Sure, they will make fun of it, probably give it poor reviews, but as long as 3DRealms manage to keep it fun, ironic and in bad taste it will still sell like hot cakes. If they play their cards right they could turn this into a stroke of genius.
Do you consider the frequent posting of short blurbs of information to a publically-accessible web site to be blogging? If so, then this would indeed qualify as blogging.
That's bullshit and you know it. If I make apache logs accessible from the web, does that make it a blog? It's definitely a log, and it's on the web. Of course, noone would call it a blog, except maybe ignorant journalists, or money-starved scientists trying to mooch off a current buzzword.
All that may be true... but does it all need to be on redundant hard disks and backed up off-site?
No, of course not, it's all realtively easy to replace. But, sacrificing 1/4th (or 1/5th with 5 disks) of capacity to be safe from a (single) HD failure sounds like a good trade-off. The only backup is a DVD copy tucked away somewhere.
As far as your Seinfeld (or porn) goes, 430GB will fit on a spindle of DVDs that costs $30, and gives you the advantage of portability and playability on your TV
Yes. But, three seconds to get to it on a RAID, versus a minute or so to find the right disk, put it in, wait for it to load, realise that the particular episode is on another disc, reload... As Jello Biafra said, Give Me Convenience or Give Me Death.:-)
Besides, the computer is connected to the TV via S-video cable, so I do watch it on the big screen. And, since the box is networked, it's portable and playable anywhere in the house (with GigE, every disk is local:-)).
Nobody is going to watch all that crap, and unless you have not got a life, you are included on that select group.
Well, it's a library. As in any library, you're probably not going to leaf through every book in it for quite some time. But, the point of the library is access. I want every Seinfeld show online. So the moment I want to watch that classic 'The sea was angry that day, my friends' moment, it's there, online and three clicks away. Or if I want to be sure of the exact quote to 'dead parrot' sketch. Or just pop in a random Red Dwarf episode, whenever I feel like it. Or put the entire Neil Young discography on and set it on shuffle, right now.
And that's why I'm building a RAID server.
BTW, I'm probably gonna go with HighPoint RR454 RAID card. It supports RAID-5 and has ATA133 interface. It may be dated, but it suppports up to 8 drives, versus SATA versions that support only 4 (for the same pricetag).
What you're proposing is essentially an evolutionary algorithm to arrive at a properly structured brain.
Basically, yes.
So be prepared to wait as long as it took for the brain to evolve in the first place.
It may take similar or larger number of generations, but the brain evolved on a liner scale, and computers evolve on an exponential scale. It's logical to expect that the exponential growth will catch up to linear relatively quickly.
It took millions of years for nature to come up with eyes. It took mere decades for engineers to come up with a light sensor.
But my point is that if a machine passes the Turing test, that doesn't *necessarily* mean anything about how the human mind works.
So? What are the consequences of this? You don't even know if my brain functions the same as yours, you just assume that, because my answers seem written by a person. What if it turned out that I am a program? Would that make my arguments any less valid? Or more?;-)
We probably won't ever be able to exactly duplicate the inner workings of a human brain. But, if our artificial brain behaves the same as a human brain in all significant aspects, why should we care?
Well, that works as long as you ignore the fact that the brain is differentiated and modular.
So are transistors in a chip. Some form ALU, some form MMU and so on... But, they are all exactly the same sort of device, same as neurons form all different parts of the brain.
But you still need to know 'okay, I need to arrange these cells in a series of columns so they can process different angled structures in my visual field', 'these cells need to be in sequence so I can have a working memory', etc.
No you don't. You just need to simulate a huge nubmer of neurons and allow them to randomly evolve their connections. Keep those that bring you closer towards a set goal. Run that simulation for a truly huge number of iterations, and chances are you'll end up with something functional.
Of course, this is grossly oversimplifying, but the basic idea is exactly that. So, again, we're down to just more computing power...
Whats even more ridiculous is the assumption that more computing power MUST lead to some sort of magic AI.
Ridiculous? But more computing power _does_ lead to intelligence. The difference between your brain and that of a mouse is just the number of neurons, or in other words, computing power. And you probably wouldn't call a mouse intelligent.
Intelligence is considered to be an emergent property, something you get by increasing quantitative, not qualitative properties. Simply put, pile up enough neurons in a brain, and you get intelligence.
From that follows that if it is possible to accurately model a single neuron (which I believe it is, or eventually will be), then the only thing standing between a computer and intelligence is computing power.
Furthermore, we don't know whether or not our face-recognition programs are using the same techniques as the mind...
We don't, but that is an implementation detail, and not relevant to the discussion. The point of the Turing test is to intentionally abstract implementation details. It stipulates that if a person cannot differentiate between a human and a computer through conversation, then said computer should be considered as intelligent as a person.
Would being intelligent as a person make a computer alive, conscious or actually a person himself is a whole other question to which Turing test does not attempt to provide an answer.
Hype or not, this is a camera that would satisfy all my needs as an amateur photographer. I don't need a true SLR, I don't need to exchange lenses. I want a good lens, a live preview and I want to be able to take a picture wherever there is enough light to see. For a thousand bucks, it doesn't get much better than this.
Don't bash it only because it's a Sony. On specs alone, it's an attractive package. Actual design could be better, especially the LCD placement. I hope that soon there will be a couple of alternatives to choose from, I'm definitely looking for a camera like this.
There are three basic types of digital sensors: Those with RGB sensors laid out in a rectangular grid; Foveon where they are stacked on top of each other; and Fuji's which has sensors which are offset from one another.
Whatever, apples and oranges. The point is the size of the sensor (and, consequently, light sensitivity), and live preview, not pixel arrangement.
Prehaps you should browse around dpreview.com to see just how ordinary that camera is. Try the side by side feature of the site.
I read the review, even before posting the comment. And, to quote from the site: The DSC-R1 is a first of a kind, the first all-in-one digital (fixed-lens digital) to utilize an APS-C size sensor. I still stand by my assertion, it is not perfect, but this camera is a breakthrough, and deserved an Slashdot article.
The author also laments that there's no macro mode, which is kind of redeundant when you've already said you can't get any closer than 13 inches. And all for $1000!
The limiting factor on any consumer camera I used was always the light sensitivity. I'd trade movie recording and 13" 'macro' for a decent ISO1600 capability. For me, that alone would extend the usability of the camera much more than being able to macro up to any distance or record a movie. The capability of taking night pictures (concerts, outdoors at night!) with sub-1s exposure is truly exciting.
Besides, lens like that for an dSLR would usually set you back more than $1000.
Sony has done nothing worth a headline here. This is pure PR - one of those planted "news" stories where some reporters got fed a story on a slow news day... maybe got sent a free camera with some marketing hype.
Geez... You really have no clue when it comes to digital cameras, right?
So, let me explain. This is a _significant_ new development in the field of consumer digital cameras, in no way an typical incremental evolution.
Its significance comes from a new type of CCD, a new development by Sony. Until now, you basically had two types of digital cameras, compact consumer cameras with tiny CCDs, and digital SLRs with huge CCDs. This is a first consumer camera with a huge CCD. They managed to marry the best of both worlds - excellent picture quality and light sensitivity (if that means anything to you, it goes up to ISO1600!) and a live preview on the cameras LCD screen. Truly a first. Expect all the other manufacturers to frantically scramble to catch up to Sony. This kind of cameras will rule the market in a few years. And, like it or not, Sony was there first.
Then again, if you read the FA you'd probably know as much. But, hey, let's not allow the facts to take the edge off of a fine groupthink induced rant.
Fine, but after spending your money on mods you'll have something very close to my Motorola wr850g that you can pick up on ebay for less than $30 and then flash with dd-wrt.
The major difference is the amount of RAM & CPU power. Sure, you can use dd-wrt for the basic stuff (I use one on my home network). But, could you squeeze L7 traffic shaping on that? How about snort? Or a dozen VPN connections? Squid?
They may be close in capabilities, but with IPCop you have no upper limits on what can you be done with it.
It makes noise and uses more power than a dedicated router.
Replace the hard disk with a Compact Flash to IDE adapter, use passive cooler for the CPU and basically you're set. You don't need P4 to run a firewall. Something like VIA's C3 will do. Setup like that doesn't draw much power, so passive PSU cooling becomes an option.
It doesn't do wireless.
Of course it does. The 'blue' interface is for wireless.
the difference being that Yahoo's version is web-based and can be accessed from any computer.
And that's why I carry my Opera RSS feeds on a USB stick (basically, you just need to move the Mail directory in Opera's app data dir). You need to have Opera installed on the machine to access it, but I rarely have the need for RSS feeds on other people's computers. As long as my laptop, my work PC and my home PC are synchronized, I'm fine.
What I can't figure out is how people are having problems figuring out BIONC.
The 'Classic' install went something like this: Double-click the exe. Next, next, next, finish. Enter your email address. Sit back and watch the pretty graphics, you're done.
Until BOINC SETI gets up to that level of user friendliness, I'm sure it can kiss most of its classic users goodbye. Average people just can't be expected to put up with the new install process. Would it really be impossible to make an installer that provides the old next-next-next experience?
No matter what I have done, I have never gotten sustained 48MB/s transfer from any IDE hard drive.
I get that when transfering GB-sized files from one disk to the other (SATAATA133), but that's far from typical transfers, like those when you're starting and application up. Slow random (track-to-track) HD read speeds ensure you don't get more then 10MB/s at best. And, if you by accident start up two big apps at once... go for a coffee break.
On the other hand, random access on flash is almost as fast as sequential access. And this is why this high-speed flash chips are so significant (not necessarily this particular model, but that generation of flash technology). If you install your OS on a flash drive that does 28MB/s on random reads, you could cut your boot and app startup times in half, which would be rather a big deal, at least for me.
Of course, there's always the issue of limited number of writes on a flash, but I guess that will get better over time, too.
Re:Did you read the article? This isn't praticle.
on
First Silicon Laser
·
· Score: 1
Why not focus research on creating something that's extremely effecient?
Well, because things usually get extremely efficient only _after_ you focus your research on them.
Slashdot Mirror
No you won't, because you clearly believe in property rights and that plan is not original and doesn't belong to you.
:-)
Oh, I believe in property rights. It's just that I don't mind stealing from the people who don't.
Who needs property rights?
Well, if you don't, then I'll stop by later to pick up your car keys.
Sure, they will make fun of it, probably give it poor reviews, but as long as 3DRealms manage to keep it fun, ironic and in bad taste it will still sell like hot cakes. If they play their cards right they could turn this into a stroke of genius.
Yeah. Just like Ion Storm did with Daikatana.
Do you consider the frequent posting of short blurbs of information to a publically-accessible web site to be blogging? If so, then this would indeed qualify as blogging.
That's bullshit and you know it. If I make apache logs accessible from the web, does that make it a blog? It's definitely a log, and it's on the web. Of course, noone would call it a blog, except maybe ignorant journalists, or money-starved scientists trying to mooch off a current buzzword.
All that may be true... but does it all need to be on redundant hard disks and backed up off-site?
:-)
:-)).
No, of course not, it's all realtively easy to replace. But, sacrificing 1/4th (or 1/5th with 5 disks) of capacity to be safe from a (single) HD failure sounds like a good trade-off. The only backup is a DVD copy tucked away somewhere.
As far as your Seinfeld (or porn) goes, 430GB will fit on a spindle of DVDs that costs $30, and gives you the advantage of portability and playability on your TV
Yes. But, three seconds to get to it on a RAID, versus a minute or so to find the right disk, put it in, wait for it to load, realise that the particular episode is on another disc, reload... As Jello Biafra said, Give Me Convenience or Give Me Death.
Besides, the computer is connected to the TV via S-video cable, so I do watch it on the big screen. And, since the box is networked, it's portable and playable anywhere in the house (with GigE, every disk is local
Nobody is going to watch all that crap, and unless you have not got a life, you are included on that select group.
Well, it's a library. As in any library, you're probably not going to leaf through every book in it for quite some time. But, the point of the library is access. I want every Seinfeld show online. So the moment I want to watch that classic 'The sea was angry that day, my friends' moment, it's there, online and three clicks away. Or if I want to be sure of the exact quote to 'dead parrot' sketch. Or just pop in a random Red Dwarf episode, whenever I feel like it. Or put the entire Neil Young discography on and set it on shuffle, right now.
And that's why I'm building a RAID server.
BTW, I'm probably gonna go with HighPoint RR454 RAID card. It supports RAID-5 and has ATA133 interface. It may be dated, but it suppports up to 8 drives, versus SATA versions that support only 4 (for the same pricetag).
What you're proposing is essentially an evolutionary algorithm to arrive at a properly structured brain.
Basically, yes.
So be prepared to wait as long as it took for the brain to evolve in the first place.
It may take similar or larger number of generations, but the brain evolved on a liner scale, and computers evolve on an exponential scale. It's logical to expect that the exponential growth will catch up to linear relatively quickly.
It took millions of years for nature to come up with eyes. It took mere decades for engineers to come up with a light sensor.
But my point is that if a machine passes the Turing test, that doesn't *necessarily* mean anything about how the human mind works.
;-)
So? What are the consequences of this? You don't even know if my brain functions the same as yours, you just assume that, because my answers seem written by a person. What if it turned out that I am a program? Would that make my arguments any less valid? Or more?
We probably won't ever be able to exactly duplicate the inner workings of a human brain. But, if our artificial brain behaves the same as a human brain in all significant aspects, why should we care?
Well, that works as long as you ignore the fact that the brain is differentiated and modular.
So are transistors in a chip. Some form ALU, some form MMU and so on... But, they are all exactly the same sort of device, same as neurons form all different parts of the brain.
But you still need to know 'okay, I need to arrange these cells in a series of columns so they can process different angled structures in my visual field', 'these cells need to be in sequence so I can have a working memory', etc.
No you don't. You just need to simulate a huge nubmer of neurons and allow them to randomly evolve their connections. Keep those that bring you closer towards a set goal. Run that simulation for a truly huge number of iterations, and chances are you'll end up with something functional.
Of course, this is grossly oversimplifying, but the basic idea is exactly that. So, again, we're down to just more computing power...
Whats even more ridiculous is the assumption that more computing power MUST lead to some sort of magic AI.
Ridiculous? But more computing power _does_ lead to intelligence. The difference between your brain and that of a mouse is just the number of neurons, or in other words, computing power. And you probably wouldn't call a mouse intelligent.
Intelligence is considered to be an emergent property, something you get by increasing quantitative, not qualitative properties. Simply put, pile up enough neurons in a brain, and you get intelligence.
From that follows that if it is possible to accurately model a single neuron (which I believe it is, or eventually will be), then the only thing standing between a computer and intelligence is computing power.
Furthermore, we don't know whether or not our face-recognition programs are using the same techniques as the mind...
We don't, but that is an implementation detail, and not relevant to the discussion. The point of the Turing test is to intentionally abstract implementation details. It stipulates that if a person cannot differentiate between a human and a computer through conversation, then said computer should be considered as intelligent as a person.
Would being intelligent as a person make a computer alive, conscious or actually a person himself is a whole other question to which Turing test does not attempt to provide an answer.
Dude, that was a knock-out. It's over, give it up.
Don't believe the hype.
Hype or not, this is a camera that would satisfy all my needs as an amateur photographer. I don't need a true SLR, I don't need to exchange lenses. I want a good lens, a live preview and I want to be able to take a picture wherever there is enough light to see. For a thousand bucks, it doesn't get much better than this.
Don't bash it only because it's a Sony. On specs alone, it's an attractive package. Actual design could be better, especially the LCD placement. I hope that soon there will be a couple of alternatives to choose from, I'm definitely looking for a camera like this.
There are three basic types of digital sensors: Those with RGB sensors laid out in a rectangular grid; Foveon where they are stacked on top of each other; and Fuji's which has sensors which are offset from one another.
Whatever, apples and oranges. The point is the size of the sensor (and, consequently, light sensitivity), and live preview, not pixel arrangement.
Prehaps you should browse around dpreview.com to see just how ordinary that camera is. Try the side by side feature of the site.
I read the review, even before posting the comment. And, to quote from the site: The DSC-R1 is a first of a kind, the first all-in-one digital (fixed-lens digital) to utilize an APS-C size sensor. I still stand by my assertion, it is not perfect, but this camera is a breakthrough, and deserved an Slashdot article.
The author also laments that there's no macro mode, which is kind of redeundant when you've already said you can't get any closer than 13 inches. And all for $1000!
The limiting factor on any consumer camera I used was always the light sensitivity. I'd trade movie recording and 13" 'macro' for a decent ISO1600 capability. For me, that alone would extend the usability of the camera much more than being able to macro up to any distance or record a movie. The capability of taking night pictures (concerts, outdoors at night!) with sub-1s exposure is truly exciting.
Besides, lens like that for an dSLR would usually set you back more than $1000.
Sony has done nothing worth a headline here. This is pure PR - one of those planted "news" stories where some reporters got fed a story on a slow news day... maybe got sent a free camera with some marketing hype.
Geez... You really have no clue when it comes to digital cameras, right?
So, let me explain. This is a _significant_ new development in the field of consumer digital cameras, in no way an typical incremental evolution.
Its significance comes from a new type of CCD, a new development by Sony. Until now, you basically had two types of digital cameras, compact consumer cameras with tiny CCDs, and digital SLRs with huge CCDs. This is a first consumer camera with a huge CCD. They managed to marry the best of both worlds - excellent picture quality and light sensitivity (if that means anything to you, it goes up to ISO1600!) and a live preview on the cameras LCD screen. Truly a first. Expect all the other manufacturers to frantically scramble to catch up to Sony. This kind of cameras will rule the market in a few years. And, like it or not, Sony was there first.
Then again, if you read the FA you'd probably know as much. But, hey, let's not allow the facts to take the edge off of a fine groupthink induced rant.
2.000.000 population, 5% internet penetration, 5.000 Ubuntu desktops -- instant 5% market share if you look at it one way.
Which way of looking would that be? Seriously, I just don't see it.
Fine, but after spending your money on mods you'll have something very close to my Motorola wr850g that you can pick up on ebay for less than $30 and then flash with dd-wrt.
The major difference is the amount of RAM & CPU power. Sure, you can use dd-wrt for the basic stuff (I use one on my home network). But, could you squeeze L7 traffic shaping on that? How about snort? Or a dozen VPN connections? Squid?
They may be close in capabilities, but with IPCop you have no upper limits on what can you be done with it.
It makes noise and uses more power than a dedicated router.
Replace the hard disk with a Compact Flash to IDE adapter, use passive cooler for the CPU and basically you're set. You don't need P4 to run a firewall. Something like VIA's C3 will do. Setup like that doesn't draw much power, so passive PSU cooling becomes an option.
It doesn't do wireless.
Of course it does. The 'blue' interface is for wireless.
the difference being that Yahoo's version is web-based and can be accessed from any computer.
And that's why I carry my Opera RSS feeds on a USB stick (basically, you just need to move the Mail directory in Opera's app data dir). You need to have Opera installed on the machine to access it, but I rarely have the need for RSS feeds on other people's computers. As long as my laptop, my work PC and my home PC are synchronized, I'm fine.
So technically, it could be used as a DoS attack on other browsers as well, though not nearly as badly as on IE.
:-)
Well, Opera just opened a small window which just sat there and did nothing. I closed it, and continued on my merry way. Score one for Opera.
What was Threshold?
You lucky bastard...
Anyone have Brent Spiner's address? He owes me 45 minutes of my life and some 200MB of bandwidth.
What I can't figure out is how people are having problems figuring out BIONC.
The 'Classic' install went something like this: Double-click the exe. Next, next, next, finish. Enter your email address. Sit back and watch the pretty graphics, you're done.
Until BOINC SETI gets up to that level of user friendliness, I'm sure it can kiss most of its classic users goodbye. Average people just can't be expected to put up with the new install process. Would it really be impossible to make an installer that provides the old next-next-next experience?
No matter what I have done, I have never gotten sustained 48MB/s transfer from any IDE hard drive.
I get that when transfering GB-sized files from one disk to the other (SATAATA133), but that's far from typical transfers, like those when you're starting and application up. Slow random (track-to-track) HD read speeds ensure you don't get more then 10MB/s at best. And, if you by accident start up two big apps at once... go for a coffee break.
On the other hand, random access on flash is almost as fast as sequential access. And this is why this high-speed flash chips are so significant (not necessarily this particular model, but that generation of flash technology). If you install your OS on a flash drive that does 28MB/s on random reads, you could cut your boot and app startup times in half, which would be rather a big deal, at least for me.
Of course, there's always the issue of limited number of writes on a flash, but I guess that will get better over time, too.
Why not focus research on creating something that's extremely effecient?
Well, because things usually get extremely efficient only _after_ you focus your research on them.