>And the key storage would be as secure as the primary key storage you're currently using, so there is no additional vulnerability there either.
Of course there is. Primary key storage is done by *you* - the only way anyone can get access to your device is to target you specifically, and manage to determine your personal key.
Secondary "back door" key storage is done by the government, and used regularly by them if it offers any actual value - and thus there are likely hundreds, if not thousands of individuals with legitimate access to that key. And the moment any *one* of them is subverted - be it by bribe, blackmail, the temptation of the enormous black-market value of that key, or just plain old incompetence, *everyone* loses any real security on their phone. People are after all almost always the weakest link in any halfway decent security system.
Not to mention, one of the prime "bad actors" encryption protects against are the "legitimate" authorities, who have a long rack record of abusing their power - encrypting phones became popular in large part in response to the revelations of illegal government surveillance programs.
Sure, cures can be a decent business model, but treatments are far better - especially for diseases which will kill you or drastically reduce quality of life. People will pay almost as much to survive one more year as they would to be cured completely, the limiting factor is how much money they have available to pay in any particular time-frame. And 30K/year for 10 years is a lot more profitable than 30K once. Plus, it's not like the patent expires and then generics take over the market - everybody wants the brand name thanks to the perceived lower quality of generics (and kickbacks to the doctors prescribing them), not to mention the expense of setting up a production line and getting your specific product approved.
Either way is worrying, since wherever it's stored is vulnerable to compromise. But other devices would mean it's stored on their server, whereas other apps could be a local private database. Not that I'd trust Google to go that route.
But then, I avoid doing anything financial on my phone. I decided it was a choice between being able to explore the app ecosystem, or use my phone for secure purposes, and the latter just wasn't that important to me.
Heck, Google themselves have offered Gboard for years as a replacement for the default on Android. It does swipe-typing, voice-to-text, etc. And without the risk of exposing all your typed information to a third party (Google presumably already spies on you as much as they want, with or without Gboard installed).
Maybe, but that presupposes a cultural shift that would probably lead MOST people to subscribe instead. Might make great sense, but culture and good sense have at best a passing relationship. Especially when marketing is involved.
Meanwhile, if self-driving cars are all that's available, poor people will buy them too - just second- (or fourth-) hand, just as they do with normal cars today. Which is why I said what I did - non-autonomous cars aren't going away until the incremental cost of self-driving systems are low enough that it makes sense to put them even in the lowest-end cars on the market.
I assume you meant to reply to the previous post, the one giving a 50-year forecast for "total AI domination" of the roads.
In which case figure, yeah, it takes 20-25 years to get (almost) all the current cars off the road. But how long does it take to get from where we're at today, to the point that self-driving is cheap and reliable enough that it offers a good value to include in ALL new cars, even the low-end, budget-minded ones?
I think 25-30 years is probably a reasonable estimate. Another 5-10 years, minimum, before self-driving reaches the point that you can confidently take a nap on the way to work during rush hour, really the minimum point at which they become truly valuable for most people. And how much longer beyond that before all the sensor packages get cheap enough to add to a $13k car without losing customers tot he "dumb-car" competition?
Perhaps so - but pulling over is a large part of the point of having a shoulder - if you lose a tire, blow your engine, or suffer any of a wide number of other problems, you HAVE to pull over, or stop in the middle of the road, which is at least as dangerous AND a public navigation hazard.
If your AI is uncertain how to handle a situation, and the passenger doesn't take control, then you now have a car cruising down the road without a competent driver - a major hazard to both the occupants and everyone else nearby. Can you honestly suggest any safer alternative than pulling over?
Continuing on under the assumption that the confusion is transitory noise is how you get people slamming into broken guardrails at full speed.
Oh really? Exactly how useful is a self driving car that requires the passengers to be alert and able to take over control at a moment's notice at all times? It's a novelty at best.
Meanwhile, any car that doesn't pull over for emergency vehicles is a public hazard and should be treated as such. In many places you're required to slow and move to the passing lane (if possible) even when passing a parked police car with its emergency lights on.
However, pulling over and stopping is one of the FIRST things any self-driving car should be able to do - that should be the default behavior anytime the passenger fails to respond to a "take control" prompt. And detecting flashing red or blue lights behind you should be a relatively easy thing to do. Combine the two, and pulling over for emergency vehicles should be extremely easy to implement.
It may be the eventual point, but nobody is there yet. If you're asleep at the wheel in any current "self-driving" car, then that's probably at least a reckless driving charge.
Yep, bad science all around. Including the fact that crocodiles are NOT descended from dinosaurs - they split off from that evolutionary branch only slightly later than the other reptiles, long before anything normally recognized as a dinosaur evolved.
My thought exactly. Crocodiles are NOT descended from dinosaurs, they pre-date them. Other reptiles had split off from the evolutionary branch that led to dinosaurs a bit sooner, but not dramatically so. Birds are pretty much the only living species that are descended from dinosaurs. (hadn't heard about the bottleneck before, I may have to investigate.)
Moreover, the last time I checked crocodiles are carnivores, and have been since before the dinosaurs arose - meaning that being able to taste plant toxins has always been an utterly useless ability for them - if they were ever able to do so, it's completely reasonable to expect they would have lost the ability sometime in the last 200 million years.
Does the wheel fight you so hard that you can't override it in an emergency? Ditto the brakes? Because if not, I would have a hard time believing that he was paying any attention to the road, which is the entire point of the whole "keep your hands on the wheel" thing.
>And then Musk's compensation package?! WTF?! >Tesla shareholders are just plain stupid.
I don't know - seems like a smart bet to me, *especially* if you don't have confidence in his long term abilities: Pay him peanuts unless he pulls an elephant out of his hat, and if he actually manages to do so, then give him a bigger slice of the elephant.
I'll have to read that once I have some time, looks interesting. But from a brief skim of the first, and your summary, it sounds like the proposition is that awareness is based in the older section of the brain, and not the newer cortex - which I would say should be self-evident, unless we assume that a large percentage of vertebrate life lacks awareness, despite observational evidence to the contrary. (Which is hardly uncommon in behavioral and related sciences)
However, at least when skimming I saw no proposal behind the physical *mechanism* of awareness, without which it is irrelevant to the conversation at hand. Arguing against baseless preconceptions is admirable, especially when widely accepted, but only brings you closer to a real answer in the sense that it rules out one place where the answer probably isn't. How much progress that translates to depends entirely on the (generally unknown) limits of the remaining search space.
Oh, there's a vast field of speculation, but if you have any hard experimental evidence on specific source of consciousness I'd love to hear it. Bjorn Merker perhaps? Seems he's published quite a bit, can you point me to the paper where he presents solid evidence as to the mechanism by which awareness is created?
Neuroscience has a lot to tell us about some of the cruder operations of the biomachinery in our skulls that seats our awareness - and the detail continues to improve along with our ability to observe and manipulate individual neurons. We can manipulate the psyche in all manner of fashions, though often not in quite the the ways we first assume - for example, recent evidence suggests that general anesthesia doesn't actually suspend awareness, just induces motor-nerve paralysis and prevents new memories from being formed.
My point is simply that our science tells us basically NOTHING about this topic. All we have is conjecture and assumption - things rightfully recognized as being notoriously detached from reality.
Um, considering that it basically renders you completely blind to the real world, I'd venture a guess that nobody but a complete idiot would use it in public - with the exception of in an arcade where everybody around you is already a self-professed dork anyway. Or at work, where it's a valuable tool, and $#@! what you look like.
Why must they? Unless you actually want them to be able physically touch each other, there's no need for them to be anywhere near each other physically when they encounter each other virtually.
Considering that they say the minimum radius is 22m (i.e. a 44x44m space) without the haptic feedback, I'd say an almost 9x reduction in radius is a bit more than "somewhat" smaller. The extra 2m are, I assume, the width of the hallway, and not directly relevant to the minimum "straight" radius.
Still, I'm having a hard time imagining a lot of compelling VR experiences that require you to keep one hand on the wall at all times.
Lots of options, but mostly very sub-optimal. I've seen - the giant "gerbil ball" models (a nuisance getting in and out of, and you pretty much require wireless VR, which isn't here yet. - the "running in a bowl wearing slippery shoes" model, which is considerably more affordable and doesn't mind wires, but requires an abdominal harness to keep you from falling on your ass, which also prevents a wide range of natural motion (no crouching or bending over too far) - and various actual powered treadmills, which are potentially more stable, but dramatically more expensive.
And of course, all suffer from the fact that you're not *actually* moving, your feet are simply sliding around underneath you, which dramatically affects the physics of motion.
So yeah, there are "solutions", but they're FAR from perfect. Much better suited to your living room, but there's plenty of room for exploring more realistic options as well, even if they're more relevant to the commercial realm.
Really? If a 5x7m space were enough to freely roam through a vast game world from end to end, I'd certainly prefer that to having to teleport between points that I could then explore within the limits of a 5x7m box.
Of course, I wouldn't want to explore the world with my hand constantly on a wall, which seems to be an important part of this trick. Without that it sounds like the minimum radius is ~22m, or a 44x44m space, probably considerably larger since at any moment you might decide to turn away from the "straight" line towards the boundary of the space.
However, that sort of space would potentially fit within a football or soccer field - a large, flat area sitting mostly unused almost anywhere in the world. Sufficiently large empty warehouses might also do the trick. It wouldn't be private VR, but at least there's a path available for infinite-world VR in a commercial setting. Something a bit more stable than having a giant gerbil ball or other omni-treadmill in your basement.
> don't have feelings because they don't have neurons That is still based on the assumption that awareness must have a mechanism behind it. If it's a fundamental property, then that is not the case - there is no (known) mechanism behind the charges, masses, etc. of fundamental particles - they are simply properties inherent in the particle.
>if you poke it in the eye with a sharp stick and the rock says "Ow!" Find me a rock with eyes and the ability to say "ow" and maybe you have the beginning of an argument. Without that, you're simply projecting your own expectations onto a scenario where they are not relevant.
If he can figure out a way to make decent bricks out of whatever he happens to be digging through, that would be pretty impressive. Similarly if he worked out a cost-effective method to produce the bricks on-site, rather than having to ship slag around.
If you think bricks are boring, you're not paying attention to the engineering or logistics surrounding them.
You pretty much have to allow them, as drivers will quite often delay coming into the pits until they're not sure they have enough gas for the next lap. Avoiding even one extra fueling stop over the course of race can gain you a LOT of positions. Deny access to the pits while a safety car is in place would mean a badly timed accident could cause a lot of uninvolved cars to drop out of the race.
That said, while a safety car is in place it seems only reasonable that they shouldn't be allowed to leave the pits ahead of their pre-pit position in the pack.
Did he say anything about "huge" though? Interlocking cinder-block size solid blocks could find a lot of use. So long as they're small enough for one man (or even two) to stack them easily they could be extremely convenient to build with.
Slashdot Mirror
>And the key storage would be as secure as the primary key storage you're currently using, so there is no additional vulnerability there either.
Of course there is. Primary key storage is done by *you* - the only way anyone can get access to your device is to target you specifically, and manage to determine your personal key.
Secondary "back door" key storage is done by the government, and used regularly by them if it offers any actual value - and thus there are likely hundreds, if not thousands of individuals with legitimate access to that key. And the moment any *one* of them is subverted - be it by bribe, blackmail, the temptation of the enormous black-market value of that key, or just plain old incompetence, *everyone* loses any real security on their phone. People are after all almost always the weakest link in any halfway decent security system.
Not to mention, one of the prime "bad actors" encryption protects against are the "legitimate" authorities, who have a long rack record of abusing their power - encrypting phones became popular in large part in response to the revelations of illegal government surveillance programs.
Sure, cures can be a decent business model, but treatments are far better - especially for diseases which will kill you or drastically reduce quality of life. People will pay almost as much to survive one more year as they would to be cured completely, the limiting factor is how much money they have available to pay in any particular time-frame. And 30K/year for 10 years is a lot more profitable than 30K once. Plus, it's not like the patent expires and then generics take over the market - everybody wants the brand name thanks to the perceived lower quality of generics (and kickbacks to the doctors prescribing them), not to mention the expense of setting up a production line and getting your specific product approved.
In other apps, or on other devices?
Either way is worrying, since wherever it's stored is vulnerable to compromise. But other devices would mean it's stored on their server, whereas other apps could be a local private database. Not that I'd trust Google to go that route.
But then, I avoid doing anything financial on my phone. I decided it was a choice between being able to explore the app ecosystem, or use my phone for secure purposes, and the latter just wasn't that important to me.
Heck, Google themselves have offered Gboard for years as a replacement for the default on Android. It does swipe-typing, voice-to-text, etc. And without the risk of exposing all your typed information to a third party (Google presumably already spies on you as much as they want, with or without Gboard installed).
Maybe, but that presupposes a cultural shift that would probably lead MOST people to subscribe instead. Might make great sense, but culture and good sense have at best a passing relationship. Especially when marketing is involved.
Meanwhile, if self-driving cars are all that's available, poor people will buy them too - just second- (or fourth-) hand, just as they do with normal cars today. Which is why I said what I did - non-autonomous cars aren't going away until the incremental cost of self-driving systems are low enough that it makes sense to put them even in the lowest-end cars on the market.
I assume you meant to reply to the previous post, the one giving a 50-year forecast for "total AI domination" of the roads.
In which case figure, yeah, it takes 20-25 years to get (almost) all the current cars off the road. But how long does it take to get from where we're at today, to the point that self-driving is cheap and reliable enough that it offers a good value to include in ALL new cars, even the low-end, budget-minded ones?
I think 25-30 years is probably a reasonable estimate. Another 5-10 years, minimum, before self-driving reaches the point that you can confidently take a nap on the way to work during rush hour, really the minimum point at which they become truly valuable for most people. And how much longer beyond that before all the sensor packages get cheap enough to add to a $13k car without losing customers tot he "dumb-car" competition?
Perhaps so - but pulling over is a large part of the point of having a shoulder - if you lose a tire, blow your engine, or suffer any of a wide number of other problems, you HAVE to pull over, or stop in the middle of the road, which is at least as dangerous AND a public navigation hazard.
If your AI is uncertain how to handle a situation, and the passenger doesn't take control, then you now have a car cruising down the road without a competent driver - a major hazard to both the occupants and everyone else nearby. Can you honestly suggest any safer alternative than pulling over?
Continuing on under the assumption that the confusion is transitory noise is how you get people slamming into broken guardrails at full speed.
Oh really? Exactly how useful is a self driving car that requires the passengers to be alert and able to take over control at a moment's notice at all times? It's a novelty at best.
Meanwhile, any car that doesn't pull over for emergency vehicles is a public hazard and should be treated as such. In many places you're required to slow and move to the passing lane (if possible) even when passing a parked police car with its emergency lights on.
However, pulling over and stopping is one of the FIRST things any self-driving car should be able to do - that should be the default behavior anytime the passenger fails to respond to a "take control" prompt. And detecting flashing red or blue lights behind you should be a relatively easy thing to do. Combine the two, and pulling over for emergency vehicles should be extremely easy to implement.
It may be the eventual point, but nobody is there yet. If you're asleep at the wheel in any current "self-driving" car, then that's probably at least a reckless driving charge.
Yep, bad science all around. Including the fact that crocodiles are NOT descended from dinosaurs - they split off from that evolutionary branch only slightly later than the other reptiles, long before anything normally recognized as a dinosaur evolved.
My thought exactly. Crocodiles are NOT descended from dinosaurs, they pre-date them. Other reptiles had split off from the evolutionary branch that led to dinosaurs a bit sooner, but not dramatically so. Birds are pretty much the only living species that are descended from dinosaurs. (hadn't heard about the bottleneck before, I may have to investigate.)
Moreover, the last time I checked crocodiles are carnivores, and have been since before the dinosaurs arose - meaning that being able to taste plant toxins has always been an utterly useless ability for them - if they were ever able to do so, it's completely reasonable to expect they would have lost the ability sometime in the last 200 million years.
Does the wheel fight you so hard that you can't override it in an emergency? Ditto the brakes? Because if not, I would have a hard time believing that he was paying any attention to the road, which is the entire point of the whole "keep your hands on the wheel" thing.
>And then Musk's compensation package?! WTF?!
>Tesla shareholders are just plain stupid.
I don't know - seems like a smart bet to me, *especially* if you don't have confidence in his long term abilities: Pay him peanuts unless he pulls an elephant out of his hat, and if he actually manages to do so, then give him a bigger slice of the elephant.
I'll have to read that once I have some time, looks interesting. But from a brief skim of the first, and your summary, it sounds like the proposition is that awareness is based in the older section of the brain, and not the newer cortex - which I would say should be self-evident, unless we assume that a large percentage of vertebrate life lacks awareness, despite observational evidence to the contrary. (Which is hardly uncommon in behavioral and related sciences)
However, at least when skimming I saw no proposal behind the physical *mechanism* of awareness, without which it is irrelevant to the conversation at hand. Arguing against baseless preconceptions is admirable, especially when widely accepted, but only brings you closer to a real answer in the sense that it rules out one place where the answer probably isn't. How much progress that translates to depends entirely on the (generally unknown) limits of the remaining search space.
Oh, there's a vast field of speculation, but if you have any hard experimental evidence on specific source of consciousness I'd love to hear it. Bjorn Merker perhaps? Seems he's published quite a bit, can you point me to the paper where he presents solid evidence as to the mechanism by which awareness is created?
Neuroscience has a lot to tell us about some of the cruder operations of the biomachinery in our skulls that seats our awareness - and the detail continues to improve along with our ability to observe and manipulate individual neurons. We can manipulate the psyche in all manner of fashions, though often not in quite the the ways we first assume - for example, recent evidence suggests that general anesthesia doesn't actually suspend awareness, just induces motor-nerve paralysis and prevents new memories from being formed.
My point is simply that our science tells us basically NOTHING about this topic. All we have is conjecture and assumption - things rightfully recognized as being notoriously detached from reality.
Um, considering that it basically renders you completely blind to the real world, I'd venture a guess that nobody but a complete idiot would use it in public - with the exception of in an arcade where everybody around you is already a self-professed dork anyway. Or at work, where it's a valuable tool, and $#@! what you look like.
Why must they? Unless you actually want them to be able physically touch each other, there's no need for them to be anywhere near each other physically when they encounter each other virtually.
Considering that they say the minimum radius is 22m (i.e. a 44x44m space) without the haptic feedback, I'd say an almost 9x reduction in radius is a bit more than "somewhat" smaller. The extra 2m are, I assume, the width of the hallway, and not directly relevant to the minimum "straight" radius.
Still, I'm having a hard time imagining a lot of compelling VR experiences that require you to keep one hand on the wall at all times.
Lots of options, but mostly very sub-optimal. I've seen
- the giant "gerbil ball" models (a nuisance getting in and out of, and you pretty much require wireless VR, which isn't here yet.
- the "running in a bowl wearing slippery shoes" model, which is considerably more affordable and doesn't mind wires, but requires an abdominal harness to keep you from falling on your ass, which also prevents a wide range of natural motion (no crouching or bending over too far)
- and various actual powered treadmills, which are potentially more stable, but dramatically more expensive.
And of course, all suffer from the fact that you're not *actually* moving, your feet are simply sliding around underneath you, which dramatically affects the physics of motion.
So yeah, there are "solutions", but they're FAR from perfect. Much better suited to your living room, but there's plenty of room for exploring more realistic options as well, even if they're more relevant to the commercial realm.
Really? If a 5x7m space were enough to freely roam through a vast game world from end to end, I'd certainly prefer that to having to teleport between points that I could then explore within the limits of a 5x7m box.
Of course, I wouldn't want to explore the world with my hand constantly on a wall, which seems to be an important part of this trick. Without that it sounds like the minimum radius is ~22m, or a 44x44m space, probably considerably larger since at any moment you might decide to turn away from the "straight" line towards the boundary of the space.
However, that sort of space would potentially fit within a football or soccer field - a large, flat area sitting mostly unused almost anywhere in the world. Sufficiently large empty warehouses might also do the trick. It wouldn't be private VR, but at least there's a path available for infinite-world VR in a commercial setting. Something a bit more stable than having a giant gerbil ball or other omni-treadmill in your basement.
> don't have feelings because they don't have neurons
That is still based on the assumption that awareness must have a mechanism behind it. If it's a fundamental property, then that is not the case - there is no (known) mechanism behind the charges, masses, etc. of fundamental particles - they are simply properties inherent in the particle.
>if you poke it in the eye with a sharp stick and the rock says "Ow!"
Find me a rock with eyes and the ability to say "ow" and maybe you have the beginning of an argument. Without that, you're simply projecting your own expectations onto a scenario where they are not relevant.
If he can figure out a way to make decent bricks out of whatever he happens to be digging through, that would be pretty impressive. Similarly if he worked out a cost-effective method to produce the bricks on-site, rather than having to ship slag around.
If you think bricks are boring, you're not paying attention to the engineering or logistics surrounding them.
You pretty much have to allow them, as drivers will quite often delay coming into the pits until they're not sure they have enough gas for the next lap. Avoiding even one extra fueling stop over the course of race can gain you a LOT of positions. Deny access to the pits while a safety car is in place would mean a badly timed accident could cause a lot of uninvolved cars to drop out of the race.
That said, while a safety car is in place it seems only reasonable that they shouldn't be allowed to leave the pits ahead of their pre-pit position in the pack.
Did he say anything about "huge" though? Interlocking cinder-block size solid blocks could find a lot of use. So long as they're small enough for one man (or even two) to stack them easily they could be extremely convenient to build with.