he said he could compress an entire movie down to 8 kilobytes.
To put that into perspective, the script for a movie is in the order of 200kB, even after a run through xz it is still around 32-64kB. So even if your decompressor has all the smarts of a Spielberg and could reconstruct the movie from just the script alone, you don't even have enough storage for the script itself. And of course there is no information in the script that would be detailed enough to reconstruct that exact movie, so you would just end up with a remake anyway.
I think compression could still be improved quite a lot when one focuses on just producing something that looks plausible instead of something that is an accurate reproduction of the original. But even in the best-case scenario 8kB is not nearly enough.
The only way this could be possible is if you were literally God and could predict and enumerate all the movies that will ever be filmed. An 8kB number should be more than enough to identifier a movies by a unique id. Predicting all possible movies would however get a little tricky.
Some parts of the Internets infrastructure are decentralized, the network on the whole however heavily depends on central authorities. If you want an IP address, you have to ask a central authority. You want a DNS name, ask a central authority. Even when you want to connect your laptop to your phone you may need to ask your phone company for permission.
With the Web it gets even worse, since almost everything people do on there is hosted on some big Google or Facebook server farm. In theory everybody could run their own Web and mail servers, but in reality everybody depends on centralized providers. Another problem with the Web is that all the addressing is location based, not content based. So you can't refer to a specific document, you can only refer to a storage location and whatever is stored there can change.
There are attempts to build a decentralized content addressable network on top of the Internet, like Freenet, Maidsafe or IPFS, but so far none of them has reached critical mass. On the raw network side mesh networks are slowly gaining some traction and they might help with the whole need for an ISP, you just connect to the nearest phone or computer and hop on from there.
If DRM goes into HTML or not won't make much of a difference, if anything it will open things up a little since you will be able to use your browser more often to watch video instead of a proprietary App. It's not like the FSF is wrong on being against DRM, but it has little chance of success here.
In general I am not really happy with much of what the FSF has been doing lately. Their intentions are noble, but when you want to actually change things you can't just say "No", you have to provide a realistic alternative. Back when Windows95/98/ME was the 'threat', GNU/Linux provided a good alternative. But with phones, tablet, web services, cloud computing and such they feel hopelessly behind.
100 words per minutes might be a stretch, but it doesn't sound all that impossible given that the speed record was set with hunt&peck typing by moving a cursor across the screen. Some fancy machine learning that could guess whole words at a time or something along the line should have no problem beating that by quite a margin. It wouldn't even need to be perfect, just close enough, to give a drastic speed up (i.e. like Tab-completion).
See this earlier work that guessed video sequences from brain activity. Getting information in a more 'holistic' form out of the brain instead by just cursor movement seems plausible.
Any old 8 or 16-bit software from decades past, if we have any of that software around today, it still works.
You are kind of ignoring the gap that existed back then between computer architectures. All your C64 programs wouldn't work on an Amiga. You couldn't read the data that you saved on 5.25" floppies in your 3.5" drive either. Each new computer generation essentially meant that you had to start all your computing from scratch, neither programs nor data could be carried over. The easy data transfer via USB or the Internet just didn't exist back then.
It took decades until we had working emulation and data formats and media that you could make work between different computer architectures.
Also if you just stick to the same hardware and OS, your software will still keep running even in the modern day. Windows XP might no longer be supported, but it runs just as fine as an old Workbench on an Amiga. There is some software that wants to be online activated, but patches for that exist in most cases and it's not that big of a deal unless you try to play a MMORPG where most of the computation happens server side.
The loss of control over personal computing and web browsing especially is completely self inflicted in the name of ease-of-use. If Mozilla wanted, they could have build a Freedom browser, but instead they build a crappy Chrome clone and to get a little bit of freedom back you have to install all kinds of third party addons (e.g. even basic things like saving video). It's mind boggling how featureless modern browsers are by default.
As for the DRM, it's tricky. If there is no standard, you either get no video or you get a proprietary plugin. The lack of a standard doesn't make DRM go away and companies have to problem breaking standards to squeeze DRM in there. I don't like DRM being a standard, but I don't think it will make things worse than they already are. On the plus side, if there is a standard it might be easier to crack.
People already use Twitter for long messages all the time, but due to the character limit they have to do it by posting images of text or other ugly workarounds. If Twitter wants to give priority to short messages, that's fine, but forcing people into workarounds is just stupid. They had the same thing already with photos, people used TwitPic and other workarounds for a long while until Twitter finally started to allow images. It hasn't killed Twitter, quite the opposite, it made it more versatile and useful. I would expect much the same happening with long text support. Not supporting more than 140 characters isn't a feature, it's a leftover from it's SMS beginnings.
Flash isn't dead yet. While most mobile webpages no longer use it, on the desktop you still see it pretty frequently. As for its impending death, that has been a long while coming:
* using Flash to design a whole website became mostly unnecessary due to HTML/CSS becoming more powerful
* using Flash for vector animation became replaced by regular video and Youtube
* using Flash as video player became unnecessary due to HTML gaining a <video> tag
I am not quite sure what happened with Flash and gaming, Newgrounds.com is still around, but you rarely hear about it anymore. Doing games in HTML with <canvas> and WebGL is now possible as well, but I don't really see those very often. I assume Unity and mobile gaming took mostly over what was once done in Flash.
However what really killed Flash was Adobe no longer supporting it. When software is full of security and performance issues, it's no surprise that people will move away from it. Flash got popular in the first place because it did things that your browser wouldn't be able to do on it's own. But while browsers got more powerful, Flash just sat there and didn't really improve much at all.
Well, yeah, without sound and with one 1fps or less. Meanwhile Steam can pump a game at 1080p@60 over the network without much problem, sound included.
Even ignoring the slowness when it comes to fast moving content. It's missing a lot of fundamental features, such as the ability to move apps between devices or screen sharing. You have to stop an app and restart it to move to another device. That you have to pipe the protocol through SSH if you want a bit of security also makes it more complicated to use than it's needs to be.
Unix is build with a one-system:many-users mindset. That was a great idea 30 years ago. Today's reality however is the other way around many-systems:single-user, everybody has a tablet, a smartphone and a PC, sometimes multiple of each. Unix provides nothing to deal with that. You can try to export your home directory via NFS, but that falls apart the moment you have the same app running on two different computer and both want to access the same file. Some programs solve the problem at the application level, i.e. bookmark syncing in Chrome, but that doesn't scale to the rest of the system.
Once upon a time X11's network transparency was its claim to fame and it did provide a bit of a solution to the many-systems:single-user problem. But today it's close to useless. Ever tried to stream a video over X11's network connection? Doesn't work. The protocol just can't deal with today's workloads. Proprietary alternatives exist that can handle this much better.
In terms of security the whole idea of giving every app that a user runs access to everything that the user can access is also foolish. But at least there is a bit of hope in fixing that. Android solves that by running each app under a different user. Ubuntu tries to solve that by sandboxing. It's all still far away from being the default way you run apps on a desktop Linux, but at least people have recognized the problem.
The big issue I have with HTML is that it's useless for publishing larger content, like books or even just multi-page articles. Thanks to hyper links it is of course possible to add some Next/Prev buttons to a webpage to represent such content, but those links are just hacks, not markup. eReader have developed their own formats (.ePub,.mobi, iBook, etc.) for accomplishing this task, but while they often are little more than a.zip with.html files, none of them are proper part of the Web and your regular web browser won't read them.
HTML had <link rel="next/prev"...> markup going back to HTML2.0, but it was never properly supported by any browser or developed into something that would be powerful enough to replace.ePub and Co. This to me is one of the big failures of the Web that nobody really talks about. The Web should be the place where you publish content, it should be the replacement of paper, but instead people are forced to use.ePub or.PDF for that task, as plain old HTML isn't doing the job.
The other elephant in the room are of course the hyper links. The Web still lacks any kind of content-addressability, it's all location based, thus when server go down or it's URL layout changes, all your hyperlinks break. Basic tasks like linking to a specific paragraph from another article are also not possible with HTML. Project Xanadu never got much traction, but it's really time for the Web to learn a thing or two from what they tried to accomplish back then.
I don't get why you think taking a portion of a GPL-licensed work means that portion wouldn't itself be GPL-licensed. Of course it is, like the whole work is.
The stance of the FSF is that game assets are not software and thus having a GPLed engine bundled with non-free assets is ok. The assets and the engine doesn't form a complete work that as whole has to be GPL licensed. Going by that interpretation taking the assets from a GPL game and including them in a proprietary work would be ok, as it's just the reverse direction of what is already common practice (see Doom, Quake, Dosbox, ScummVM). The GPL would apply to the art, but not to the engine.
The license follows the graphics no matter how you get them.
That's not the issue, the graphics will obviously stay GPL, the issue is if it does force other parts of the program to fall under the GPL. In a statically linked C project the issue is clear because all the used library become part of the resulting executable. In a lot of other context the coupling between the GPL code and the proprietary code is not so tight. Using a GPLed grep in a proprietary shell script is for example considered ok, but using a GPLed grep() function loaded from a dynamic library would be considered a violation going by the common interpretation, even so it's essentially the thing. The proprietary code wouldn't need to contain any GPLed pieces, it would just call them. In the case of graphics the only thing that couples the code with the assets would be a filename.
Things get even more complicated when you are dealing indirect couplings by third parties, not by the distributor of the proprietary application. For example assume a proprietary application provides support for themes and then a user builds a theme from GPLed assets. Is that a violation or not? What about plugins or scripts?
This whole situation really isn't clear and even the official GPL FAQ is a little vague when it comes to how to deal with plugins and basically leaves the user with a "it depends".
If you use GPL code, you take on that license. It's really that simple.
The only area where the GPL is clear is statically linked C code, everything else is very open to interpretation. Take for example a GPLed game. Now somebody comes along, takes all the graphics and releases a proprietary game with them. Is that a violation or not? What if he doesn't actually include the graphics, but loads them from the web or straight out of your git repository or their fork of your git repository? What if it's not graphics, but interpreted code code? At what point do the assets require the rest of the code released under GPL? You can go with "never", "always" or "up to a judge's interpretation".
Even the FSF can't manage to get a conflict free interpretation out of it. On one side, if you use one of their GPLed libraries, your main app shall be released as GPL as well. But on the other side they want the right to clone proprietary APIs without adhering to the license. So what is it? Shall APIs have copyright or not? The FSF wants it both ways.
Did Youtube demand money from her in exchange for fire, theft, and kneecap insurance?
Yes, that's one of the main points of her open letter. Youtube has a system in place, Content ID, to stop piracy and it works quite well. The crux is that they only allow it's use to musicians who have agreed to license their content to them or at least that's assumed, as they don't publish any rules. Everybody else gets left in the dust and isn't allowed into Content ID and thus their content can be shared on Youtube without permission. Which according to her argument violates the requirements for "Safe Harbor" protection and makes Youtube guilty of mass copyright infringement, as that "Safe Harbor" law requires technical measures to be made available to everybody.
I think before we see any large social changes we will see a lot of laws and regulation put into action against such technology. Social networks for example can lock away access to peoples photos behind a login and when logins are only given out to people who have verified their identity spidering becomes much harder. That's not even far future, many services already require a mobile phone number for id and some countries don't give you a mobile phone number anonymously. Collecting the data will still happen, but it will be much more troublesome and could be made illegal on top. So whenever somebody would offer a public search service, they could be shutdown relatively fast (unless Bitcoin and Tor make it commercially viable to put in the effort).
Given the advances in computer graphics I could also imagine that social networks will get flooded with lots of fake data. Face swap yourself into all kinds of photos and it will become much harder to find who you really are, since lots of information about you online will be fake. So maybe we end up with a general distrust about data we found online about a person.
I'd blame the OS instead. Giving each process full access to the system just isn't a good way to do things and constantly leads to problems like this. Python can stop some those problems, but it provides by no means a secure sandbox. If you access the filesystem in Python, you still have full access to the filesystem. In cases such as this the process should be limited to exactly the data it needs to get the job done, meaning an input image, an output location and a bunch of configuration parameter.
The lack of a packaging format for third party apps has been one of the biggest and most persistent problems in the Linux landscape for ages. I have no idea of the quality of the work Ubuntu is doing here and it does seem to duplicate the work going on with xdg-app, but I really wouldn't mind getting rid of tarballs and shelf extracting shell scripts. The monolithic dependency trees that package managers require at the moment just don't scale and never provided a good way for third party apps to plug into them (just adding random third party repos is inherently fragile and insecure).
I have done that, but the update didn't carry over the applications from Windows7 for some reason, it presented a mostly blank install. The only two application it did carry over where some old copies of Word and Excel, both of which failed to run in Windows10. The rollback to Windows7 however seems to have worked. Given how aggressively Microsoft is pushing Windows10 update I would have expected a better tested upgrade routine.
Yep, that is my experience as well. When you click the "download and install later" option, that's it, the update will now be carried out and you have no way to cancel it. The dialog box that is presented to you before the final update does not have a cancel button or a close button or any other means to not carry the installation out, you can delay the installation by some days, but you have to set a date for the install, there is no "ask me later".
I don't think modern games would be a good choice for an AI training, as most modern games are extremely simplistic and build in such a way that the player can hardly fail at all. You have endless respawns, navigation markers and all that stuff to help you. They often also have level up mechanics that could be exploited by an AI. Old games like Doom and Quake seem to be a much better fit, as in those you have to actually navigate on your own instead of just following a magic quest marker. Those games also tend to have direct player control instead of the fly-by-wire you have in Assassins Creed where the character walks on his own and player input is just a lose suggestion for where he should go.
Price per GB has resumed dropping [cbsistatic.com] since the effect of the Thailand flooding and HDD consolidation in 2011-2012.
The most frustrating part isn't so much the price per gigabyte, as that is back to pre-flood levels, but that you only get that good GB/$ rate in the $100 price range, while you used to get that in the $50 range. If you only need 1 or 2TB you are still paying quite a bit more then five years ago, it's only with 3TB-8TB drives that you get a slightly better rate then pre-flood.
Back in 2011 I could get a 1.5TB drive for 45€, now five years later the best I can get is 3TB for 90€. Double the storage for double the price. If I just want to spend 50€ I only get 1TB. It's nice that we now have 6TB and 8TB drives, but they aren't cheap and so far haven't really lowered the price of the smaller drives and given how long this has already taken I am not even sure if HDDs will ever get cheap again before SSDs will take that space.
Slashdot Mirror
he said he could compress an entire movie down to 8 kilobytes.
To put that into perspective, the script for a movie is in the order of 200kB, even after a run through xz it is still around 32-64kB. So even if your decompressor has all the smarts of a Spielberg and could reconstruct the movie from just the script alone, you don't even have enough storage for the script itself. And of course there is no information in the script that would be detailed enough to reconstruct that exact movie, so you would just end up with a remake anyway.
I think compression could still be improved quite a lot when one focuses on just producing something that looks plausible instead of something that is an accurate reproduction of the original. But even in the best-case scenario 8kB is not nearly enough.
The only way this could be possible is if you were literally God and could predict and enumerate all the movies that will ever be filmed. An 8kB number should be more than enough to identifier a movies by a unique id. Predicting all possible movies would however get a little tricky.
Some parts of the Internets infrastructure are decentralized, the network on the whole however heavily depends on central authorities. If you want an IP address, you have to ask a central authority. You want a DNS name, ask a central authority. Even when you want to connect your laptop to your phone you may need to ask your phone company for permission.
With the Web it gets even worse, since almost everything people do on there is hosted on some big Google or Facebook server farm. In theory everybody could run their own Web and mail servers, but in reality everybody depends on centralized providers. Another problem with the Web is that all the addressing is location based, not content based. So you can't refer to a specific document, you can only refer to a storage location and whatever is stored there can change.
There are attempts to build a decentralized content addressable network on top of the Internet, like Freenet, Maidsafe or IPFS, but so far none of them has reached critical mass. On the raw network side mesh networks are slowly gaining some traction and they might help with the whole need for an ISP, you just connect to the nearest phone or computer and hop on from there.
If DRM goes into HTML or not won't make much of a difference, if anything it will open things up a little since you will be able to use your browser more often to watch video instead of a proprietary App. It's not like the FSF is wrong on being against DRM, but it has little chance of success here.
In general I am not really happy with much of what the FSF has been doing lately. Their intentions are noble, but when you want to actually change things you can't just say "No", you have to provide a realistic alternative. Back when Windows95/98/ME was the 'threat', GNU/Linux provided a good alternative. But with phones, tablet, web services, cloud computing and such they feel hopelessly behind.
100 words per minutes might be a stretch, but it doesn't sound all that impossible given that the speed record was set with hunt&peck typing by moving a cursor across the screen. Some fancy machine learning that could guess whole words at a time or something along the line should have no problem beating that by quite a margin. It wouldn't even need to be perfect, just close enough, to give a drastic speed up (i.e. like Tab-completion).
See this earlier work that guessed video sequences from brain activity. Getting information in a more 'holistic' form out of the brain instead by just cursor movement seems plausible.
Any old 8 or 16-bit software from decades past, if we have any of that software around today, it still works.
You are kind of ignoring the gap that existed back then between computer architectures. All your C64 programs wouldn't work on an Amiga. You couldn't read the data that you saved on 5.25" floppies in your 3.5" drive either. Each new computer generation essentially meant that you had to start all your computing from scratch, neither programs nor data could be carried over. The easy data transfer via USB or the Internet just didn't exist back then.
It took decades until we had working emulation and data formats and media that you could make work between different computer architectures.
Also if you just stick to the same hardware and OS, your software will still keep running even in the modern day. Windows XP might no longer be supported, but it runs just as fine as an old Workbench on an Amiga. There is some software that wants to be online activated, but patches for that exist in most cases and it's not that big of a deal unless you try to play a MMORPG where most of the computation happens server side.
The loss of control over personal computing and web browsing especially is completely self inflicted in the name of ease-of-use. If Mozilla wanted, they could have build a Freedom browser, but instead they build a crappy Chrome clone and to get a little bit of freedom back you have to install all kinds of third party addons (e.g. even basic things like saving video). It's mind boggling how featureless modern browsers are by default.
As for the DRM, it's tricky. If there is no standard, you either get no video or you get a proprietary plugin. The lack of a standard doesn't make DRM go away and companies have to problem breaking standards to squeeze DRM in there. I don't like DRM being a standard, but I don't think it will make things worse than they already are. On the plus side, if there is a standard it might be easier to crack.
People already use Twitter for long messages all the time, but due to the character limit they have to do it by posting images of text or other ugly workarounds. If Twitter wants to give priority to short messages, that's fine, but forcing people into workarounds is just stupid. They had the same thing already with photos, people used TwitPic and other workarounds for a long while until Twitter finally started to allow images. It hasn't killed Twitter, quite the opposite, it made it more versatile and useful. I would expect much the same happening with long text support. Not supporting more than 140 characters isn't a feature, it's a leftover from it's SMS beginnings.
Flash isn't dead yet. While most mobile webpages no longer use it, on the desktop you still see it pretty frequently. As for its impending death, that has been a long while coming:
* using Flash to design a whole website became mostly unnecessary due to HTML/CSS becoming more powerful
* using Flash for vector animation became replaced by regular video and Youtube
* using Flash as video player became unnecessary due to HTML gaining a <video> tag
I am not quite sure what happened with Flash and gaming, Newgrounds.com is still around, but you rarely hear about it anymore. Doing games in HTML with <canvas> and WebGL is now possible as well, but I don't really see those very often. I assume Unity and mobile gaming took mostly over what was once done in Flash.
However what really killed Flash was Adobe no longer supporting it. When software is full of security and performance issues, it's no surprise that people will move away from it. Flash got popular in the first place because it did things that your browser wouldn't be able to do on it's own. But while browsers got more powerful, Flash just sat there and didn't really improve much at all.
It already began back in 1979, humans getting killed by robots is nothing new.
Yes it does.
Well, yeah, without sound and with one 1fps or less. Meanwhile Steam can pump a game at 1080p@60 over the network without much problem, sound included.
Even ignoring the slowness when it comes to fast moving content. It's missing a lot of fundamental features, such as the ability to move apps between devices or screen sharing. You have to stop an app and restart it to move to another device. That you have to pipe the protocol through SSH if you want a bit of security also makes it more complicated to use than it's needs to be.
Unix is build with a one-system:many-users mindset. That was a great idea 30 years ago. Today's reality however is the other way around many-systems:single-user, everybody has a tablet, a smartphone and a PC, sometimes multiple of each. Unix provides nothing to deal with that. You can try to export your home directory via NFS, but that falls apart the moment you have the same app running on two different computer and both want to access the same file. Some programs solve the problem at the application level, i.e. bookmark syncing in Chrome, but that doesn't scale to the rest of the system.
Once upon a time X11's network transparency was its claim to fame and it did provide a bit of a solution to the many-systems:single-user problem. But today it's close to useless. Ever tried to stream a video over X11's network connection? Doesn't work. The protocol just can't deal with today's workloads. Proprietary alternatives exist that can handle this much better.
In terms of security the whole idea of giving every app that a user runs access to everything that the user can access is also foolish. But at least there is a bit of hope in fixing that. Android solves that by running each app under a different user. Ubuntu tries to solve that by sandboxing. It's all still far away from being the default way you run apps on a desktop Linux, but at least people have recognized the problem.
The big issue I have with HTML is that it's useless for publishing larger content, like books or even just multi-page articles. Thanks to hyper links it is of course possible to add some Next/Prev buttons to a webpage to represent such content, but those links are just hacks, not markup. eReader have developed their own formats (.ePub, .mobi, iBook, etc.) for accomplishing this task, but while they often are little more than a .zip with .html files, none of them are proper part of the Web and your regular web browser won't read them.
HTML had <link rel="next/prev"...> markup going back to HTML2.0, but it was never properly supported by any browser or developed into something that would be powerful enough to replace .ePub and Co. This to me is one of the big failures of the Web that nobody really talks about. The Web should be the place where you publish content, it should be the replacement of paper, but instead people are forced to use .ePub or .PDF for that task, as plain old HTML isn't doing the job.
The other elephant in the room are of course the hyper links. The Web still lacks any kind of content-addressability, it's all location based, thus when server go down or it's URL layout changes, all your hyperlinks break. Basic tasks like linking to a specific paragraph from another article are also not possible with HTML. Project Xanadu never got much traction, but it's really time for the Web to learn a thing or two from what they tried to accomplish back then.
I don't get why you think taking a portion of a GPL-licensed work means that portion wouldn't itself be GPL-licensed. Of course it is, like the whole work is.
The stance of the FSF is that game assets are not software and thus having a GPLed engine bundled with non-free assets is ok. The assets and the engine doesn't form a complete work that as whole has to be GPL licensed. Going by that interpretation taking the assets from a GPL game and including them in a proprietary work would be ok, as it's just the reverse direction of what is already common practice (see Doom, Quake, Dosbox, ScummVM). The GPL would apply to the art, but not to the engine.
The license follows the graphics no matter how you get them.
That's not the issue, the graphics will obviously stay GPL, the issue is if it does force other parts of the program to fall under the GPL. In a statically linked C project the issue is clear because all the used library become part of the resulting executable. In a lot of other context the coupling between the GPL code and the proprietary code is not so tight. Using a GPLed grep in a proprietary shell script is for example considered ok, but using a GPLed grep() function loaded from a dynamic library would be considered a violation going by the common interpretation, even so it's essentially the thing. The proprietary code wouldn't need to contain any GPLed pieces, it would just call them. In the case of graphics the only thing that couples the code with the assets would be a filename.
Things get even more complicated when you are dealing indirect couplings by third parties, not by the distributor of the proprietary application. For example assume a proprietary application provides support for themes and then a user builds a theme from GPLed assets. Is that a violation or not? What about plugins or scripts?
This whole situation really isn't clear and even the official GPL FAQ is a little vague when it comes to how to deal with plugins and basically leaves the user with a "it depends".
If you use GPL code, you take on that license. It's really that simple.
The only area where the GPL is clear is statically linked C code, everything else is very open to interpretation. Take for example a GPLed game. Now somebody comes along, takes all the graphics and releases a proprietary game with them. Is that a violation or not? What if he doesn't actually include the graphics, but loads them from the web or straight out of your git repository or their fork of your git repository? What if it's not graphics, but interpreted code code? At what point do the assets require the rest of the code released under GPL? You can go with "never", "always" or "up to a judge's interpretation".
Even the FSF can't manage to get a conflict free interpretation out of it. On one side, if you use one of their GPLed libraries, your main app shall be released as GPL as well. But on the other side they want the right to clone proprietary APIs without adhering to the license. So what is it? Shall APIs have copyright or not? The FSF wants it both ways.
Seen John Carmack's recent tweet?
Did Youtube demand money from her in exchange for fire, theft, and kneecap insurance?
Yes, that's one of the main points of her open letter. Youtube has a system in place, Content ID, to stop piracy and it works quite well. The crux is that they only allow it's use to musicians who have agreed to license their content to them or at least that's assumed, as they don't publish any rules. Everybody else gets left in the dust and isn't allowed into Content ID and thus their content can be shared on Youtube without permission. Which according to her argument violates the requirements for "Safe Harbor" protection and makes Youtube guilty of mass copyright infringement, as that "Safe Harbor" law requires technical measures to be made available to everybody.
I think before we see any large social changes we will see a lot of laws and regulation put into action against such technology. Social networks for example can lock away access to peoples photos behind a login and when logins are only given out to people who have verified their identity spidering becomes much harder. That's not even far future, many services already require a mobile phone number for id and some countries don't give you a mobile phone number anonymously. Collecting the data will still happen, but it will be much more troublesome and could be made illegal on top. So whenever somebody would offer a public search service, they could be shutdown relatively fast (unless Bitcoin and Tor make it commercially viable to put in the effort).
Given the advances in computer graphics I could also imagine that social networks will get flooded with lots of fake data. Face swap yourself into all kinds of photos and it will become much harder to find who you really are, since lots of information about you online will be fake. So maybe we end up with a general distrust about data we found online about a person.
I'd blame the OS instead. Giving each process full access to the system just isn't a good way to do things and constantly leads to problems like this. Python can stop some those problems, but it provides by no means a secure sandbox. If you access the filesystem in Python, you still have full access to the filesystem. In cases such as this the process should be limited to exactly the data it needs to get the job done, meaning an input image, an output location and a bunch of configuration parameter.
The lack of a packaging format for third party apps has been one of the biggest and most persistent problems in the Linux landscape for ages. I have no idea of the quality of the work Ubuntu is doing here and it does seem to duplicate the work going on with xdg-app, but I really wouldn't mind getting rid of tarballs and shelf extracting shell scripts. The monolithic dependency trees that package managers require at the moment just don't scale and never provided a good way for third party apps to plug into them (just adding random third party repos is inherently fragile and insecure).
I have done that, but the update didn't carry over the applications from Windows7 for some reason, it presented a mostly blank install. The only two application it did carry over where some old copies of Word and Excel, both of which failed to run in Windows10. The rollback to Windows7 however seems to have worked. Given how aggressively Microsoft is pushing Windows10 update I would have expected a better tested upgrade routine.
Yep, that is my experience as well. When you click the "download and install later" option, that's it, the update will now be carried out and you have no way to cancel it. The dialog box that is presented to you before the final update does not have a cancel button or a close button or any other means to not carry the installation out, you can delay the installation by some days, but you have to set a date for the install, there is no "ask me later".
I don't think modern games would be a good choice for an AI training, as most modern games are extremely simplistic and build in such a way that the player can hardly fail at all. You have endless respawns, navigation markers and all that stuff to help you. They often also have level up mechanics that could be exploited by an AI. Old games like Doom and Quake seem to be a much better fit, as in those you have to actually navigate on your own instead of just following a magic quest marker. Those games also tend to have direct player control instead of the fly-by-wire you have in Assassins Creed where the character walks on his own and player input is just a lose suggestion for where he should go.
Price per GB has resumed dropping [cbsistatic.com] since the effect of the Thailand flooding and HDD consolidation in 2011-2012.
The most frustrating part isn't so much the price per gigabyte, as that is back to pre-flood levels, but that you only get that good GB/$ rate in the $100 price range, while you used to get that in the $50 range. If you only need 1 or 2TB you are still paying quite a bit more then five years ago, it's only with 3TB-8TB drives that you get a slightly better rate then pre-flood.
Back in 2011 I could get a 1.5TB drive for 45€, now five years later the best I can get is 3TB for 90€. Double the storage for double the price. If I just want to spend 50€ I only get 1TB. It's nice that we now have 6TB and 8TB drives, but they aren't cheap and so far haven't really lowered the price of the smaller drives and given how long this has already taken I am not even sure if HDDs will ever get cheap again before SSDs will take that space.
Pebble uses an e-paper display, not an e-ink display, and e-paper is a polarisier-free LCD.