Stopping malware is not a priority for advertising companies. The priority is to do whatever they can to help advertisers, because advertisers give them money. Money focuses people's priorities (including mine).
It is actually a priority. Google's ad-ranking system takes into account not just the revenue potential from an ad click but also "ad quality", a metric that considers various aspects of the ad, the site to which it links, and more, all related to the user experience. Because Google knows that it's important that when users click on a Google ad they have a good experience. Otherwise, they'll click less. Given that Google only gets paid when they click, that's directly bad for revenue. It likely also reduces the value of the ad to the advertiser, since users who do click may arrive more skeptical of what they'll find, and be less likely to buy. So advertisers will bid less, and that's bad for revenue.
To help the advertisers, Google provides feedback on what they can do to improve their ad quality metric, because it's one of two levels advertisers have to control how often their ad is shown (the other is how much money they bid for each click). Google also provides details statistics to enable advertisers to calculate their ROI from advertising on Google, which will quickly show the damage from any degradation in user trust in Google ads.
I don't know the details of what happened here, and although I could search the bug database to find out, if I knew the details I couldn't post. That said, I strongly suspect that what happened here is that the situation was more complicated than is presented in the article and that there were very good reasons why it took the Google ads team so long to address the issue. Because bad ads are bad for Google.
(Disclaimer: I work for Google, but don't speak for Google. My description of the ad ranking system is public knowledge and the rest is only my opinions.)
I don't believe the USA is more violent then it was before
It's actually as safe as it's ever been, safer even than the "Leave It to Beaver" 50s, and the decline in violence and crime is continuing. It's possible that some of the improvement in child safety is due to hypervigilant parents, but I suspect not much. Most of it is just that the nation is more... "civilized" is the best word I can come up with. It's still more dangerous and violent than many other developed nations, but in a better place than it has been, and heading the right direction.
I believe that people are just more aware of bad shit that happens because you have a non stop stream of information, pictures and videos coming from various sources.
Yup. Our perceptions are badly skewed by media. Our inbuilt mechanism for judging risk is heavily biased towards shocking narratives, and it's also observation-frequency biased. In evolutionary terms, those make sense. Without the range-extending capabilities of technology, our observations were limited to the personal, so observation frequency made sense. For rarer but more severe risks, the information communicated by others also provided a pretty good measure of frequency, since the aggregate perceptive range of our acquaintances and their acquaintances, etc., was pretty small.
That's clearly not the world we live in today.
Of course, we do have excellent tools for judging risk, vastly better than anything our ancestors had. Statistical methods provide a more accurate, more precise and more nuanced view of relative risk than anything our "gut" could ever do. If we use them.
In this case, these children's parents clearly do make use of the statistical tools available to us today, correctly judging the relative risk of their children walking as being lower than driving in an automobile. The CPS agency, not so much.
So what yer sayin' is that y'all got niggers runnin' loose. I kin see why yer afeard a lettin' God-fearin' white children walk theah. Them black bucks is mighty unperdictable.
Wait, this is the 1930s, right?
(Sorry if I misinterpreted your diversity comment, but it wouldn't surprise me if there is an element of racism at work here.)
I have some questions if historical Jesus existed, but the idea that he popped over to North America is absurd.
Why? If he can come back from the dead, certainly traveling to a different continent isn't so difficult.
If he preached to the Native Americans, he did a pretty shitty job of it, and the whole thing is absurd.
Given that the civilizations of that period vanished (archaeologically and per the history given in the Book of Mormon), what remnants of that visit would you expect to find?
the idea that some guy thousands of years later wrote a book in Elizabethan English is absurd on its face
That's not actually what the Book of Mormon purports to be. It claims to have been written between 600 BC and 200 AD, and translated in the 19th century, to English... in a "scriptural style", mimicking the Bible that Joseph Smith knew. So your complaint is that he chose to use that style, rather than his contemporary language? Okay, but that's a pretty weak criticism.
What are you saying is absurd? That there was a prophet? All religions based on Judaism believe in prophets. Or something in particular about Joseph Smith as a prophet?
I mean the talk that the Garden of Eden was in North America
So where was it, then? Or are you taking the position that it didn't exist but is just allegorical?
Or believing that a guy with prior convictions for fraud found some scriptures written by God himself inside a cave
That sentence is chock full of misinformation. Joseph Smith was never convicted of fraud (he was charged with banking fraud, but that was later -- all of his various charges came after he had published the Book of Mormon, and most were vague, like "disorderly person", because the people didn't like what he was saying -- and was a charge trumped up because they didn't like how the church members were organized), the Book of Mormon wasn't written by God himself but by a series of prophets, the same as the Bible, and it wasn't found in a cave.
The Catholic Church also says a lot of absurdities as well like claiming that St. Peter is the founder of the Church when the *real* founder was the Emperor Constantine for one.
Not much of a Catholic if you don't believe in the Apostolic succession:-)
Or maybe one day people will be able to go more than a 1/2 day without a "quick email check" (or whatever they "need" to do online - all the time). Seriously people, learn to disconnect.
Why? You can only stare at the cheap hotel room art for so long. TV sucks. Books are good, but the net has a lot more. In some locales it's worthwhile to leave the hotel and find other stuff to do, but in a lot of places I travel for business, there really isn't much point in that.
When evaluating a hotel room, I rate the importance of Internet service just below the importance of having a bed. If your Internet service doesn't work, I'm leaving.
Wow, I'm amazed they would charge their guests for wifi access. Even the cheapest, sleaziest, motels have "free" wifi. Are they the only hotel doing this? Reminds me of the time when some motels made their televisions coin-operated.
It varies widely across the levels of hotels, but there are some patterns:
The really sleazy motels generally charge for Internet (some don't even have it).
Stepping up to stuff like Comfort Inn, Econolodge, etc., wifi is generally free. Their customers are price-sensitive and are likely to be annoyed at being asked to pay extra for much of anything.
The next step up is the lower tier of business travel hotels, like Hampton Inn and such. They generally have free wifi, same rationale as the previous. However, many of them offer a premium service with higher bandwidth, perhaps a external IP, because some business travelers need it and will often pay.
The next step up is the higher tier of business travel hotels, like Marriott Residence Inn, Embassy Suites, etc., vary. Pretty much the same situation as the lower tier, but a higher percentage of them charge even for "basic" service. They nearly always have free Internet in their business center.
Luxury hotels mostly charge for Internet. Luxury hotels nickel and dime you for every damned thing they can think of, I suppose on the theory that if you're willing to pay $350+ per night for a room, you won't bother to look at the bill and notice an extra $50 per day in extra charges.
I don't often stay at "destination" hotels ($700 per night and up, usually), in fact my sample size is two stays at the same Ritz Carlton (the one in Half Moon Bay), but what I saw there was that Wifi was free again. Not, I'm sure, because they think the patrons are sensitive to the price, but because making them go through some hassle to get onto the net is unacceptable customer service and clashes with the bowing and scraping that is de rigeur in every other part of the experience.
All of this is in relation to hotels in the US. International travelers can expect it to be all over the place. I stayed at a fairly nice place in Zurich that didn't offer Internet at all, and a fleabag in Santiago that had outstanding Internet. That's not saying anything about Zurich or Santiago, either; different hotels in the same area were different.
Until it's on a store shelf it isn't for sale. It was never advertized as a consumer product. It wasn't even promoted. To get it you had to go out of your way to even find out where you were supposed to get the damn thing.
Yes, it's very well hidden, on the "devices" page in the Google Play store: https://play.google.com/store/..., right below the Nest devices and right above Chromebooks.
Here's a thought. What if you accidently keep one bit of information that could be turned against you when out of context, and you diligently deleted the very documents that would have shown the redeeming context?
Then you explain the context, and have the relevant people testify about their recollection of it. With no documentation to trip them up, and with the benefit of hindsight. This actually happens a lot, when some of the documentation still exists while other documentation has already been deleted.
If one of your employees step out of line and produces something that could be turned against you, then act on it. If the paper trail shows you did, then you have nothing to fear from sensible people.
That assumes you recognize it before it becomes a problem. Which, besides being difficult on its own, raises the question of who the "you" is. An e-mail communication between two employees, neither of whom sees a problem, won't ever come to the attention of management, much less the legal staff. Not until a discovery search, at which point it's far too late. Acting at that point could actually make things worse.
Perhaps you want management and legal to be responsible for reading every e-mail, chat and document sent or received by any employee? That would be even more insanely expensive than trawling through terabytes of old data because in the discovery case the searchers at least have some notion of what they're looking for and can ignore much irrelevant documentation. Not to mention the chilling effect on employee productivity and/or morale.
Never trust a company which goes to great length to cover up their past.
Agreed. But having a reasonable retention policy is not going to "great length".
Those will attack you anyway, with or without evidence.
Without evidence, suits get dismissed and media gets bored and moves on.
Stop feeding them.
Yes, that's exactly the point of a retention policy, to eliminate litigious lawyer-feed.
The machine I'm working on will use emotions - or a rough facsimile of emotions. Are they necessary for intelligence? - I don't know, but I do know that they make the whole design a lot simpler and more logical.
Interesting. Can you elaborate on how they simplify the design? I'm not sure it's a meaningful data point anyway, because since we don't understand general intelligence we can't know what a successful design will look like or require. But I'm interested to hear your reasoning.
Besides these machines will have to work with humans - they will have to understand emotions to understand us..
It's not necessary to have some characteristic in order to understand it and work with it.
There is a problem with that argument - namely that it is possible to make an argument that even the most basic creatures - single cells - show hints of both the most basic awareness and survival instinct
Yes, that's why I said the claim is "very unlikely":-)
Emotions are totally logical - you merely have to understand them - emotions are the heart of the human and animal behaviour control system.
I don't think anyone understands what emotions are. They clearly are at the center of the human behavior control system, and it's pretty clear why the evolved the way they did. That says nothing about whether or not AI would have anything analogous.
The Strong AI needs a 'motivator' - and emotions by definition are a great motivator.
But by no means the only motivator. If your contention is that emotion is a necessary component of intelligence, I'm interested to hear your reasoning.
The problem comes when the machine tries to read human emotions - or when the human tries to read the machines emotions.. That will certainly be an interesting time... and we may learn a lot.
We will learn a huge amount about intelligence, thought and emotion as we develop AI, yes.
The traditional answer is that the emotions are read through some form of 'psychic aura', the current scientific answer is that they are conveyed through subliminal signals - posture, expression, pheromones, voice tension, eye signals. As a reductionist my answer is : insufficient data either way
There's plenty of data. It's perfectly possible to discern emotion via media which place the viewer at a great physical and temporal distance, unless you're going to argue that a video camera is capable of capturing the 'psychic aura'.
If huge corporations started following some basic legal and ethical guidelines, they wouldn't have to worry so much about old documents getting leaked. If your business strategy is to f##k your customers and/or your partners, sooner or later you will pay for it, documents or no documents.
This is an appealing idea, but it's just not true in the world we live in. People make mistakes, say things they shouldn't, joke about things they shouldn't. Worse, attorneys are masters at finding evil implications and subtexts even where they don't exist. It's very easy for companies who are really trying to be good corporate citizens, serve their customers well, treat partners fairly and generally behave well to end up shafted by something that turned up in discovery in a malicious lawsuit.
To reduce the chance of that, the company's lawyers need to go through all of the retained data with a fine-toothed comb, checking every item to see if it needs to be turned over, and if some of it should be redacted. This process gets insanely expensive when retained data is measured in terabytes.
This is why good companies with competent legal counsel define and implement retention and destruction policies. Having a standard policy that you always follow means that data destroyed per policy is not destruction of evidence, but just housekeeping. Setting that policy to be relatively aggressive about destruction reduces the quantity of material available for discovery, so the attorneys don't have to wade through all of it. Even better, there is NO chance that something in deleted data can be twisted or taken out of context, or that some employee's joke or fit of anger could cost the company a huge settlement.
Of course, unethical companies are generally really aggressive about destroying old data, for exactly the reason that per-policy destruction of evidence is not, legally, destruction of evidence. One way to tell the difference is to look at the policy. Most companies retain data for six months to a year, and recommend to employees that they be careful what they say in stored communications. Shady companies often have policies mandating destruction in as little as 30 days, as well as a strong corporate culture of never putting anything important in storable form.
Will it be as toothless as HIPAA or SOX, where the only person thrown in jail on Sarbanes-Oxley was guy who fished up one too many groupers?
If you think that HIPAA and SOX are toothless, you don't know anything about them. The number of people thrown in jail is far from the only valid metric. Spend some time working in corporate worlds that manage medical or financial information and see just how terrified everyone is of violating them. In the relevant industries you can get almost anything done, regardless of whether it makes sense, if you can make a vaguely believable argument that HIPAA or SOX requires it.
If enforced, where is there proof that the hole was discovered, and what date? I'm sure a H-1B will be darn sure to keep mum when he/she actually found the breach in order to not be deported.
From an enforcement perspective, the date will be the date on the first documented discussion, or the date recalled by a whistleblower. This sort of stuff tends to always generate an e-mail trail.
What is a breach? Is someone duping gold on ClicheQuest considered a breach? A warp hack? What about a web server showing the FTP server's links? The courts can be clogged for years of lawyers deliberating this... and when it comes to technical issues, courts tend to side with what side has the most lawyers.
Sure, for any situation there are edge cases. But who cares whether gold-duping is considered a breach? A laptop full of names and social security numbers walking out the door is clearly a breach, and that's what we care about. But, regardless, legislation actually tends to be quite careful about defining such things. That care is a lot of what makes the law hard to read.
What happens when a breach and trade secrets smack into each other? A court erroring one way, and businesses can have their secret sauce dumped out by clever lawyers. Another way, and every breach can be covered up as a trade secret.
Trade secret law cannot be used to hide information from courts. They'll simply request the data and seal it. If it's dumped out by lawyers that will only be because the lawyers for the owner of the secrets were negligent. Filing the motions needed to protect such data is their job.
Who is going to fund enforcement?
The Department of Justice, same as all federal laws. Sure, a future president could direct the DoJ not to bother, just as Obama has directed them not to pursue pot smokers, but in this case that would be a really hard move to justify politically.
I assumed always that our self-preservation came about because we have consciousness.
That seems very unlikely. This would imply that creatures that don't have consciousness lack the instinct for self-preservation. That would mean we should see a lot of lower life forms that don't try to protect themselves. It would also seem to imply that our self-preservation should focus primarily on us as individuals, and not on our family or species.
If we instead look at self-preservation as an evolutionarily-derived imperative, it's pretty clear that we should expect all organisms to protect their genes, since those that didn't would be more likely to get selected out. Note "genes", not "self", except to the extent that protecting the self protects the replication of the genes. That provides a much better explanation of observed behavior, particularly the strong tendency of humans to defend their families and their tribes (however tribes are constituted) even at the expense of their own lives, but to defend themselves over just about anything else.
And if the instinct for self-preservation is a result of evolutionary forces, then AI that is created by us rather than evolved will be very unlikely to have that instinct. Unless we create it via competitive evolution-style methods.
A robot without self-awareness could follow a rule but would not have any internal feelings about that rule. Without those feelings, rules alone won't work. Philosophy majors take over this discussion...
Why do you think that self-awareness implies "feelings"? Emotions seem also to be the result of survival imperatives: love and affection serve to encourage procreation and protection of offspring, and binds us into mutually-supporting communities of various sizes; anger and hate are important responses to dissuade non-cooperation in said communities; fear and pain serve to help us to protect ourselves; and so on. For any emotion you can name, evolutionary pressures explain it. Of course the fact that an explanation can be found doesn't mean the explanation is correct, but in order for one idea to explain so much, that idea must have extraordinary "reach"... which also exposes the idea to correspondingly many opportunities for falsification. This gives us strong reason to believe it.
And, again, AIs developed by non-competitive processes have no reason to develop these various emotions... though it could empirically derive the dynamics which drove their development, and therefore logically choose to act as though it did have them.
Philosophy majors take over this discussion
Sorry, math/CS major here. Though I am reading Russell's History of Western Philosophy.
Anonymous, verifiable voting that allows the voter to check their vote was counted correctly, but not prove to anyone else how they voted, is possible. See the Punchscan system. Some more recent research has also shown how this can be done electronically (Punchscan uses paper ballots).
Of course, that still leaves open the door for coercion/payment at time of vote, but if that were as large a problem as often claimed we couldn't trust absentee and other mail-in ballots.
Let's face it, how many children from ghetto neighborhoods are working at Google?
Not to detract from your point, but there are a fair number of people from ghetto and poor rural neighborhoods working at Google. I'd estimate that about 5% of American Google engineers come from a background that could be described that way. That's just a guess based on personal observation, but I think it's probably not too far off the mark. My current team has a much higher percentage of people from low-income backgrounds -- probably 50% -- but it's an atypical team in many ways.
It's interesting to speculate on the causality in that correlation. The obvious expectation is that the emotional state engendered by a big and unavoidable expense causes a reduction in intelligence, and that it's the relative scale of the expense which causes the difference between poor and wealthy people. However, it's also possible that the ability to continue thinking clearly in the fact of disastrous expense is what enables people to build and preserve wealth. In fact, I think resilience of that sort is clearly a big factor in wealth.
The researchers should try scaling the size of the disastrous expense relative to the subjects' wealth.
That's going to kill the resale value of the existing Leafs, so if you want a short-range electric vehicle at a good price, there are going to be some great deals in the next two years.
That's why I leased my LEAF. Not because I predicted this particular change, but because I knew significant improvement would be coming. EV technology is improving rapidly.
Slashdot Mirror
Stopping malware is not a priority for advertising companies. The priority is to do whatever they can to help advertisers, because advertisers give them money. Money focuses people's priorities (including mine).
It is actually a priority. Google's ad-ranking system takes into account not just the revenue potential from an ad click but also "ad quality", a metric that considers various aspects of the ad, the site to which it links, and more, all related to the user experience. Because Google knows that it's important that when users click on a Google ad they have a good experience. Otherwise, they'll click less. Given that Google only gets paid when they click, that's directly bad for revenue. It likely also reduces the value of the ad to the advertiser, since users who do click may arrive more skeptical of what they'll find, and be less likely to buy. So advertisers will bid less, and that's bad for revenue.
To help the advertisers, Google provides feedback on what they can do to improve their ad quality metric, because it's one of two levels advertisers have to control how often their ad is shown (the other is how much money they bid for each click). Google also provides details statistics to enable advertisers to calculate their ROI from advertising on Google, which will quickly show the damage from any degradation in user trust in Google ads.
I don't know the details of what happened here, and although I could search the bug database to find out, if I knew the details I couldn't post. That said, I strongly suspect that what happened here is that the situation was more complicated than is presented in the article and that there were very good reasons why it took the Google ads team so long to address the issue. Because bad ads are bad for Google.
(Disclaimer: I work for Google, but don't speak for Google. My description of the ad ranking system is public knowledge and the rest is only my opinions.)
I don't believe the USA is more violent then it was before
It's actually as safe as it's ever been, safer even than the "Leave It to Beaver" 50s, and the decline in violence and crime is continuing. It's possible that some of the improvement in child safety is due to hypervigilant parents, but I suspect not much. Most of it is just that the nation is more... "civilized" is the best word I can come up with. It's still more dangerous and violent than many other developed nations, but in a better place than it has been, and heading the right direction.
I believe that people are just more aware of bad shit that happens because you have a non stop stream of information, pictures and videos coming from various sources.
Yup. Our perceptions are badly skewed by media. Our inbuilt mechanism for judging risk is heavily biased towards shocking narratives, and it's also observation-frequency biased. In evolutionary terms, those make sense. Without the range-extending capabilities of technology, our observations were limited to the personal, so observation frequency made sense. For rarer but more severe risks, the information communicated by others also provided a pretty good measure of frequency, since the aggregate perceptive range of our acquaintances and their acquaintances, etc., was pretty small.
That's clearly not the world we live in today.
Of course, we do have excellent tools for judging risk, vastly better than anything our ancestors had. Statistical methods provide a more accurate, more precise and more nuanced view of relative risk than anything our "gut" could ever do. If we use them.
In this case, these children's parents clearly do make use of the statistical tools available to us today, correctly judging the relative risk of their children walking as being lower than driving in an automobile. The CPS agency, not so much.
The county is pretty diverse
So what yer sayin' is that y'all got niggers runnin' loose. I kin see why yer afeard a lettin' God-fearin' white children walk theah. Them black bucks is mighty unperdictable.
Wait, this is the 1930s, right?
(Sorry if I misinterpreted your diversity comment, but it wouldn't surprise me if there is an element of racism at work here.)
I don't buy it. You're claiming to be decades ahead of the rest of the industry.
I have some questions if historical Jesus existed, but the idea that he popped over to North America is absurd.
Why? If he can come back from the dead, certainly traveling to a different continent isn't so difficult.
If he preached to the Native Americans, he did a pretty shitty job of it, and the whole thing is absurd.
Given that the civilizations of that period vanished (archaeologically and per the history given in the Book of Mormon), what remnants of that visit would you expect to find?
the idea that some guy thousands of years later wrote a book in Elizabethan English is absurd on its face
That's not actually what the Book of Mormon purports to be. It claims to have been written between 600 BC and 200 AD, and translated in the 19th century, to English... in a "scriptural style", mimicking the Bible that Joseph Smith knew. So your complaint is that he chose to use that style, rather than his contemporary language? Okay, but that's a pretty weak criticism.
The list goes on.
Is the rest of the list equally weak?
Joseph Smith?
What are you saying is absurd? That there was a prophet? All religions based on Judaism believe in prophets. Or something in particular about Joseph Smith as a prophet?
I mean the talk that the Garden of Eden was in North America
So where was it, then? Or are you taking the position that it didn't exist but is just allegorical?
Or believing that a guy with prior convictions for fraud found some scriptures written by God himself inside a cave
That sentence is chock full of misinformation. Joseph Smith was never convicted of fraud (he was charged with banking fraud, but that was later -- all of his various charges came after he had published the Book of Mormon, and most were vague, like "disorderly person", because the people didn't like what he was saying -- and was a charge trumped up because they didn't like how the church members were organized), the Book of Mormon wasn't written by God himself but by a series of prophets, the same as the Bible, and it wasn't found in a cave.
The Catholic Church also says a lot of absurdities as well like claiming that St. Peter is the founder of the Church when the *real* founder was the Emperor Constantine for one.
Not much of a Catholic if you don't believe in the Apostolic succession :-)
Or maybe one day people will be able to go more than a 1/2 day without a "quick email check" (or whatever they "need" to do online - all the time). Seriously people, learn to disconnect.
Why? You can only stare at the cheap hotel room art for so long. TV sucks. Books are good, but the net has a lot more. In some locales it's worthwhile to leave the hotel and find other stuff to do, but in a lot of places I travel for business, there really isn't much point in that.
When evaluating a hotel room, I rate the importance of Internet service just below the importance of having a bed. If your Internet service doesn't work, I'm leaving.
Wow, I'm amazed they would charge their guests for wifi access. Even the cheapest, sleaziest, motels have "free" wifi. Are they the only hotel doing this? Reminds me of the time when some motels made their televisions coin-operated.
It varies widely across the levels of hotels, but there are some patterns:
The really sleazy motels generally charge for Internet (some don't even have it).
Stepping up to stuff like Comfort Inn, Econolodge, etc., wifi is generally free. Their customers are price-sensitive and are likely to be annoyed at being asked to pay extra for much of anything.
The next step up is the lower tier of business travel hotels, like Hampton Inn and such. They generally have free wifi, same rationale as the previous. However, many of them offer a premium service with higher bandwidth, perhaps a external IP, because some business travelers need it and will often pay.
The next step up is the higher tier of business travel hotels, like Marriott Residence Inn, Embassy Suites, etc., vary. Pretty much the same situation as the lower tier, but a higher percentage of them charge even for "basic" service. They nearly always have free Internet in their business center.
Luxury hotels mostly charge for Internet. Luxury hotels nickel and dime you for every damned thing they can think of, I suppose on the theory that if you're willing to pay $350+ per night for a room, you won't bother to look at the bill and notice an extra $50 per day in extra charges.
I don't often stay at "destination" hotels ($700 per night and up, usually), in fact my sample size is two stays at the same Ritz Carlton (the one in Half Moon Bay), but what I saw there was that Wifi was free again. Not, I'm sure, because they think the patrons are sensitive to the price, but because making them go through some hassle to get onto the net is unacceptable customer service and clashes with the bowing and scraping that is de rigeur in every other part of the experience.
All of this is in relation to hotels in the US. International travelers can expect it to be all over the place. I stayed at a fairly nice place in Zurich that didn't offer Internet at all, and a fleabag in Santiago that had outstanding Internet. That's not saying anything about Zurich or Santiago, either; different hotels in the same area were different.
I can coexist with people who have a religion I think is patently absurd (I'm looking at you, Mormons)
Out of curiosity, which Mormon beliefs do you find absurd? I'm mostly interested to see whether they're actual Mormon beliefs, or urban legends.
But you have to 1 - already know that something called Google Glass exists.
Either that or notice it while perusing the other devices Google has for sale.
2 - know what the hell it is.
Either that or read the description on the Play site.
3 - be willing to shell out a fortune for an in-development toy.
Granted on the fortune. $1500 is expensive.
Until it's on a store shelf it isn't for sale. It was never advertized as a consumer product. It wasn't even promoted. To get it you had to go out of your way to even find out where you were supposed to get the damn thing.
Yes, it's very well hidden, on the "devices" page in the Google Play store: https://play.google.com/store/..., right below the Nest devices and right above Chromebooks.
Here's a thought. What if you accidently keep one bit of information that could be turned against you when out of context, and you diligently deleted the very documents that would have shown the redeeming context?
Then you explain the context, and have the relevant people testify about their recollection of it. With no documentation to trip them up, and with the benefit of hindsight. This actually happens a lot, when some of the documentation still exists while other documentation has already been deleted.
If one of your employees step out of line and produces something that could be turned against you, then act on it. If the paper trail shows you did, then you have nothing to fear from sensible people.
That assumes you recognize it before it becomes a problem. Which, besides being difficult on its own, raises the question of who the "you" is. An e-mail communication between two employees, neither of whom sees a problem, won't ever come to the attention of management, much less the legal staff. Not until a discovery search, at which point it's far too late. Acting at that point could actually make things worse.
Perhaps you want management and legal to be responsible for reading every e-mail, chat and document sent or received by any employee? That would be even more insanely expensive than trawling through terabytes of old data because in the discovery case the searchers at least have some notion of what they're looking for and can ignore much irrelevant documentation. Not to mention the chilling effect on employee productivity and/or morale.
Never trust a company which goes to great length to cover up their past.
Agreed. But having a reasonable retention policy is not going to "great length".
Those will attack you anyway, with or without evidence.
Without evidence, suits get dismissed and media gets bored and moves on.
Stop feeding them.
Yes, that's exactly the point of a retention policy, to eliminate litigious lawyer-feed.
The machine I'm working on will use emotions - or a rough facsimile of emotions. Are they necessary for intelligence? - I don't know, but I do know that they make the whole design a lot simpler and more logical.
Interesting. Can you elaborate on how they simplify the design? I'm not sure it's a meaningful data point anyway, because since we don't understand general intelligence we can't know what a successful design will look like or require. But I'm interested to hear your reasoning.
Besides these machines will have to work with humans - they will have to understand emotions to understand us..
It's not necessary to have some characteristic in order to understand it and work with it.
Is he sorry that they created a monster or is he just sorry that they got caught and now their credibility is in the trash can?
He's sorry that they continued supporting it after the flaws were discovered. He regrets that they were so obvious.
There is a problem with that argument - namely that it is possible to make an argument that even the most basic creatures - single cells - show hints of both the most basic awareness and survival instinct
Yes, that's why I said the claim is "very unlikely" :-)
Emotions are totally logical - you merely have to understand them - emotions are the heart of the human and animal behaviour control system.
I don't think anyone understands what emotions are. They clearly are at the center of the human behavior control system, and it's pretty clear why the evolved the way they did. That says nothing about whether or not AI would have anything analogous.
The Strong AI needs a 'motivator' - and emotions by definition are a great motivator.
But by no means the only motivator. If your contention is that emotion is a necessary component of intelligence, I'm interested to hear your reasoning.
The problem comes when the machine tries to read human emotions - or when the human tries to read the machines emotions.. That will certainly be an interesting time... and we may learn a lot.
We will learn a huge amount about intelligence, thought and emotion as we develop AI, yes.
The traditional answer is that the emotions are read through some form of 'psychic aura', the current scientific answer is that they are conveyed through subliminal signals - posture, expression, pheromones, voice tension, eye signals. As a reductionist my answer is : insufficient data either way
There's plenty of data. It's perfectly possible to discern emotion via media which place the viewer at a great physical and temporal distance, unless you're going to argue that a video camera is capable of capturing the 'psychic aura'.
If huge corporations started following some basic legal and ethical guidelines, they wouldn't have to worry so much about old documents getting leaked. If your business strategy is to f##k your customers and/or your partners, sooner or later you will pay for it, documents or no documents.
This is an appealing idea, but it's just not true in the world we live in. People make mistakes, say things they shouldn't, joke about things they shouldn't. Worse, attorneys are masters at finding evil implications and subtexts even where they don't exist. It's very easy for companies who are really trying to be good corporate citizens, serve their customers well, treat partners fairly and generally behave well to end up shafted by something that turned up in discovery in a malicious lawsuit.
To reduce the chance of that, the company's lawyers need to go through all of the retained data with a fine-toothed comb, checking every item to see if it needs to be turned over, and if some of it should be redacted. This process gets insanely expensive when retained data is measured in terabytes.
This is why good companies with competent legal counsel define and implement retention and destruction policies. Having a standard policy that you always follow means that data destroyed per policy is not destruction of evidence, but just housekeeping. Setting that policy to be relatively aggressive about destruction reduces the quantity of material available for discovery, so the attorneys don't have to wade through all of it. Even better, there is NO chance that something in deleted data can be twisted or taken out of context, or that some employee's joke or fit of anger could cost the company a huge settlement.
Of course, unethical companies are generally really aggressive about destroying old data, for exactly the reason that per-policy destruction of evidence is not, legally, destruction of evidence. One way to tell the difference is to look at the policy. Most companies retain data for six months to a year, and recommend to employees that they be careful what they say in stored communications. Shady companies often have policies mandating destruction in as little as 30 days, as well as a strong corporate culture of never putting anything important in storable form.
Wow, vastly different experience than I had in the financial and health care industries. You should have blown the whistle on them. It's not too late.
Will it be as toothless as HIPAA or SOX, where the only person thrown in jail on Sarbanes-Oxley was guy who fished up one too many groupers?
If you think that HIPAA and SOX are toothless, you don't know anything about them. The number of people thrown in jail is far from the only valid metric. Spend some time working in corporate worlds that manage medical or financial information and see just how terrified everyone is of violating them. In the relevant industries you can get almost anything done, regardless of whether it makes sense, if you can make a vaguely believable argument that HIPAA or SOX requires it.
If enforced, where is there proof that the hole was discovered, and what date? I'm sure a H-1B will be darn sure to keep mum when he/she actually found the breach in order to not be deported.
From an enforcement perspective, the date will be the date on the first documented discussion, or the date recalled by a whistleblower. This sort of stuff tends to always generate an e-mail trail.
What is a breach? Is someone duping gold on ClicheQuest considered a breach? A warp hack? What about a web server showing the FTP server's links? The courts can be clogged for years of lawyers deliberating this... and when it comes to technical issues, courts tend to side with what side has the most lawyers.
Sure, for any situation there are edge cases. But who cares whether gold-duping is considered a breach? A laptop full of names and social security numbers walking out the door is clearly a breach, and that's what we care about. But, regardless, legislation actually tends to be quite careful about defining such things. That care is a lot of what makes the law hard to read.
What happens when a breach and trade secrets smack into each other? A court erroring one way, and businesses can have their secret sauce dumped out by clever lawyers. Another way, and every breach can be covered up as a trade secret.
Trade secret law cannot be used to hide information from courts. They'll simply request the data and seal it. If it's dumped out by lawyers that will only be because the lawyers for the owner of the secrets were negligent. Filing the motions needed to protect such data is their job.
Who is going to fund enforcement?
The Department of Justice, same as all federal laws. Sure, a future president could direct the DoJ not to bother, just as Obama has directed them not to pursue pot smokers, but in this case that would be a really hard move to justify politically.
I assumed always that our self-preservation came about because we have consciousness.
That seems very unlikely. This would imply that creatures that don't have consciousness lack the instinct for self-preservation. That would mean we should see a lot of lower life forms that don't try to protect themselves. It would also seem to imply that our self-preservation should focus primarily on us as individuals, and not on our family or species.
If we instead look at self-preservation as an evolutionarily-derived imperative, it's pretty clear that we should expect all organisms to protect their genes, since those that didn't would be more likely to get selected out. Note "genes", not "self", except to the extent that protecting the self protects the replication of the genes. That provides a much better explanation of observed behavior, particularly the strong tendency of humans to defend their families and their tribes (however tribes are constituted) even at the expense of their own lives, but to defend themselves over just about anything else.
And if the instinct for self-preservation is a result of evolutionary forces, then AI that is created by us rather than evolved will be very unlikely to have that instinct. Unless we create it via competitive evolution-style methods.
A robot without self-awareness could follow a rule but would not have any internal feelings about that rule. Without those feelings, rules alone won't work. Philosophy majors take over this discussion...
Why do you think that self-awareness implies "feelings"? Emotions seem also to be the result of survival imperatives: love and affection serve to encourage procreation and protection of offspring, and binds us into mutually-supporting communities of various sizes; anger and hate are important responses to dissuade non-cooperation in said communities; fear and pain serve to help us to protect ourselves; and so on. For any emotion you can name, evolutionary pressures explain it. Of course the fact that an explanation can be found doesn't mean the explanation is correct, but in order for one idea to explain so much, that idea must have extraordinary "reach"... which also exposes the idea to correspondingly many opportunities for falsification. This gives us strong reason to believe it.
And, again, AIs developed by non-competitive processes have no reason to develop these various emotions... though it could empirically derive the dynamics which drove their development, and therefore logically choose to act as though it did have them.
Philosophy majors take over this discussion
Sorry, math/CS major here. Though I am reading Russell's History of Western Philosophy.
Anonymous, verifiable voting that allows the voter to check their vote was counted correctly, but not prove to anyone else how they voted, is possible. See the Punchscan system. Some more recent research has also shown how this can be done electronically (Punchscan uses paper ballots).
Of course, that still leaves open the door for coercion/payment at time of vote, but if that were as large a problem as often claimed we couldn't trust absentee and other mail-in ballots.
Nissan's 2016 LEAF is going to have a 200+-mile range, and will also be sub-$30K.
You meant a 125 mile range (200km). Unit of measure is important.
http://insideevs.com/nissan-exec-reliable-125-miles-of-range-coming-to-leaf-by-2016/
Hmm. The articles I've seen were talking about a 300 km range.
Let's face it, how many children from ghetto neighborhoods are working at Google?
Not to detract from your point, but there are a fair number of people from ghetto and poor rural neighborhoods working at Google. I'd estimate that about 5% of American Google engineers come from a background that could be described that way. That's just a guess based on personal observation, but I think it's probably not too far off the mark. My current team has a much higher percentage of people from low-income backgrounds -- probably 50% -- but it's an atypical team in many ways.
It's interesting to speculate on the causality in that correlation. The obvious expectation is that the emotional state engendered by a big and unavoidable expense causes a reduction in intelligence, and that it's the relative scale of the expense which causes the difference between poor and wealthy people. However, it's also possible that the ability to continue thinking clearly in the fact of disastrous expense is what enables people to build and preserve wealth. In fact, I think resilience of that sort is clearly a big factor in wealth.
The researchers should try scaling the size of the disastrous expense relative to the subjects' wealth.
That's going to kill the resale value of the existing Leafs, so if you want a short-range electric vehicle at a good price, there are going to be some great deals in the next two years.
That's why I leased my LEAF. Not because I predicted this particular change, but because I knew significant improvement would be coming. EV technology is improving rapidly.