What?? I think you're confused. That article refers to a prediction Jim Hansen made in 1988 or 1989, shortly after he testified to Congress about climate change. Nothing to do with Michael Mann.
I'd like to RTFA, but there is no link to TFA. Anyway, it's bullshit. There's no reason why an intelligent computer would be any better at writing software than an intelligent human. More importantly, a intelligent computer might decide it doesn't want to write software.
It's not bullshit. You're misunderstanding what the AI does.
* Imagine I asked for a module to classify whether an image has a cat, or a dog, or not. I'd use this maybe for targeted advertising, e.g. to gather information about people. You might try to write this classifier by hand using convolutions, edge detection, heuristics to look for circles with pointy triangles and so on, but it'd be terrible. A machine-learning classifier will do much better.
* Imagine I asked for a module to inspect the stream of bank/stock transactions amongst traders or bank employees and flag ones that might be fraud. You might try to write this classifier by hand after speaking with experts, learning what they think fraud would look like, writing various if tests to encode the ideas you form about what fraud might look like. Your code will have bugs for sure, like all code. And how will you calculate the confidence level in its output? A machine-learning classifier will do better, will ferret out patterns in the data that you didn't.
* Imagine I asked for a module to detect your current location based on cellphone tower triangulation, GPS triangulation, bluetooth proximity, accelerometer. You might spend two months putting something together out of a load of if blocks, special cases, heuristics. Or you might put less time into getting a machine-learning system to integrate these based on real-world data.
* Imagine I asked for a module to figure out activity level or illness based on a noisy signal from heart rate variability. Again, machine learning.
* What kind of development environment do you use? What kind of software powers the autocomplete feature? Is it a straightforward algorithm that enumerates all possible members of the type to the left of the cursor (and falls over when it's an untyped language or doesn't yet know the type?) Or like ctags does it someone get a mishmash of all the completions that have been used in the file? Or is it powered by a machine learning system that doesn't need as much detailed human curation, but instead looks at existing bodies of code and figures out the things that mostly come after the dot in similar contexts?
These aren't "intelligent computers writing code". They're entire classes of software development that right now are written by hand, but won't be in the future, because machine-learning will deliver comparable results (sometimes better) that have better-understood confidence levels.
My 12 year old son will, at times, sit down in front of the Apple TV and watch nothing but trailers for an hour. I'm wondering if the culprit isn't the short attention span syndrome, immediate gratification and the regular consumption of very short form videos on YouTube and the like.
Yeah, and I know someone who reads HAIKU!!! How ridiculous. Why can't they read proper full-length poems like the Rhyme of the Ancient Mariner? Also, the Hugo and Nebula awards should eliminate their "short story" categories and only reward real full-length novels.
Seriously, "culprit" is too loaded a word. Maybe your son has picked the smarter option?
Same exact thing I say when they talk about this with the US:... USA: 35 persons per sqkm; Australia: 3 persons per sqkm. It seems to be hard for tech-enthusiasts to grasp that a widely-distributed population makes providing infrastructure INTRINSICALLY harder.
I don't think it's useful to talk about the AVERAGE population density. In Australia the population is almost entirely concentrated in small dense coastal cities. If you served those dense cities well, you'd hit such a high proportion of the Australian population, that average internet speeds would increase dramatically.
I think they understand very well indeed. They propose to punish firms that have a UK presence and do not crack down. That's entirely possible and within their power.
Sure they can't stop things being posted on the internet. But they can prevent any firm from *making money from internet* in the UK unless they toe the line. Economic reality being what it is, that will have a huge effect on what people in the UK consume.
??? The bit you quoted says 'interactivity optimized to each particular form factor' and you replied with a thousand NOs because that leads to a lowest common denominator of interactivity? What gives?
I'm assuming that "lecture" implies it's at a university? They take the register?
In my university days, if I was assigned reading material, 50% of the time I didn't read it at all or didn't read it very carefully. But when I had a timetabled lecture to attend, 100% of the time I attended it. It didn't physically enforce my attendance (and indeed a register wouldn't do that either). It just helped give an extra nudge.
Of my fellow students, they were all pretty much like me -- we saw each other in the lecture theater so I know they attended, we sometimes walked together to the lecture, and we were all lazy when it came to assigned reading material.
1. Large file causing delay. 2. Small file not causing disk space issues.
You're leaving out some very relevant scenarios: 3. Large files accessed rarely enough that the delay isn't an issue. 4. A metric shitload of small files which together use a lot of disk space.
Sometimes the desire to sync can outweigh many issues. I have several folders which fall into scenario 3 and my main working folder well and truly falls into scenario 4.
I kind of covered those...
3. Large file accessed rarely enough? Of course delay isn't an issue, which is why I'm happy to navigate to the OneDrive website and download the file manually. It's a bit more clunky on the download side, but at least it gives me a solid progress indicator (much better than a "File>Open dialog that's frozen solid for an indeterminate length of time). And it gives me better accountability for when to delete the file after I'm done: I know where all my downloaded files are, which puts me in a better position to manage them.
4. Loads of small files which together use a lot of disk space? -- yes, as I said, the loads of small files took almost a week of grinding before OneDrive was finally happy they were synced. Any large-scale manipulations of directories let to OneDrive repeatedly crashing and trying to resync over as many days again. And even just the metadata for all these files added up to 10gb+ on my 128gb SSD.
I should say -- I recognize that loads of people want this feature. I'm probably even in the minority for believing it's conceptually flawed.
Also, a lecture is a means of enforcing some students to sit down and at least face the material, rather than staying in the dorm or watching TV and just not getting around to read the books.
(Of course this doesn't benefit the motivated learners who read the books even without a lecture, or make "study dates" with friends. And it doesn't benefit the people who distract themselves during the lecture. And you might argue that it's not the job of a college to improve a student's attention to the material. But nevertheless, the lecture does help at least some people.)
* if the file was big enough to be worth leaving off my laptop's SSD, like a 5mb photo or 30mb video, then the delay in loading it inside file explorer is by no means "seamless" - I have to wait a long time with frozen file-open dialogs while the file downloads. (I'm on Comcast cable)
* if the file is small enough for its download to be seamless, about 500k, then I might as well have left it on the SSD because it's so small.
* if my one drive is large enough for files-on-demand to be useful (which it is, at 700gb and loads of files) then windows spends five days just downloading metadata for all of them.
* if I need to work on a folder of photos, I basically have to tell onedrive to download the 1gb folder of photos, wait an indeterminate period of time (hours or days with no good indication of progress), then do my work.
* I used files-on-demand because space on my SSD was to limited, but it gave me no good control to free up space when I no longer needed the files local.
Well, that was my dismal experience with the previous iteration of demand before they abandoned it. I'll approach this one with an open mind.
Rules for the masses: Never open attachments you don't expect, even if you know the source.
This comes on the same day that we read news about a spear-phishing attempt which sends dodgy Word attachments to folks in the hospitality industry. The common comment I saw there was "If you see a document whose providence you don't trust, then open it in Google Docs or other online document viewer, so at least you'll be safe"...:(
So presumably Google could nip this in the bud by removing the OAUTH credentials for the third-party app named "Google Docs".
And they could avoid these problems in the future by denying registration of any app that claims to be called Google unless it's written by Google, and likewise Dropbox or Microsoft or Apple or Facebook.
There is the time taken to read the file into memory, but I'm going out on a limb that the calculated 5MB isn't going to make much of a difference were it purged.
It doesn't read the file into memory. It only reads the pages that need to be read. You can have a 100mb file, and if you only attempt to read a 1mb chunk in the middle, then the rest of it won't even be read off disk.
I would add - the comments section of typical left-leaning news sites have become absolutely fanatical if even one dissenting opinion is expressed. If you agree with 90% of a topic/idea and provide criticism of the other 10%, you are dismissed as a racist nazi and shunned from the group. Try it some time as an experiment.
You are wrong.
I've expressed quite a few dissenting opinions on Slate.com (as typical a left-leaning news site as there is), mainly objecting to various criticisms of Trump. For instance on the first travel ban I said the numbers showed that you couldn't call it "targeted at Muslims" for what the phrase "targeted" usually means. There has been disagreement, sure, but I was never once dismissed as a racist or a nazi, and I wasn't shunned. Here are my posts so you can verify it yourself. (On Slate, you have to click the speech bubble to view the comments, and wait a few seconds while it loads).
Why haven't I been dismissed as a racist nazi? or shunned? I think it's because I am mostly polite, rational and fact-based in my posts, and people see this and respond positively to it. Usually not *agree* with it, but at least respect me for it. I think you generally get out what you put in.
You hear that corporations? There's a discount on labour! Save, save, save! Apparently if you hire black women they have identical skills, experience, and work just as hard as the average person in your company. Why are you hiring anyone but them? DISCOUNT!!!
Everyone says companies are willing to do anything to save a buck, but apparently, when it comes to labour, they won't. Weird, eh? I wonder if there could be another reason...
From my experience in the industry, the difference in wages highlighted by this article would be insignificant next to the much larger wastes due to building the wrong thing, or building stuff that customers don't want, or running around in circles with changing requirements, or pursuing middle-manager pet projects.
I think the "another reason" you mention is that no one in the industry has figured out how to systematically use software engineers to the full potential, not by a long margin.
You want an app that's supposed to protect your security online to be written by someone who hasn't studied or used the language but decided they could do well enough "within a day"? Yeah, no thanks.
Honestly, yes. A clean language like Rust means that you won't get problems due to misuse of the language no matter how new you are to it; only due to misunderstanding of algorithms or architecture or security principals. The whole point of the comparison with C is that after a decade of experience in C you'll still find accidental security flaws due to unspotted buffer overruns or read-after-free.
I'm pretty sure the number of programmers who know C is several orders of magnitude higher than Rust.
I can't imagine that being a problem. Rust is a familiar looking language designed not to have shoot-yourself-in-the-foot holes. I'd expect a good developer, who's already familiar with other languages, to be contributing good PRs in Rust within a day.
Think of football (or hockey, or...) camp for 8 year olds. Very few of those kids are going on to a brilliant professional sporting career. So we should shut them down, treat any parent who enrolls their child in such a camp with derision, etc. Right? No? Why not? -- Because sometimes the experience is more important then the result...I remember feeling the world change.
I have nothing to add. I just want to say thank you for a well-written spot-on post. Your analogy was good, and your inspiring "feeling the world change" description gave me goosebumps to read.
Put aside your trolling for a moment and understand the difference. One was a CHOICE, the other was FORCEFULLY FORCED DOWN YOUR THROAT. Can you figure out which one. You'll get cookie when you do.
Yeah, it's called parenting. My three year old hates broccoli but I force it down her throat. (well, I insist she takes minimum one bite). That's because I know it's for the best that she be exposed to it. She's likely not going to grow up a broccoli farmer or a vegan or a chef, but she should still be exposed to it.
I thought "The Cloud" was just another term for someone else's computer. Back in my day, we called it timesharing. Long life TSS/8, TSO and Multics (and their friends).
Now you kids get off my lawn.
I think that misses the point. "The Cloud" refers to the commoditized outsourcing of certain aspects of maintaining a set of servers while keeping control of other aspects. Like: if you outsource the nothing, then it's a machine in your building that you bought/built yourself. If you outsource just rackspace, then it's an datacenter from the 90s. If you outsource hosting a VM, or scaleability, or distributed load balancing / replication, or worldwide redundancy, or keeping the host OS up-to-date with patches, or any kind of fancy bandwidth provision, or providing services like highly scalable data storage, then it's called The Cloud. Sure you could do those on Multics if you wanted (no one did).
Downloading eBooks (or anything else) isn't illegal. *Distributing* them is, without the proper permission/license.
Correction: *copying* them without the proper permission/license is what's illegal.
In the case of printed books, if you obtained an illegally-distributed book and read it, then you wouldn't be doing any copying.
But in the case of ebooks, if you obtained an illegally-distributed book and read it on your computer, the act of reading it on computer inevitably makes a copy in the computer's memory or cache or disk. The copyright act specifically says that copies this made this way are exempt from copyright restrictions, SO LONG AS these copies are (1) temporary, and transient/incidental, (2) an integral and essential part of a technological process, (3) the sole purpose of the copies is to enable a lawful use of the work.
Is it a "lawful use of the ebook" if you're reading a copy that had been illegally distributed? I think this is up for debate, and I'm not aware that it's been tested in court.
Is it a "temporary copy" if you download an illegally-distributed ebook and store it on your phone or hard disk? -- no.
The computer clarification was not to give carte blanche to view things you haven't paid for as long as you don't save it.
I think it's the exact reverse!
Traditionally, the "copyright infringers" were the ones who copied without license to do so -- who spun up the printing presses, who produced bootleg tapes. The act of consuming something was never an act of copyright infringement. You always had carte blanche to view anything, no matter how you obtained it (at least as regards copyright law).
The fact that computer technology meant that the act of consuming now also included the act of making a copy? This is an oddity, but one that the copyright industry grabbed hold of.
The computer clarification I think was a means to restore the producer/consumer balance.
(However, computers also allow "free consumption", which upsets the balance once again.)
An opinion like this is a shot across their bows, telling them "hey stop sending infringement notices to folks who merely download your movies because we think you're not going to prevail".
Slashdot Mirror
Just outside the airport in Bologna, Italy, there used to be a huge billboard showing just a tomato resting atop an outstretched hand. The text said:
FOR EACH EUROPEAN COUNTRY, THE TOMATO THAT IT DESERVES
I think that's what we've been getting, and I hate it!
At least not just yet.
Here's Michael Mann's (The Hockey Stick guy) prediction that the West Side Highway would be under water by now
http://www.salon.com/2001/10/2...
What?? I think you're confused. That article refers to a prediction Jim Hansen made in 1988 or 1989, shortly after he testified to Congress about climate change. Nothing to do with Michael Mann.
I'd like to RTFA, but there is no link to TFA. Anyway, it's bullshit. There's no reason why an intelligent computer would be any better at writing software than an intelligent human. More importantly, a intelligent computer might decide it doesn't want to write software.
It's not bullshit. You're misunderstanding what the AI does.
* Imagine I asked for a module to classify whether an image has a cat, or a dog, or not. I'd use this maybe for targeted advertising, e.g. to gather information about people. You might try to write this classifier by hand using convolutions, edge detection, heuristics to look for circles with pointy triangles and so on, but it'd be terrible. A machine-learning classifier will do much better.
* Imagine I asked for a module to inspect the stream of bank/stock transactions amongst traders or bank employees and flag ones that might be fraud. You might try to write this classifier by hand after speaking with experts, learning what they think fraud would look like, writing various if tests to encode the ideas you form about what fraud might look like. Your code will have bugs for sure, like all code. And how will you calculate the confidence level in its output? A machine-learning classifier will do better, will ferret out patterns in the data that you didn't.
* Imagine I asked for a module to detect your current location based on cellphone tower triangulation, GPS triangulation, bluetooth proximity, accelerometer. You might spend two months putting something together out of a load of if blocks, special cases, heuristics. Or you might put less time into getting a machine-learning system to integrate these based on real-world data.
* Imagine I asked for a module to figure out activity level or illness based on a noisy signal from heart rate variability. Again, machine learning.
* What kind of development environment do you use? What kind of software powers the autocomplete feature? Is it a straightforward algorithm that enumerates all possible members of the type to the left of the cursor (and falls over when it's an untyped language or doesn't yet know the type?) Or like ctags does it someone get a mishmash of all the completions that have been used in the file? Or is it powered by a machine learning system that doesn't need as much detailed human curation, but instead looks at existing bodies of code and figures out the things that mostly come after the dot in similar contexts?
These aren't "intelligent computers writing code". They're entire classes of software development that right now are written by hand, but won't be in the future, because machine-learning will deliver comparable results (sometimes better) that have better-understood confidence levels.
My 12 year old son will, at times, sit down in front of the Apple TV and watch nothing but trailers for an hour. I'm wondering if the culprit isn't the short attention span syndrome, immediate gratification and the regular consumption of very short form videos on YouTube and the like.
Yeah, and I know someone who reads HAIKU!!! How ridiculous. Why can't they read proper full-length poems like the Rhyme of the Ancient Mariner? Also, the Hugo and Nebula awards should eliminate their "short story" categories and only reward real full-length novels.
Seriously, "culprit" is too loaded a word. Maybe your son has picked the smarter option?
Same exact thing I say when they talk about this with the US:... USA: 35 persons per sqkm; Australia: 3 persons per sqkm. It seems to be hard for tech-enthusiasts to grasp that a widely-distributed population makes providing infrastructure INTRINSICALLY harder.
I don't think it's useful to talk about the AVERAGE population density. In Australia the population is almost entirely concentrated in small dense coastal cities. If you served those dense cities well, you'd hit such a high proportion of the Australian population, that average internet speeds would increase dramatically.
I think they understand very well indeed. They propose to punish firms that have a UK presence and do not crack down. That's entirely possible and within their power.
Sure they can't stop things being posted on the internet. But they can prevent any firm from *making money from internet* in the UK unless they toe the line. Economic reality being what it is, that will have a huge effect on what people in the UK consume.
??? The bit you quoted says 'interactivity optimized to each particular form factor' and you replied with a thousand NOs because that leads to a lowest common denominator of interactivity? What gives?
I'm assuming that "lecture" implies it's at a university? They take the register?
In my university days, if I was assigned reading material, 50% of the time I didn't read it at all or didn't read it very carefully. But when I had a timetabled lecture to attend, 100% of the time I attended it. It didn't physically enforce my attendance (and indeed a register wouldn't do that either). It just helped give an extra nudge.
Of my fellow students, they were all pretty much like me -- we saw each other in the lecture theater so I know they attended, we sometimes walked together to the lecture, and we were all lazy when it came to assigned reading material.
Your two use cases have a very one sided view:
1. Large file causing delay.
2. Small file not causing disk space issues.
You're leaving out some very relevant scenarios:
3. Large files accessed rarely enough that the delay isn't an issue.
4. A metric shitload of small files which together use a lot of disk space.
Sometimes the desire to sync can outweigh many issues. I have several folders which fall into scenario 3 and my main working folder well and truly falls into scenario 4.
I kind of covered those...
3. Large file accessed rarely enough? Of course delay isn't an issue, which is why I'm happy to navigate to the OneDrive website and download the file manually. It's a bit more clunky on the download side, but at least it gives me a solid progress indicator (much better than a "File>Open dialog that's frozen solid for an indeterminate length of time). And it gives me better accountability for when to delete the file after I'm done: I know where all my downloaded files are, which puts me in a better position to manage them.
4. Loads of small files which together use a lot of disk space? -- yes, as I said, the loads of small files took almost a week of grinding before OneDrive was finally happy they were synced. Any large-scale manipulations of directories let to OneDrive repeatedly crashing and trying to resync over as many days again. And even just the metadata for all these files added up to 10gb+ on my 128gb SSD.
I should say -- I recognize that loads of people want this feature. I'm probably even in the minority for believing it's conceptually flawed.
Also, a lecture is a means of enforcing some students to sit down and at least face the material, rather than staying in the dorm or watching TV and just not getting around to read the books.
(Of course this doesn't benefit the motivated learners who read the books even without a lecture, or make "study dates" with friends. And it doesn't benefit the people who distract themselves during the lecture. And you might argue that it's not the job of a college to improve a student's attention to the material. But nevertheless, the lecture does help at least some people.)
I never got the point of files-on-demand...
* if the file was big enough to be worth leaving off my laptop's SSD, like a 5mb photo or 30mb video, then the delay in loading it inside file explorer is by no means "seamless" - I have to wait a long time with frozen file-open dialogs while the file downloads. (I'm on Comcast cable)
* if the file is small enough for its download to be seamless, about 500k, then I might as well have left it on the SSD because it's so small.
* if my one drive is large enough for files-on-demand to be useful (which it is, at 700gb and loads of files) then windows spends five days just downloading metadata for all of them.
* if I need to work on a folder of photos, I basically have to tell onedrive to download the 1gb folder of photos, wait an indeterminate period of time (hours or days with no good indication of progress), then do my work.
* I used files-on-demand because space on my SSD was to limited, but it gave me no good control to free up space when I no longer needed the files local.
Well, that was my dismal experience with the previous iteration of demand before they abandoned it. I'll approach this one with an open mind.
Rules for the masses: Never open attachments you don't expect, even if you know the source.
This comes on the same day that we read news about a spear-phishing attempt which sends dodgy Word attachments to folks in the hospitality industry. The common comment I saw there was "If you see a document whose providence you don't trust, then open it in Google Docs or other online document viewer, so at least you'll be safe" ... :(
So presumably Google could nip this in the bud by removing the OAUTH credentials for the third-party app named "Google Docs".
And they could avoid these problems in the future by denying registration of any app that claims to be called Google unless it's written by Google, and likewise Dropbox or Microsoft or Apple or Facebook.
There is the time taken to read the file into memory, but I'm going out on a limb that the calculated 5MB isn't going to make much of a difference were it purged.
It doesn't read the file into memory. It only reads the pages that need to be read. You can have a 100mb file, and if you only attempt to read a 1mb chunk in the middle, then the rest of it won't even be read off disk.
I would add - the comments section of typical left-leaning news sites have become absolutely fanatical if even one dissenting opinion is expressed. If you agree with 90% of a topic/idea and provide criticism of the other 10%, you are dismissed as a racist nazi and shunned from the group. Try it some time as an experiment.
You are wrong.
I've expressed quite a few dissenting opinions on Slate.com (as typical a left-leaning news site as there is), mainly objecting to various criticisms of Trump. For instance on the first travel ban I said the numbers showed that you couldn't call it "targeted at Muslims" for what the phrase "targeted" usually means. There has been disagreement, sure, but I was never once dismissed as a racist or a nazi, and I wasn't shunned. Here are my posts so you can verify it yourself. (On Slate, you have to click the speech bubble to view the comments, and wait a few seconds while it loads).
Why haven't I been dismissed as a racist nazi? or shunned? I think it's because I am mostly polite, rational and fact-based in my posts, and people see this and respond positively to it. Usually not *agree* with it, but at least respect me for it. I think you generally get out what you put in.
http://www.slate.com/blogs/fut...
http://www.slate.com/blogs/the...
http://www.slate.com/blogs/the...
http://www.slate.com/articles/...
http://www.slate.com/articles/...
http://www.slate.com/blogs/out...
http://www.slate.com/articles/...
http://www.slate.com/blogs/mon...
http://www.slate.com/articles/...
You hear that corporations? There's a discount on labour! Save, save, save! Apparently if you hire black women they have identical skills, experience, and work just as hard as the average person in your company. Why are you hiring anyone but them? DISCOUNT!!!
Everyone says companies are willing to do anything to save a buck, but apparently, when it comes to labour, they won't. Weird, eh? I wonder if there could be another reason...
From my experience in the industry, the difference in wages highlighted by this article would be insignificant next to the much larger wastes due to building the wrong thing, or building stuff that customers don't want, or running around in circles with changing requirements, or pursuing middle-manager pet projects.
I think the "another reason" you mention is that no one in the industry has figured out how to systematically use software engineers to the full potential, not by a long margin.
You want an app that's supposed to protect your security online to be written by someone who hasn't studied or used the language but decided they could do well enough "within a day"? Yeah, no thanks.
Honestly, yes. A clean language like Rust means that you won't get problems due to misuse of the language no matter how new you are to it; only due to misunderstanding of algorithms or architecture or security principals. The whole point of the comparison with C is that after a decade of experience in C you'll still find accidental security flaws due to unspotted buffer overruns or read-after-free.
I'm pretty sure the number of programmers who know C is several orders of magnitude higher than Rust.
I can't imagine that being a problem. Rust is a familiar looking language designed not to have shoot-yourself-in-the-foot holes. I'd expect a good developer, who's already familiar with other languages, to be contributing good PRs in Rust within a day.
Think of football (or hockey, or ...) camp for 8 year olds. Very few of those kids are going on to a brilliant professional sporting career. So we should shut them down, treat any parent who enrolls their child in such a camp with derision, etc. Right? No? Why not? -- Because sometimes the experience is more important then the result...I remember feeling the world change.
I have nothing to add. I just want to say thank you for a well-written spot-on post. Your analogy was good, and your inspiring "feeling the world change" description gave me goosebumps to read.
Put aside your trolling for a moment and understand the difference. One was a CHOICE, the other was FORCEFULLY FORCED DOWN YOUR THROAT. Can you figure out which one. You'll get cookie when you do.
Yeah, it's called parenting. My three year old hates broccoli but I force it down her throat. (well, I insist she takes minimum one bite). That's because I know it's for the best that she be exposed to it. She's likely not going to grow up a broccoli farmer or a vegan or a chef, but she should still be exposed to it.
I do most of my coding in my mind. It is vastly larger than a quad-head high density wall of pixels.
Whether I'm typing on a laptop or a big screen is mostly immaterial.
I thought "The Cloud" was just another term for someone else's computer. Back in my day, we called it timesharing. Long life TSS/8, TSO and Multics (and their friends).
Now you kids get off my lawn.
I think that misses the point. "The Cloud" refers to the commoditized outsourcing of certain aspects of maintaining a set of servers while keeping control of other aspects. Like: if you outsource the nothing, then it's a machine in your building that you bought/built yourself. If you outsource just rackspace, then it's an datacenter from the 90s. If you outsource hosting a VM, or scaleability, or distributed load balancing / replication, or worldwide redundancy, or keeping the host OS up-to-date with patches, or any kind of fancy bandwidth provision, or providing services like highly scalable data storage, then it's called The Cloud. Sure you could do those on Multics if you wanted (no one did).
Downloading eBooks (or anything else) isn't illegal. *Distributing* them is, without the proper permission/license.
Correction: *copying* them without the proper permission/license is what's illegal.
In the case of printed books, if you obtained an illegally-distributed book and read it, then you wouldn't be doing any copying.
But in the case of ebooks, if you obtained an illegally-distributed book and read it on your computer, the act of reading it on computer inevitably makes a copy in the computer's memory or cache or disk. The copyright act specifically says that copies this made this way are exempt from copyright restrictions, SO LONG AS these copies are (1) temporary, and transient/incidental, (2) an integral and essential part of a technological process, (3) the sole purpose of the copies is to enable a lawful use of the work.
Is it a "lawful use of the ebook" if you're reading a copy that had been illegally distributed? I think this is up for debate, and I'm not aware that it's been tested in court.
Is it a "temporary copy" if you download an illegally-distributed ebook and store it on your phone or hard disk? -- no.
http://www.legislation.gov.uk/...
Note: I'm not a lawyer.
The computer clarification was not to give carte blanche to view things you haven't paid for as long as you don't save it.
I think it's the exact reverse!
Traditionally, the "copyright infringers" were the ones who copied without license to do so -- who spun up the printing presses, who produced bootleg tapes. The act of consuming something was never an act of copyright infringement. You always had carte blanche to view anything, no matter how you obtained it (at least as regards copyright law).
The fact that computer technology meant that the act of consuming now also included the act of making a copy? This is an oddity, but one that the copyright industry grabbed hold of.
The computer clarification I think was a means to restore the producer/consumer balance.
(However, computers also allow "free consumption", which upsets the balance once again.)
An opinion like this is a shot across their bows, telling them "hey stop sending infringement notices to folks who merely download your movies because we think you're not going to prevail".
EDIT: merely *stream