As long as it stays locked up in one place, there is the potential to implement controls on it, whether that's through government regulation or through soft power, used to make the owners of the data provide controls in order to avoid the threat of regulation. Or, in the case of Google, through significant internal pressure to do the right thing -- which is the source of the Takeout, deletion and opt-out controls that Google offers.
Of course, there are also significant societal risks in having so much data in one place. One risk is that if Google or Facebook suffered a serious data breach, we're back in the "out of control" situation. Another is that such large piles of data are an attractive target for the other huge concentration of social power, government, and government is also prone to abuse.
Monetize can include renting access to the data; renting analytic access to the data without allowing actual access to the underlying data
I don't see the harm, assuming it's aggregated and anonymized. If the analytic system provides data about individuals, then it can be extracted and we're back into the out-of-control situation.
plus plenty of other nefarious activities
Like?
If your starting point is that this data is a monetisable asset, you're going to end up doing bad things with it.
https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/162/Slippery-Slope
Also, you're moving the goalposts here. Your original claim was that there was no difference between monetizing and selling. I pointed out there's a large difference. Now you're essentially saying that monetizing is also bad. Which is fine, you can try to make that argument (though it does reek of a slippery slope fallacy), but it's a different argument than you started with.
If step one of your plan to replace Facebook is everyone running family servers, your plan is doomed from the start. Most families don't have anyone capable of doing that, and hardly any families have anyone capable of doing it well -- keeping the machine running, updated, and properly secured.
There's a remote chance that it could work if there were a competitive network of service providers who ran the servers. For example, if ISPs did it, the way they all used to run email servers. It might also be somewhat possible if cloud providers operated and maintained the servers. In both cases, though, I think it would just lead the cloud providers to exploit economies of scale by putting up one big infrastructure for all of their users, and to compete by offering features that others don't have... then network effects would kick in and one of them would become dominant and you'd just have a new Facebook.
I think the bottom line is that widely-used services that are subject to intense network effects are natural monopolies. And natural monopolies require regulation.
You're not thinking it through. If we were paying the google to protect our privacy, then the google would have an actual incentive to do it, but the money is coming from corporate cancers that want to have it all. The only concern of the corporate cancers is creating the image that our privacy is being protected.
No, you're not thinking it through. Were Google to sell the data to "the corporate cancers", the latter would have it and would not need to buy it again; Google could only sell new data as they receive it. If Google keeps the data, though, they can monetize its use indefinitely. Plus, the fact is that the data plus Google's analytical abilities are actually of greater value to Google's customers than they could extract from the data themselves. They mostly know this.
In addition, people at Google really aren't assholes (well, some are, like in any population, but fewer than in most) and profit not only isn't their only motive, it's not even the primary motive. Not under the current leadership, anyway. Also, I can tell you that if Google were to start selling data, you wouldn't have to guess or extrapolate. Lots of Google employees would be shouting it from the rooftops. And quitting. Probably in that order.
Monetize vs sell: a distinction without a difference here.
It's a huge difference, assuming the service doesn't provide APIs that allow what Cambridge Analytica did. As long as all of the data is locked up in one place there is the possibility of control, and the likelihood that it will be used only in non-harmful ways. Once it leaks, absolutely anything can happen.
It feels different when it's topsy-turvey, with you turvey instead of topsy, doesn't it?
It feels exactly the same level of ludicrous both ways to me. Actually, it doesn't "feel" any way; I try to think with my brain.
If the purpose of your post was simply to illustrate that profiling is stupid, then we agree. But that's not how it came across. It came across as a serious proposal and you are a sufficiently assholish and irrational leftist that I found that completely believable.
So they got some heavy duty PR, invented a word "jaywalking", pushed this into the press, and lobbied intensely that it was all really the fault of the pedestrians.
Not really, at least with respect to the origin of the term. And it seems it was more automobile clubs than the auto industry that pushed it. https://en.wikipedia.org/wiki/...
The bomber was white, Christian, home-schooled, anti-LGBT and conservative. This fits the profile of almost all domestic terrorists in the US.
That profile also fits millions of non-terrorists. With a sufficiently loose definition of anti-LGBT, it fits a large minority of people where I live. If you remove "home-schooled" as well (and with the aforementioned loose definition) it fits almost everyone.
Your statement is akin to saying that the FBI should be watching all Arab Muslims who go to Mosque daily.
Profiling is questionable in any case, but it's just stupid when the 99.999% of the people who fit the profile aren't a problem.
I think you've proved my point rather succinctly actually.
No, he refuted your point quite succinctly. I'm perfectly happy to call a white guy a terrorist. Timothy McVeigh and Dylan Roof come immediately to mind. But the fact is that terrorism is politically-motivated by definition, and in the absence of a political motive Conditt's actions weren't terrorism. As far as we know, he was just a serial killer.
Replying to cancel accidental mod. Intended to pick "Funny".
a minor crime
You are holding a weapon (a rock). Game over.
And you're holding a weapon (a gun). So does that mean that when you draw yours to shoot the rock-thrower, I'm justified in drawing mine to shoot you? (And, note that while mine is an itty bitty .380, or sometimes a 9mm, not a .357 Magnum, it'll do the job.)
Intent matters here, a lot. Someone throwing a rock at an unoccupied car is not threatening anyone with serious injury or death. Someone throwing the same rock at a person probably is. In no state in the US is deadly force legally justified for preventing vandalism. In only one state (Texas) is it justified even for preventing theft, and then only if there's no other option, the property would not be recoverable and if it's nighttime (and that Texas law is a travesty and should be repealed).
Dude, people like you give responsible gun owners a bad name. Please stop it.
Oh, and by the way, I should point out that in the situation described I absolutely would be legally justified in putting a bullet into your head to stop you from shooting the rock thrower. Deadly force is justified in every state in the union if it's necessary to prevent the unlawful use of deadly force against another. Think about that. Hard.
Hmm. Interesting. Can you cite some case law where the term "protect" was successfully applied to shooting someone for throwing rocks at a car window?
Here's the law: http://codes.findlaw.com/tx/pe...
And, no, it wouldn't cover this case, unless it was clear that the intent of the rock-throwing was to steal the vehicle. The law allows use of deadly force when necessary to prevent "arson, burglary, robbery, aggravated robbery, theft". Note that this only applies during the nighttime, and there are some other caveats as well.
I know just enough to know that one should use the right caliber for the right job and that the tactical advantage of having a gun is lost as soon as someone is aware that you do in fact have a gun.
Right, because the only way to utilize a gun is as a deterrent; it's not as if they shoot real bullets or anything.
Actually, the value of a gun as a deterrent is dependent upon the potential aggressor knowing that you have it. If they don't know about it, it can't be a deterrent, right? (Unless enough people carry concealed that it becomes a general deterrent).
The flip side is that if a potential aggressor knows about your gun, and is determined to attack anyway, then they can try to plan to somehow neutralize you or your gun before you can bring it into play. If the aggressor has a gun, their obvious solution is to shoot you first, assuming they're willing to take the big step up in legal risk of actually using deadly force. If they're an active shooter they'll have no problem with that. If they're a mugger... probably not.
This is a commonly-cited reason for preferring open carry over concealed carry. The fact is that most potential assailants really don't want to tangle with someone with a gun and if they see that their potential target is armed, they'll choose a different target. An obvious exception is those whose goal is to steal the gun, but that's an easily-countered problem.
So, open carry is a strategic advantage, but possibly a tactical disadvantage. I say "possibly" because the lawful carrier is almost always in a reactive mode. The aggressor initiates, and if that means presenting a gun or a knife someone who is carrying concealed probably finds themselves in a position where they can't draw without being shot or stabbed unless they can first create a distraction, or unless they can draw as part of an action expected by the assailant. For example, if a mugger asks for your wallet and you can retrieve a pocket gun instead. A wallet can also make a good distraction. Drop it or toss it and take a step back.
(Though, honestly, if the mugger will be happy with the wallet give them the damned wallet. Whatever is in it is not worth a life and even if you're the sort of asshole who doesn't care about that, whatever is in it is less than what it will cost you in time, inconvenience and probably cash to deal with the fallout of having shot and possibly killed someone.)
They absolutely can, if that is their job as a professional test driver, and they have someone else monitoring the systems.
I don't think this is true. Being a "professional" doesn't change human nature, especially if the human is being asked to remain attentive but inactive for long stretches of time. But there are solutions.
A common practice early in Google's testing was to have three people in the car. A safety driver behind the wheel, an engineer in back to monitor the system and a random Google employee in the front passenger seat who was getting a "fun ride", to experience the coolness of a self-driving car. But... that random employee wasn't just cargo. The engineers realized that the safety drivede a tendency to get complacent, even with adequate breaks. So the "extra" employee provided an additional set of eyes for whom the experience was novel and interesting and who could be expected not only to remain focused on the driving environment but also to stimulate the attention of the safety driver, who acted as "tour guide", pointing out the things the vehicle was reacting to.
I got such a "fun ride" a few years ago. It was rather impressive. What convinced me of the promise of the self-driving tech was when the car noticed and reacted to a cyclist that no human could have seen. I had already realized, intellectually, that the perfect attentiveness of a computer plus the superhuman abilities of LIDAR + RADAR promised dramatically safer "drivers", but seeing it in action drove the point home.
Right, because that's totally a doable thing.
It totally is. Others in the industry (e.g. Google/Waymo) have done it. They use a combination of starting first in safer conditions (closed environments; safer, slower roads; multiple, attentive safety drivers on short shifts) and massive use of simulation. The thing about self-driving systems is that you can take a real data feed and replay it as many times as you want, and you can also alter it or generate fake data to provide a realistic situation that is more challenging.
There's a reason that Google has been working on this problem for years. You have take a methodical, incremental approach to be safe. Uber's trying to take shortcuts, "move fast and break things", and that's not the right approach in a situation where you're operating in the middle of population centers.
Sorry, guys, we're not going to the moon. Gotta get these rocket thingies 100% safe first in controlled conditions.
You'll note that they didn't test or launch the rockets from downtown Tempe. NASA was always very careful to ensure that the only people at risk were those who volunteered for it.
H1Bs are currently handed out by a lottery. Instead, hand them out based on the salary of the employee, highest-paid first.
Really need an H1B employee? Gonna have to pay more.
This will heavily bias H1B towards the east and west coast.
Lottery unfortunately is the most fair way to do it.
You could apply regional adjustments. Relative cost of living is well-known and there are standard adjustment tables that big corporations use when they move employees around. Just apply those to normalize the salaries, then sort.
Solution: Government or public university work. Job security and 4-5 wks of vaca are features, not bugs.
It should also be pointed out that "tech" is a very broad swath of job categories. Tech support, operations support and software development are all very different kinds of jobs, and very different job markets.
Indeed but the question here is did the autonomous vehicle even drive as safely as a careful and considerate driver? It wouldn't surprise me if uber cars were not as well programmed as google cars, a reflection of the companies programming those vehicles.
I agree. I was responding to the OP's hypothetical in which it was found that the vehicle was not at fault, meaning it did at least as well as a human could have done, not claiming that was what happened. And I definitely agree that Google has proceeded cautiously, methodically and responsibly, while it wouldn't surprise me at all if Uber has cut a lot of corners.
And while on the subject, please feel free to discuss non-paying Internships...
In software engineering if someone offers you an unpaid internship you laugh at them. Long and loud.
then we've learned that self-driving does not equal infallible
What will we learn next? That water is wet?
No one with a brain has ever claimed that self-driving cars are or ever could be infallible. Nor do they need to be. They only need to be better than the average human, which is a low bar. And, then, over time they'll get better and better as the algorithms and sensors are refined. In a few decades the NTSB will be combing through car wrecks with something akin to the same scrutiny they apply now to plane crashes because the wrecks will be so rare.
But they'll obviously never be infallible.