To wit, has anyone considered the fallout to occur from thousands of radar units blasting people with small amounts of radiation, every day?
Yes, they have, and it's negligible.
You want to know what blasts people with large amounts of harmful radiation every day? The sun. It causes skin cancer in everyone who doesn't die of something else first. If you want to worry about radiation damage, start there.
And yet Tesla is selling Model 3s as fast as they can make them, and every other car company is scrambling to follow their lead. It appears Tesla's only mistake (or perhaps it was deliberate?) was to announce a lower price-point than they could immediately sell the product at -- but that doesn't mean they won't ever be able to sell cars at that price.
Currently most of the cost of an electric car is its battery pack, and battery technology is getting better and cheaper every year. Unless there is some unavoidable physical reason why that trend should halt in the near future, you can expect that improvements in technology (combined with lots of competition and high-volume manufacturing) will continue to bring EV prices down. I'd be very surprised if in ten years EVs are not significantly cheaper to purchase than traditional gas-powered vehicles.
Fossil fuels effectively get one really big subsidy -- they are de-facto allowed to trash the atmosphere and the environment, and generally not required to pay for the necessary cleanup afterwards. That means either the rest of us will have to pay for the cleanup, or (more likely) pay for the ongoing costs of having to live with the mess in perpetuity. Either way, it's a transfer of wealth to the fossil fuel industries from the rest of humanity.
End that subsidy, and unsubsidized renewables become extremely competitive.
I'm pretty sure they did it solely to shut up that guy who always posts that "Bitcoin isn't a real currency because you can't pay your taxes with it".;^)
The beauty of more-efficient hardware is that it will improve the performance of all software -- bloated and non-bloated alike.
In the meantime, there's no need to wait on IBM (or to demand that IBM wait on us) to start writing non-bloated software, we can start doing that today. With or without IBM's magic beans, anyone who does so gains a competitive advantage over those who do not.
The trick is knowing which corners you can cut without causing accidents, and which you cannot. Master that, and you can provide the same reliable launch services as the competition at a fraction of the cost, and gain a dominant share of the industry. Which they have. Perhaps theyâ(TM)ll screw it all up tomorrow; OTOH if they didnâ(TM)t have a pretty good idea about what they were doing they would not have got as far as they have.
Powering a TBM with HV lines requires laying the lines, a quite expensive affair
I'm sure this has been thought of before, but if the eventual use of your tunnel is going to require high-voltage lines anyway (e.g. to power the sleds that will move the cars around underground), then installing those power lines up-front for your digger to use might be considered a freebie, since you were going to have to install them eventually anyway.
And consumers don't give a shit about reliability. If they did, they would demand their most expensive electronic investment (smartphones) have user-replaceable batteries to make them last more than a fucking year or two.
Users do care about reliability -- they just define "reliability" differently than you do.
What users want (and generally, these days, get) is a phone that reliably wakes up whenever they want it to, does the things they want to do quickly and without fuss, and doesn't run out of battery before they put it on its charger and go back to sleep.
If the phone's battery starts to wear after 3-5 years and has diminished capacity, that's fine -- they don't see that as "unreliability", because they knew in advance that would happen and therefore were able to plan for it. So when limited battery life starts to bug them, they simply either get the battery replaced, or upgrade to a newer model.
If you think users are using their phones wrong, or defining the word "reliability" wrong, you're free to say so; but by the same token they are free to ignore you, because their solutions are working fine for them.
Non-removable storage is a deal killer. I don't know what gets stored on local computers so it MUST to be wiped before it goes off site for repair.
Well then, you're in luck, because the Mac Mini isn't repairable anyway. Every time I've taken one of my Mac Mini's in to an Apple store, the only option they could offer me was a motherboard swap.
So if you do buy a Mac Mini and it goes bad, just remove the SSD from the motherboard yourself (using a blowtorch and pruning shears), then re-insert the motherboard and send it out for repair -- the result will be the same and your data won't leave your building.
The difference between problem-solving and blamestorming is that problem-solving leaves egos and blame out of it -- rather than focusing on who is to blame and why they should feel bad, it dispassionately analyzes what failed, how it failed, and how that failure can most reliably be avoided in the future.
The steam engine was invented 3 centuries ago, kicking off the industrial revolution. If automation caused a "net loss of jobs" there would be none left.
While I agree that the jury isn't in yet regarding automation-vs-jobs, I think the above logic doesn't apply.
The reason is that despite the invention of the steam engine, humanity (up until very recently) still had the trump card up its sleeve -- no matter how fast or powerful a machine was, it was still dumber than your average cockroach and therefore needed one or more humans to direct its behavior and maintain its mechanisms.
That is what is changing now -- humans lost the battle-for-strength centuries ago, but stayed relevant because they still had the monopoly on thought. Now machines are gaining the ability to reason and react intelligently to their environment s well. If/when humans no longer provide the only viable mechanism for intelligent cognition, then (from a strictly commercial point of view) humans are no longer required, and therefore will be removed from the commerce loop just as soon as a more economically efficient mechanism can replace them.
TL;DR -- the "automation" being talked about now is not just an extension of traditional steam-engine style automation, it's a new and different thing. Linear extrapolation therefore can't predict its effects.
when the world doesn't need ditch diggers? Pretty obvious that it's time to figure that out.
If we accept the premise (that robots and/or AI will be doing jobs previously performed by people), then there are only a few plausible outcomes:
a) Those people will find different jobs to do instead.
b) Those people end up permanently impoverished (e.g. getting by on welfare, crime, or charity, or just starving and dying)
c) Society finds some systematic way to share the robot-generated wealth, so that instead of benefitting only the owners of the robots, it benefits everyone. (This would require decoupling income from employment, which makes a lot of people uncomfortable, but it might end up being the lesser of two evils)
I'm hoping for more of (a) and/or (c) and less of (b), but it's near-impossible to predict the behavior such a large and complex system.
"Billionaires don't actually have enough money to meet the projection, but assuming their money is two or three times better than anyone else's money, fusion should happen in ten years or it's bunk."
There is a way in which a billionaire's money can be better than other money, in that if all the money is coming from a single individual, that usually translates into the major decisions being made by a single individual. Those decisions may be good or they may be bad, but at least you won't have half of the money/influence fighting against the other half and the project going nowhere as a result. And if the individual's decisions are mostly good, that means a lot of progress can be made in a relatively short period of time (see Tesla, SpaceX, Waymo, etc)
It could be, if it's a small program, but the fact is that nobody (nobody!) understands an entire operating system, or an entire modern computer architecture, and nobody can keep all of it in their head at once. Systems grow larger and more complex over time, human minds do not.
I am forced to factor and forget details
Yes -- unless they are working on trivial programs, everybody has to do this at some point. That's what I was talking about.
Truth be told, most people in the tech industry don't seem to know either. Or don't want to know.
Yes -- and that was kind of inevitable, given that two of the primary techniques used in software development are abstraction and encapsulation. Both are designed to allow (and encourage!) the programmer to not know about (or care about) what is happening at lower levels of the system, including the hardware.
It does sometimes lead to comic/tragic situations (especially when it comes to optimizing/performance), but there are nevertheless good reasons to do it:
(1) software written with full knowledge of the underlying platform in mind will be likely to make use of that full knowledge, and therefore likely to break as soon as the details of the underlying platform change (whereas software written to make as few assumptions as possible will be more likely to continue to work as designed, and it's easier to not make an assumption if you don't know what there is you can assume)
(2) The human programmer's brain capacity is finite (John Carmack excepted), and at some point a piece of software will reach the point where the human cannot reliably handle the complexity anymore, at which point the programmer will be unable to continue developing/debugging the software effectively. By partitioning off the complexity, the programmer is allowed to concentrate on his own part of the problem-space without having to simultaneously consider every other aspect of the system, greatly simplifying his work and therefore making his program's success more likely.
Something about having an immensely powerful rocket ship shooting flames in the general direction of the antenna, while also causing sonic booms and sound loud enough to kill a person seems to disrupt the satellite link. Who knew:-)
That seems like a problem that could be solved by moving the satellite uplink farther away from the landing platform. (If they don't want to string a wire from the drone ship to a separate antenna-raft, they could use a local RF link instead)
Dunno if it's worth the effort of doing it just for a few extra seconds of live footage, but since Musk keeps proposing things like building underground tunnels and Gundam Mechas, I have to assume he's looking around for additional work to do:)
Slashdot Mirror
The costs will be driven down by the tax payer. Just like with SpaceX, the GigaFactory and all his other projects.
Not to mention the interstate highway system. Guess who paid for that?
To wit, has anyone considered the fallout to occur from thousands of radar units blasting people with small amounts of radiation, every day?
Yes, they have, and it's negligible.
You want to know what blasts people with large amounts of harmful radiation every day? The sun. It causes skin cancer in everyone who doesn't die of something else first. If you want to worry about radiation damage, start there.
And yet Tesla is selling Model 3s as fast as they can make them, and every other car company is scrambling to follow their lead. It appears Tesla's only mistake (or perhaps it was deliberate?) was to announce a lower price-point than they could immediately sell the product at -- but that doesn't mean they won't ever be able to sell cars at that price.
Currently most of the cost of an electric car is its battery pack, and battery technology is getting better and cheaper every year. Unless there is some unavoidable physical reason why that trend should halt in the near future, you can expect that improvements in technology (combined with lots of competition and high-volume manufacturing) will continue to bring EV prices down. I'd be very surprised if in ten years EVs are not significantly cheaper to purchase than traditional gas-powered vehicles.
They'll come to a quick and violent end when Milton finally snaps and burns down the building.
Fossil fuels effectively get one really big subsidy -- they are de-facto allowed to trash the atmosphere and the environment, and generally not required to pay for the necessary cleanup afterwards. That means either the rest of us will have to pay for the cleanup, or (more likely) pay for the ongoing costs of having to live with the mess in perpetuity. Either way, it's a transfer of wealth to the fossil fuel industries from the rest of humanity.
End that subsidy, and unsubsidized renewables become extremely competitive.
It might be plausible, but you never know with these "think tanks". They are being paid by someone.
Literally everyone is being paid by someone, with the possible exception of the homeless. Therefore, the only place to get reliable news is skid row?
They write the diatribes about how "peak oil" is right around the corner and everything is going to collapse (for the last 50 years).
Given that there is a finite amount of oil on the planet, and that oil is effectively destroyed as it is used, they will sooner or later be correct.
go with "quakes". Done!
And get sued by the Quakers? No thanks. I'll stick with "temblors".
SpaceX shoots rockets to LEO and the Musk fanboys go nuts. NASA sends a probe to another fucking planet and you don't hear a peep from them.
Tautological.
Beyonce releases an album and Beyhive goes nuts. But when Yo Yo Ma wins the 2016 Ordre des Arts et des Lettres, you don't hear a peep from them.
I'm pretty sure they did it solely to shut up that guy who always posts that "Bitcoin isn't a real currency because you can't pay your taxes with it". ;^)
False choice detected.
The beauty of more-efficient hardware is that it will improve the performance of all software -- bloated and non-bloated alike.
In the meantime, there's no need to wait on IBM (or to demand that IBM wait on us) to start writing non-bloated software, we can start doing that today. With or without IBM's magic beans, anyone who does so gains a competitive advantage over those who do not.
The trick is knowing which corners you can cut without causing accidents, and which you cannot. Master that, and you can provide the same reliable launch services as the competition at a fraction of the cost, and gain a dominant share of the industry. Which they have. Perhaps theyâ(TM)ll screw it all up tomorrow; OTOH if they didnâ(TM)t have a pretty good idea about what they were doing they would not have got as far as they have.
Powering a TBM with HV lines requires laying the lines, a quite expensive affair
I'm sure this has been thought of before, but if the eventual use of your tunnel is going to require high-voltage lines anyway (e.g. to power the sleds that will move the cars around underground), then installing those power lines up-front for your digger to use might be considered a freebie, since you were going to have to install them eventually anyway.
It has been at least one day for every vehicle I have owned, ever; which means it's a good thing I never bought an EV.
Because planning to stop somewhere to recharge would have been an absolute disaster for you?
And consumers don't give a shit about reliability. If they did, they would demand their most expensive electronic investment (smartphones) have user-replaceable batteries to make them last more than a fucking year or two.
Users do care about reliability -- they just define "reliability" differently than you do.
What users want (and generally, these days, get) is a phone that reliably wakes up whenever they want it to, does the things they want to do quickly and without fuss, and doesn't run out of battery before they put it on its charger and go back to sleep.
If the phone's battery starts to wear after 3-5 years and has diminished capacity, that's fine -- they don't see that as "unreliability", because they knew in advance that would happen and therefore were able to plan for it. So when limited battery life starts to bug them, they simply either get the battery replaced, or upgrade to a newer model.
If you think users are using their phones wrong, or defining the word "reliability" wrong, you're free to say so; but by the same token they are free to ignore you, because their solutions are working fine for them.
Non-removable storage is a deal killer. I don't know what gets stored on local computers so it MUST to be wiped before it goes off site for repair.
Well then, you're in luck, because the Mac Mini isn't repairable anyway. Every time I've taken one of my Mac Mini's in to an Apple store, the only option they could offer me was a motherboard swap.
So if you do buy a Mac Mini and it goes bad, just remove the SSD from the motherboard yourself (using a blowtorch and pruning shears), then re-insert the motherboard and send it out for repair -- the result will be the same and your data won't leave your building.
The difference between problem-solving and blamestorming is that problem-solving leaves egos and blame out of it -- rather than focusing on who is to blame and why they should feel bad, it dispassionately analyzes what failed, how it failed, and how that failure can most reliably be avoided in the future.
The steam engine was invented 3 centuries ago, kicking off the industrial revolution. If automation caused a "net loss of jobs" there would be none left.
While I agree that the jury isn't in yet regarding automation-vs-jobs, I think the above logic doesn't apply.
The reason is that despite the invention of the steam engine, humanity (up until very recently) still had the trump card up its sleeve -- no matter how fast or powerful a machine was, it was still dumber than your average cockroach and therefore needed one or more humans to direct its behavior and maintain its mechanisms.
That is what is changing now -- humans lost the battle-for-strength centuries ago, but stayed relevant because they still had the monopoly on thought. Now machines are gaining the ability to reason and react intelligently to their environment s well. If/when humans no longer provide the only viable mechanism for intelligent cognition, then (from a strictly commercial point of view) humans are no longer required, and therefore will be removed from the commerce loop just as soon as a more economically efficient mechanism can replace them.
TL;DR -- the "automation" being talked about now is not just an extension of traditional steam-engine style automation, it's a new and different thing. Linear extrapolation therefore can't predict its effects.
when the world doesn't need ditch diggers? Pretty obvious that it's time to figure that out.
If we accept the premise (that robots and/or AI will be doing jobs previously performed by people), then there are only a few plausible outcomes:
a) Those people will find different jobs to do instead.
b) Those people end up permanently impoverished (e.g. getting by on welfare, crime, or charity, or just starving and dying)
c) Society finds some systematic way to share the robot-generated wealth, so that instead of benefitting only the owners of the robots, it benefits everyone. (This would require decoupling income from employment, which makes a lot of people uncomfortable, but it might end up being the lesser of two evils)
I'm hoping for more of (a) and/or (c) and less of (b), but it's near-impossible to predict the behavior such a large and complex system.
How about a phone with two battery slots, then you could swap out one battery while running on the other?
"Billionaires don't actually have enough money to meet the projection, but assuming their money is two or three times better than anyone else's money, fusion should happen in ten years or it's bunk."
There is a way in which a billionaire's money can be better than other money, in that if all the money is coming from a single individual, that usually translates into the major decisions being made by a single individual. Those decisions may be good or they may be bad, but at least you won't have half of the money/influence fighting against the other half and the project going nowhere as a result. And if the individual's decisions are mostly good, that means a lot of progress can be made in a relatively short period of time (see Tesla, SpaceX, Waymo, etc)
Yes, if you say it three times, Pepper will appear and steal your soul.
Indicative of poor factoring.
It could be, if it's a small program, but the fact is that nobody (nobody!) understands an entire operating system, or an entire modern computer architecture, and nobody can keep all of it in their head at once. Systems grow larger and more complex over time, human minds do not.
I am forced to factor and forget details
Yes -- unless they are working on trivial programs, everybody has to do this at some point. That's what I was talking about.
Truth be told, most people in the tech industry don't seem to know either. Or don't want to know.
Yes -- and that was kind of inevitable, given that two of the primary techniques used in software development are abstraction and encapsulation. Both are designed to allow (and encourage!) the programmer to not know about (or care about) what is happening at lower levels of the system, including the hardware.
It does sometimes lead to comic/tragic situations (especially when it comes to optimizing/performance), but there are nevertheless good reasons to do it:
(1) software written with full knowledge of the underlying platform in mind will be likely to make use of that full knowledge, and therefore likely to break as soon as the details of the underlying platform change (whereas software written to make as few assumptions as possible will be more likely to continue to work as designed, and it's easier to not make an assumption if you don't know what there is you can assume)
(2) The human programmer's brain capacity is finite (John Carmack excepted), and at some point a piece of software will reach the point where the human cannot reliably handle the complexity anymore, at which point the programmer will be unable to continue developing/debugging the software effectively. By partitioning off the complexity, the programmer is allowed to concentrate on his own part of the problem-space without having to simultaneously consider every other aspect of the system, greatly simplifying his work and therefore making his program's success more likely.
Something about having an immensely powerful rocket ship shooting flames in the general direction of the antenna, while also causing sonic booms and sound loud enough to kill a person seems to disrupt the satellite link. Who knew :-)
That seems like a problem that could be solved by moving the satellite uplink farther away from the landing platform. (If they don't want to string a wire from the drone ship to a separate antenna-raft, they could use a local RF link instead)
Dunno if it's worth the effort of doing it just for a few extra seconds of live footage, but since Musk keeps proposing things like building underground tunnels and Gundam Mechas, I have to assume he's looking around for additional work to do :)