If Facebook ends up close to $38 at the end of the day, it will be a rare example of the stock having been priced correctly at the start.
This assumes that the stock could be sold for less than $38. I'm not sure that's even possible in an IPO scenario like this (not saying it isn't, saying I don't know).
Who would buy the stock for $38-$42 then turn around and sell it the same day for less?
There is no doubt - for most of what IBM is selling, APAC and the Middle East are the growth markets. So it does make sense to move centers of business from the country with low single digit growth and a saturated market to the developing areas with strong growth and lots of greenspace
On the other hand, there's clearly a lot of outsourcing fever going on, and IBM may be participating in that but couching it in the "International" part of its namesake. It's hard to say.
Your view on democracy sounds great in theory, but American-style democracy just doesn't work like that.
Here's an example. Georgia just changed its law about selling beer and wine at the grocery store on Sundays. Previously, it had been illegal.. for decades.
Under new law, cities and counties could hold referendums to allow/disallow sunday sales in their area. In almost all of the cities/counties that held a vote, Sunday sales won in a landslide.. typically 70+% approving.
Now, this new law nearly died in the 2011 state congress thanks to special interest groups like the Christian Coalition. In fact, Sunday sales were kept illegal for decades due to their efforts, even though it is clearly what the majority wanted. This tells me the Christian Coalition and groups like it successfully suppressed democracy for DECADES, against the wishes of a 70+% majority!
I'd say the timing has to do with the Ceglia case.
With a recent judicial decision to sanction Ceglia, Facebook most likely believes they are in the endgame of that suit and can move forward with an IPO without the suit causing problems.
Think about it. Would you file for a $75B IPO (or whatever bullshit figure they've come up with) if there was an open suit with someone claiming they own half of the founder's shares, and backing it up with a signed contract?
I pretty much agree with your conclusions, but have a couple quibbles.
Cyber warfare is also called a denial of service attack, and is fundamentally different from cyber espionage.
Cyber warfare is not anonymous, done by a large number of IP addresses, and can't be defended against.
You're right that the purpose of an attack is to disrupt the target machines, but it won't always be a DDOS attack. Stuxnet was a worm with a targeted payload; if the German researcher hadn't found it and gone public, Iran might not have ever figured out why their centrifuges weren't working as anticipated.
But Stuxnet doesn't fit neatly into cyberwarfare. I'd say it fits better into a third category: Covert Action, which is, by design, deniable.
But, yeah, I totally agree that this is unnecessary saber rattling.
Kudos to wiredmikey (and the ed?) for capturing that attribution of an attack is the key sticking point for military response.
Attributing attacks in a packet switched network like the Internet is just a fantasy.. Sure, you can trace an attack back to, say, China, but how do you know the attack originated there? You don't, unless China cooperates and gives your forensics experts access to their networks. Which probably will not happen.
So the hawks want to shore up some credibility for attribution. Here is the plan, from the linked DoD PDF:
This research focuses on two primary areas: developing new ways to trace the physical source of an attack, and seeking to assess the identity of the attacker via behavior-based algorithms.
Nice try Pentagon, but statistically-powered voodoo does not overcome the problem here: that the attacking machines could be controlled from anywhere, possibly even through teh 7 proxies. Lulz.
Maybe we should listen to the National Research Council when they write "deterrence of cyberattacks by the threat of in-kind response has limited applicability." (NRC Report, p.5)
I'll close with a suggestion: why not, instead of focusing on how and when we get to launch attacks, focus on bettering our defenses?
A lot of the machine learning algorithms we use today are based on statistical or classification techniques that are mathematically connected to neural networks, and their development has in part been inspired by them.
If you are saying these techniques were borne from mathematical properties of biological neural networks, you are just wrong. Get a ML textbook - it's all about curve fitting, probability theory, decision theory, information theory, statistics, optimization.
Many of our machine vision and hearing algorithms are based on phenomenon that have been observed in the brain's visual and auditory cortex. The differences of Gaussians in SIFT, or the wavelets in SURF for example.
Wrong again. You're zero for two. Actually, if you can cite a biology paper concluding cortex uses wavelets, I'll give you this one. Good luck.
Have we got a machine that wakes up one day, says hello and asks for a cheeseburger? No, of course not. That's kind of the end goal, isn't it?
No, that's not the goal of ML or AI, and that has nothing to do with anything I've written. Quit with the hyperbole already.. If you can't back up your dumb ass opinions, just STFU.
... claiming that progress is not due at all to studying how the brain does...
Thanks for the link, but it's still a pretty simple neural model. Just not as simple as many other common models, which is why they take great care to call it "biologically inspired." But, the focus of the research is on simulation, not intelligence.
To the original point, the researchers have simulated a better approximation of NNs without shedding any light on the "computational tricks" that make brains so smart. While the paper makes clear that this is a model that can be used to test neural theories, the goal of the project was to make a large scale simulation more feasible. The researchers did not claim to make better learning algorithms.
I disagree that we "have had a fair bit of success copying" the success of brains. Most machine learning these days (computer vision, etc.) does not even use neural models, and really reflects the amount of engineering effort and math that has gone into solving the problems. Some of the neural simulators I've read about have helped understand neural physics and chemistry, but not necessarily intelligence. We do, of course, have ANNs that accomplish very specific recognition or control tasks, but we do not have ANNs that, say, extract auditory features and build mental models of sound patterns.
And no, I am not the original AC. I generally agree with the last paragraph of that post (that we should look for what tricks make brains powerful instead of just copying neurons better), although I don't quite agree this is the cargo cult phenomon.
There aren't a lot of details on IBMs artificial neural networks, but generally ANNs only model a few characteristics of actual brains. It's very superficial.
For example, the central auditory system in the mammalian brain includes many different types of neurons with very different sizes, shapes, and response properties. These are organized into tissues that are further organized into circuits. There is a significant architecture there.
To contrast, many ANNs use a simple model of a neuron (input, weight, fire or don't fire), with a simplistic architecture based on a few layers. The learning algorithms have little to do with the neural analogy and much more to do with statistical optimization.
Seems like a reasonable technical approach, but the problem is clearly with adoption.
AFAIK, IBM does not make wireless access points, and it's probably going to be hard to get the IEEE to adopt the mechanism (esp. if patented and restricted) as part of the 802.11x standards.
I'd say FB is somewhere around the peak. In not too long (maybe post-IPO), the media will turn on it, and in a couple years FB will settle out as useful for the reasons described in other posts here.
FWIW, I cancelled my account a long time ago due to the privacy shenanigans and personal repulsion at FB's business practices and ethically challenged CEO.
Seems like this is exactly what a primary care physician is supposed to do.
You know, a doctor who knows you and your medical history, and tells you if something may be wrong based on your doctor visits? Maybe suggests tests or scans proactively?
A tool like this will not have as much knowledge available as an actual doctor, and will fail to catch things a real doctor would possibly catch.
Wow, what a horrible site full of misinformation and straw man arguments.
This site was funded by the Bradley Foundation, who also funded hard-right "think tank" groups such as PNAC, the Heritage Foundation, the American Enterprise Institute, and the Federalist Society. The authors affirm they are a network of "pro-life" groups.
The site begins by linking belief in overpopulation to efforts to kill the poor and promote Chinese abortions, then proceeds with meaningless factoids (all the humans on earth could fit in Texas) to conclude that overpopulation is a myth.
The only legitimate argument on the site is that the Earth can produce enough food, although the argument relies on petrochemical fertilizers, and does not acknowledge constraints on the petrochemicals.
The site does not even acknowledge concerns about the high risk of global diseases, the massive amounts of waste products and pollution from industry and agriculture, or constraints on energy and water supplies. Oh, and nobody has even mentioned that there might not be enough jobs for everyone in the world.
In one section, the authors "prove" the Earth's population will peak around 8 B in 30 years and begin to decline by linking to the UN Population DB and telling you to use the "low variant" model. They don't tell you that the other three models (constant fertility, medium, high) all show the population continuing to rise for the duration of the model (present - 2050).
Talk about selective information. What a crock of shit.
No, I have driven through rural areas in my state, and very few of these people are farmers.
They are just too stupid to compete for real jobs and too dumb to live anywhere near civilization. So they all work shitty jobs at shitty fast food restaurants and eat snacks all day.
Then they vote for "small government republicans" so they can keep their guns and feel all Christy, yet expect city taxpayers to subsidize their roads, social programs, healthcare costs, and apparently, communications infrastructure.
I also tried recompressing the test image, and was only able to see the watermark around the 60-65 quality range for JPEG.
However, the goal presented in the paper is to provide a "targeted mark" that will be visible given a predictable recompression scheme. So, if you know a particular photo site (or WAP gateway) recompresses at 60, you can mark the picture to target quality rating of 60.
The paper concludes with "untargeted marks" as a future research opportunity. The authors describe an untargeted mark as one that would appear with some degree of certainty without knowing the recompression parameters of site X or WAP gateway Y.
As others have posted, there should be ways to recompress without the mark showing. Think of this technique as an exploit for JPEG quantization.. There will be workarounds to prevent the exploit.
There are obviously a lot of problems with these ad services, but maybe there is some value to the Security-for-Ads business model.
The enterprise has an arsenal of security technology that, for the most part, has not made it to the consumer space. This makes consumer-owned computers very easy targets, and that has given rise to botnets.
Either ISPs can give away this kind of security (e.g. IPS, botnet detection) for free, or consumers can pay for it. But, consumers will not pay for it. Maybe supporting network security for consumers with ads isn't such a strange idea.
That said, Kindsight does not seem to have much of a security focus. The most detail I could find on their website are vague references to "advanced threat detection technologies," and none of the positions in their job listings include security expertise.
Slashdot Mirror
If Facebook ends up close to $38 at the end of the day, it will be a rare example of the stock having been priced correctly at the start.
This assumes that the stock could be sold for less than $38. I'm not sure that's even possible in an IPO scenario like this (not saying it isn't, saying I don't know).
Who would buy the stock for $38-$42 then turn around and sell it the same day for less?
There is no doubt - for most of what IBM is selling, APAC and the Middle East are the growth markets. So it does make sense to move centers of business from the country with low single digit growth and a saturated market to the developing areas with strong growth and lots of greenspace
On the other hand, there's clearly a lot of outsourcing fever going on, and IBM may be participating in that but couching it in the "International" part of its namesake. It's hard to say.
Your view on democracy sounds great in theory, but American-style democracy just doesn't work like that.
Here's an example. Georgia just changed its law about selling beer and wine at the grocery store on Sundays. Previously, it had been illegal.. for decades.
Under new law, cities and counties could hold referendums to allow/disallow sunday sales in their area. In almost all of the cities/counties that held a vote, Sunday sales won in a landslide.. typically 70+% approving.
Now, this new law nearly died in the 2011 state congress thanks to special interest groups like the Christian Coalition. In fact, Sunday sales were kept illegal for decades due to their efforts, even though it is clearly what the majority wanted. This tells me the Christian Coalition and groups like it successfully suppressed democracy for DECADES, against the wishes of a 70+% majority!
I'd say the timing has to do with the Ceglia case.
With a recent judicial decision to sanction Ceglia, Facebook most likely believes they are in the endgame of that suit and can move forward with an IPO without the suit causing problems.
Think about it. Would you file for a $75B IPO (or whatever bullshit figure they've come up with) if there was an open suit with someone claiming they own half of the founder's shares, and backing it up with a signed contract?
I pretty much agree with your conclusions, but have a couple quibbles.
Cyber warfare is also called a denial of service attack, and is fundamentally different from cyber espionage.
Cyber warfare is not anonymous, done by a large number of IP addresses, and can't be defended against.
You're right that the purpose of an attack is to disrupt the target machines, but it won't always be a DDOS attack. Stuxnet was a worm with a targeted payload; if the German researcher hadn't found it and gone public, Iran might not have ever figured out why their centrifuges weren't working as anticipated.
But Stuxnet doesn't fit neatly into cyberwarfare. I'd say it fits better into a third category: Covert Action, which is, by design, deniable.
But, yeah, I totally agree that this is unnecessary saber rattling.
Kudos to wiredmikey (and the ed?) for capturing that attribution of an attack is the key sticking point for military response.
Attributing attacks in a packet switched network like the Internet is just a fantasy.. Sure, you can trace an attack back to, say, China, but how do you know the attack originated there? You don't, unless China cooperates and gives your forensics experts access to their networks. Which probably will not happen.
So the hawks want to shore up some credibility for attribution. Here is the plan, from the linked DoD PDF:
Nice try Pentagon, but statistically-powered voodoo does not overcome the problem here: that the attacking machines could be controlled from anywhere, possibly even through teh 7 proxies. Lulz.
Maybe we should listen to the National Research Council when they write "deterrence of cyberattacks by the threat of in-kind response has limited applicability." (NRC Report, p.5)
I'll close with a suggestion: why not, instead of focusing on how and when we get to launch attacks, focus on bettering our defenses?
This is pathetic.
oh yeah? I know what words mean, dipshit.
Why don't you explain why this is ambition but not arrogance.
looks like the Chinese are almost as arrogant as the Americans now.
A lot of the machine learning algorithms we use today are based on statistical or classification techniques that are mathematically connected to neural networks, and their development has in part been inspired by them.
If you are saying these techniques were borne from mathematical properties of biological neural networks, you are just wrong. Get a ML textbook - it's all about curve fitting, probability theory, decision theory, information theory, statistics, optimization.
Many of our machine vision and hearing algorithms are based on phenomenon that have been observed in the brain's visual and auditory cortex. The differences of Gaussians in SIFT, or the wavelets in SURF for example.
Wrong again. You're zero for two. Actually, if you can cite a biology paper concluding cortex uses wavelets, I'll give you this one. Good luck.
Have we got a machine that wakes up one day, says hello and asks for a cheeseburger? No, of course not. That's kind of the end goal, isn't it?
No, that's not the goal of ML or AI, and that has nothing to do with anything I've written. Quit with the hyperbole already.. If you can't back up your dumb ass opinions, just STFU.
... claiming that progress is not due at all to studying how the brain does ...
That's not at all what I said. Learn to read.
Thanks for the link, but it's still a pretty simple neural model. Just not as simple as many other common models, which is why they take great care to call it "biologically inspired." But, the focus of the research is on simulation, not intelligence.
To the original point, the researchers have simulated a better approximation of NNs without shedding any light on the "computational tricks" that make brains so smart. While the paper makes clear that this is a model that can be used to test neural theories, the goal of the project was to make a large scale simulation more feasible. The researchers did not claim to make better learning algorithms.
I disagree that we "have had a fair bit of success copying" the success of brains. Most machine learning these days (computer vision, etc.) does not even use neural models, and really reflects the amount of engineering effort and math that has gone into solving the problems. Some of the neural simulators I've read about have helped understand neural physics and chemistry, but not necessarily intelligence. We do, of course, have ANNs that accomplish very specific recognition or control tasks, but we do not have ANNs that, say, extract auditory features and build mental models of sound patterns.
And no, I am not the original AC. I generally agree with the last paragraph of that post (that we should look for what tricks make brains powerful instead of just copying neurons better), although I don't quite agree this is the cargo cult phenomon.
There aren't a lot of details on IBMs artificial neural networks, but generally ANNs only model a few characteristics of actual brains. It's very superficial.
For example, the central auditory system in the mammalian brain includes many different types of neurons with very different sizes, shapes, and response properties. These are organized into tissues that are further organized into circuits. There is a significant architecture there.
To contrast, many ANNs use a simple model of a neuron (input, weight, fire or don't fire), with a simplistic architecture based on a few layers. The learning algorithms have little to do with the neural analogy and much more to do with statistical optimization.
Seems like a reasonable technical approach, but the problem is clearly with adoption.
AFAIK, IBM does not make wireless access points, and it's probably going to be hard to get the IEEE to adopt the mechanism (esp. if patented and restricted) as part of the 802.11x standards.
Looks like the team there recognizes this as a key challenge. See the bottom of this post: A new solution to wireless security issues
Hype Cycle
I'd say FB is somewhere around the peak. In not too long (maybe post-IPO), the media will turn on it, and in a couple years FB will settle out as useful for the reasons described in other posts here.
FWIW, I cancelled my account a long time ago due to the privacy shenanigans and personal repulsion at FB's business practices and ethically challenged CEO.
The PISA compares the most affluent, best-educated cities in China (Shanghai, Hong Kong) to the whole population of the US.
Criticism of PISA
Try adding in all the Chinese hillbillies, or removing all the American hillbillies, and you might have a meaningful comparison.
Seems like this is exactly what a primary care physician is supposed to do.
You know, a doctor who knows you and your medical history, and tells you if something may be wrong based on your doctor visits? Maybe suggests tests or scans proactively?
A tool like this will not have as much knowledge available as an actual doctor, and will fail to catch things a real doctor would possibly catch.
Wow, what a horrible site full of misinformation and straw man arguments.
This site was funded by the Bradley Foundation, who also funded hard-right "think tank" groups such as PNAC, the Heritage Foundation, the American Enterprise Institute, and the Federalist Society. The authors affirm they are a network of "pro-life" groups.
The site begins by linking belief in overpopulation to efforts to kill the poor and promote Chinese abortions, then proceeds with meaningless factoids (all the humans on earth could fit in Texas) to conclude that overpopulation is a myth.
The only legitimate argument on the site is that the Earth can produce enough food, although the argument relies on petrochemical fertilizers, and does not acknowledge constraints on the petrochemicals.
The site does not even acknowledge concerns about the high risk of global diseases, the massive amounts of waste products and pollution from industry and agriculture, or constraints on energy and water supplies. Oh, and nobody has even mentioned that there might not be enough jobs for everyone in the world.
In one section, the authors "prove" the Earth's population will peak around 8 B in 30 years and begin to decline by linking to the UN Population DB and telling you to use the "low variant" model. They don't tell you that the other three models (constant fertility, medium, high) all show the population continuing to rise for the duration of the model (present - 2050).
Talk about selective information. What a crock of shit.
No, I have driven through rural areas in my state, and very few of these people are farmers.
They are just too stupid to compete for real jobs and too dumb to live anywhere near civilization. So they all work shitty jobs at shitty fast food restaurants and eat snacks all day.
Then they vote for "small government republicans" so they can keep their guns and feel all Christy, yet expect city taxpayers to subsidize their roads, social programs, healthcare costs, and apparently, communications infrastructure.
I also tried recompressing the test image, and was only able to see the watermark around the 60-65 quality range for JPEG.
However, the goal presented in the paper is to provide a "targeted mark" that will be visible given a predictable recompression scheme. So, if you know a particular photo site (or WAP gateway) recompresses at 60, you can mark the picture to target quality rating of 60.
The paper concludes with "untargeted marks" as a future research opportunity. The authors describe an untargeted mark as one that would appear with some degree of certainty without knowing the recompression parameters of site X or WAP gateway Y.
As others have posted, there should be ways to recompress without the mark showing. Think of this technique as an exploit for JPEG quantization.. There will be workarounds to prevent the exploit.
There is no trust relationship between cars, especially if there is no non-repudiation mechanism built in.
This system would create as many problems as it solves.
And out of thousands of patents there are destined to be some idiotic ones...
That argument works both ways:
Out of thousands of patents there are destined to be some useful ones.
I don't think we have data on how many IBM patents are stupid, so you're guessing at that 1% figure.
Maybe a better stat would be how many patents are licensed or productized, and how many patents sit on the shelf and are never used.
I can't prove it, but I suspect most IBM patents are never used, other than to lock up Intellectual Property. That's why I called it a land grab.
There is no doubt that IBM has some world-class labs that produce leading research and some amazing technology.
At the same time, IBM makes headlines (and slashdot articles) regularly for stupid patents and stupidly obvious patents. I won't bother linking all of them, but here is one from two months ago: http://entertainment.slashdot.org/story/10/09/15/2235251/IBM-Patents-Choose-Your-Own-Adventure-Movies
Remember that when you read about how many patents IBM gains and holds. This is a land grab.
There are obviously a lot of problems with these ad services, but maybe there is some value to the Security-for-Ads business model.
The enterprise has an arsenal of security technology that, for the most part, has not made it to the consumer space. This makes consumer-owned computers very easy targets, and that has given rise to botnets.
Either ISPs can give away this kind of security (e.g. IPS, botnet detection) for free, or consumers can pay for it. But, consumers will not pay for it. Maybe supporting network security for consumers with ads isn't such a strange idea.
That said, Kindsight does not seem to have much of a security focus. The most detail I could find on their website are vague references to "advanced threat detection technologies," and none of the positions in their job listings include security expertise.
If this is news, then things are really slowing down in the security industry.