The C# (or Java or JavaScript) programmer who doesn't understand memory management has memory leaks all over his (or her) code (or they are not writing anything substantial).
True, but they'll have a lot less of those leaks. You have to admit that keeping references to objects you no longer need is less of a memory leak problem than forgetting to explicitly free and losing the reference.
I agree with your point, but I don't think you could even get any real work done in C without understanding part of the memory management.
As somebody who specifically works with undergraduate students with a Java background, the first few C++ programs I get from (almost) every new intern while training them is full of "Foo* bar = new Foo();", without the accompanying deletes. The way that happens is pretty clear: they're used to instantiating instances with new, they get a compiler error that indicates they should have an asterix next to they type when they do that, so they add an asterix.
Granted, the process with C would be a bit different because malloc and free don't look like a keyword they're familiar with, but a lot of the same problems apply. Usually allocating memory in a function and returning the pointer which never gets freed. Sometimes they delete a pointer but keep references to it, and the test programs works perfectly ok because the process hasn't overwritten the memory marked for deletion by that point yet, etc.
I love C, but there's no denying that it is a lot easier to create vulnerabilities from ignorance in that language than in a garbage collected language with actual array objects that know their size instead of a head address in contiguously allocated memory, etc. Understanding that helps us train better programmers, because we can hammer that early on.
I would actually be more interested in seeing them break apart into hardware / software / licensed entertainment.
I don't think the mobile computing vs. desktop computer is a big deal. The real issue is that the same people making the hardware are making the software. If you can buy yourself an iphone and decide whether to install android or ios, then it's all good.
If I was an owner of a business there is no way in heck that I would want to belong to a 'scheduling service' that results in a computerized voice calling my employees constantly. What a time waster for the person taking calls.
Which is why they need to not ask permission and stop identifying themselves as a computer, and give it enough voice variation that you can't clue in by that. The entire point of the duplex demo was that it sounded human.
You want me to not use such a service, have online services. If the only way to make an appointment is to make a phone call, and I have a system that will do that for me, I'll be more than glad to shift the burden. If there's on website where I can do it, that's better, but if you want to remain in the 20th century, that's your problem.
How does anyone know it's a joke until they arrest the kid and question him and investigate?
Yeah, question the knothead. And investigate as needed
Although I do agree with you that when that questioning and investigating is the first step is you believe there's a possible danger there, I do fear that as a whole our society is getting very bad at understanding the nuances of context. Well, there have always been people with trouble doing that, but it's a bigger issue when police has that problem.
It's incredibly clear the post was making fun of the fact Siri's lack of understanding of the question and its implications led it to give it a response that was apparently a willingness to help a potential school shooter. Ironically, the kid found out human beings can be just as stupid as Siri regarding natural language comprehension. Siri understood his school shooting request as just someone looking for a school. People understood his humorous post as a potential threat.
I know we have a path for improving natural language understanding in AI software. How do we tackle improving humans (or at least governement organizations made up of humans)? Because that's the real problem here.
Well, The Orville episodes are deliberately social commentary. That episode was clearly inspired by China's new social credit system as well as the US latest propensity to treat popular opinion as facts.
Exactly Musk (and he is smart enough to know better) is conflating intelligence with access to information. They two are not the same.
Although you are correct, it's more difficult to draw the line than you'd think. You show up to your grandparent's place, they ask you to fix the computer. Maybe they have a Mac, and you don't use a Mac. But you have experience, you know your away around computers, you know how to Google. And you have their problem fixed up in a jiffy. They immediately say, "our grandson, he's so smart..."
Turns out knowledge, experience, and access to information allow you to accomplish things that people without those things can't accomplish. Similarly to how greater intelligence might let you accomplish things others can't. I say might, because a "dumber" person who applied themselves to acquire a lot of knowledge and experience will accomplish more than a highly intelligent person who is lazy and lacks ambition. And now you ask, "is the other person more intelligent, because they planned their life better?"
In some ways, the answer is yes, and now we're getting into different kinds of intelligence. You take a Rain Man style idiot-savant who can do very complex math in their heads, but can't take care of themselves. Are they more intelligent than you? In mathematics, sure. But now you have a chip in your head, and that hardware can do the complex math for you. Did you become more intelligent? Chess programs are unbeatable by our best chess Grandmasters now, are they intelligent? We used to claim that would be a big milestone in hard AI, but it turned we can make a really dumb, but specialized machine for that task. So we moved the goal posts. I had the opportunity to play Lubomir Ftacnik in a group game when he came to my University's chess club, and the most impressive part wasn't that he beat everyone in the simultaneous game. It's that he sat down with each person later and went over their game with them, talking abut different things they could have done. Every move. From memory. I've later come to find out this is common among chess Grandmasters. Is that type of chess eidetic memory sufficient to make someone into a grandmaster? No. Is it beneficial? Apparently so, since so many of them have it. And that's easy to supplement you with.
You can't give just anyone access to information and suddenly expect them to be smart. Stupid is as Stupid does. There are lot of smart phone running around this country and all I have to do is flip on the news for 5min to confirm its NOT making people wiser, if anything they are just letting people do more stupid faster.
Yeah, but now it's not a smartphone, it's in your brain. What if you don't see it as data, but see it as thought. When you're playing chess, it computes a bunch of possible moves, but you're not consciously aware of all those choices, just the ones it deemed best and fed them back to you. When you make a decision, it computes which has the best expected value, and now you get a "good feeling" about the option with the best expected value. This is the best case scenario, imagine the dystopian one. Instead of getting a good feeling about the option with the best expected value, you get a good feeling about the option the chip manufacturer gets paid to promote.
Basically, it's not the question of what is intelligence and what is not that bothers me. I think the chip can be undoubtedly be used to make us all either more intelligent or sufficiently emulating higher intelligence, whatever that means. What bothers me is that this is root access to my personality. I can be subtly controlled, not know it, and not care once I do find out.
That may be the case in some places, but not all...This has driven up the going rate for farm labor, which is now pushing $12/hr, sometimes going as high as $15/hr. The demand for labor and decent pay has brought in an influx of Mexican workers, and definitely not all of it them are legal.
When the demand for labor is outpacing the supply of labor, that's a good argument for increasing the number of visas we grant. I still think we have the same problem with hiring illegal immigrants. Even if they're earning the same hourly rate as legal workers, these employers will still prefer them for the lack of payroll taxes and other employment costs. The employees are still not paying into social security and have incentives to avoid reporting workplace abuse for fear of deportation.
But you're right in that I didn't touch upon the case where illegal immigrant labor is *not* competing against full benefit job position because the demand for it as high. Congress should certainly consider whether our current immigration limits might be too low and whether we need to raise them.
Why? Because they're hard working, they stay out of trouble, and they don't leave for greener pastures as soon as they see a potentially better option.
You don't see this is a problem? Have you asked yourself why someone wouldn't leave for a better opportunity when one presents itself (statistically speaking, that is. Of course individuals may be satisfied with what they have and lack ambition, but when you talk about an entire group of people as having that trait, you should look closer). Once you answer that question, you might ask yourself what those employers mean by hard working.
When I was an undergraduate at university, there was a group of graduate students under a South American professor known to work nights weekends, and basically not having a life outside of school. It was later found out he had a penchant for hiring people on student visas, and threaten them with revoking that visa if they didn't work hard enough. Could he do that? Of course not. They didn't know that. For the longest time no one reported him, because hey feared repercussions. They couldn't easily change advisors, like a US or permanent resident could. And their apparently hard-working nature wasn't a case of better work ethic than Americans, as it was a case of them thinking they didn't have the same labor protection as Americans.
Immigration, legal and illegal, is benefiting both our countries in this regard
I welcome hard working immigrants with open arms, and I see the benefits of even illegal immigration to our society just as you do. I agree with everything you've said. My problem is that I also see the danger of looking the other way to reap those benefits by allowing a labor black market. But sure, if we think we need to increase current immigration limits, let's do it.
Sure, you can take the free market approach of "if they're not profitable, let them die", but that's the same as saying, "I hate Mexicans so much I want to see both our countries poorer."
Well, I think if a business is not profitable because there's no market for it, it should die (or really, we should direct a pivot into a different field: I don't want the people owning those business to go bankrupt and for people to lose jobs, I want the market to shift into something different that works, and we can structure a system to help them do that). If a business is not profitable because the supply of labor is too low and current government regulations are keeping us from importing labor, that's a problem with the current regulations. We can fix that cheaply and easily.
Would you be supportive of immigration controls that are effective, such as random ID checks and fines for employers of illegal immigrants?
I'm very much in favor of cripplingly high fines for employers of illegal immigrants. The way I see it, they are the cause of the biggest problems with said illegal immigration. If employers are hiring illegals instead of Americans, they're doing so because they can hire them for less than minimum wage while not paying for required benefits and employment taxes. This creates a second-class citizen situation: yes, we get cheaper products, but we do so because we're supporting a type of slave labor where illegal immigrants are forced to earn significantly below the cost of living for their region, which is why you see them having to group up several families in a one-family house. They don't complain about any abuse or safety violations at their work place because they fear deportation is found out.
Random ID checks, not so much. It's unconstitutional to perform a warrant-less search, and this is what it amounts to. If you have cause to perform a check on someone's resident / citizenship status, then you perform it, such as when hiring a new employee.
That said, I'm not a Democrat. So gauging my opinion on the above isn't a representative sample of that if it's what you're looking for.
Besides, even if it's not perfect, a one-time $5 billion is peanuts compared to the cost of hosting illegal immigrants. Even the liberal politifact says the costs is between $43 to $279 billion per year [politifact.com]. Over the lifetime of the wall, which is probably 20 years or more, that's 0.0008% to 0.005%. So the wall only has to be 0.005% effective to save us money, which it certainly will be. Heck, even Trump's rhetoric about building the wall is more than 0.005% effective.
The $5 billion isn't for a complete border wall. It's what he's asking to build a section of it right now. The estimates are at $25 billion. And it's not a sunk cost. It's not like you build the wall then don't do anything for 20 years. You have maintenance, you have patrolling. Smugglers build tunnels to get past existing patrols. People vandalize existing barriers to get through right now.
Most importantly, even if you're right, and it would save us money, as I've stated, it's not the hot path for illegal immigration. If you apply those funds elsewhere, you can save more money. Trump talked about the cost of drug trafficking, but the majority of that cost would go away for free if we ended the drug war and just legalized all drugs. That would do away with enforcement costs, leaving only the societal costs. Taxation of those sales can be used to offset those societal costs.
Basically, it's not just a question of whether there are positive gains from investing the money on a border wall. Even if we have the money, there's an opportunity cost to not investing it someplace else with a higher return on investment. You'd think a businessman like Trump would understand that concept.
Which means we should be spending less and taxing more, not spending $5 billion on something that's not even effective at preventing what it's designed to prevent. Considering the majority of illegal immigration is a result of overstayed visas, the majority of the drug traffic happens at legal ports of entry, and the known or suspected terrorists that have attempted to enter the US have been caught at airports, it seems to me the return on investment on building a border wall is nowhere near the $5 billion he's asking for.
This is slashdot, so in programmer terms, this is premature optimization. Sure, there are thousands of people illegally crossing the border, but if you actually run a profiler you'll see that's not the even the hot path, so why are we proposing to spend so much of resources on that instead of elsewhere?
You're not wrong that they're pushing for it due to business opportunities, but telemetry and advertising aren't the only business opportunities that come with fast internet. Everybody is hopping on the streaming media bandwagon. And if you're Microsoft instead of Netflix or Hulu, they provide services like azure and skype, office 365, all of which benefit from broadband.
You're not wrong that increased access to broadband is good for megacorporations. You're way off base to imply it's one-sided, and especially in the implication that their benefit is entirely for services that go against the end-user interest.
"by the 24th century, nobody will care." -- bwhaha how naive can you be to consider such a thing.
It's a fictional show set in an utopian future (at least during the Roddenberry era, before DS9). Naive optimism was the point.
Nobody thinks prejudice will be completely eradicated from human society, but those of us who wish for that future work toward it. Roddenberry did it by showing his vision of what such a society could be like. To be fair, I'd say we have been getting better, and one would hope the trend will continue. The world is certainly better for ethnic minorities, homosexuals, and even bald men today than it was in the 1950s, for instance.
Not even that is going to be simple. There are some deaf people that don't want their own deaf children to get cochlear implants [theatlantic.com] or to have other types of procedures that could restore their hearing. If they don't want that, odds are they won't accept a genetic fix to prevent the problem from developing in the first place.
That's an individual decision, and it's up to the parents. If they don't want to consider a genetic fix because they don't consider it a problem, that's fine, but it shouldn't stop the research from helping others who do want to correct the issue in their children.
I personally see the decision to not add a genetic fix as more ethically ambiguous than the research for a fix. You could remove one stumbling block from your child's life and chose not to. That said, if you're confident you can look your adult child in the eye and explain your reasoning with conviction if they ever come asking you chose not to, then, again, it's your choice. All you can do as a parent is make the choices you think are right.
Suppose for sake of argument that sexual preference has a genetic control (I don't believe that this is the case, but this is for the sake of argument) and some parent doesn't want (or does, as some people today may well do) their child to be a homosexual. Is that something that's permissible to "fix"?
I give the same answer here as the rumored answer to why Jean-Luc Picard is bald in the 24th century: surely they have come up with a cure by then. The answer is, "by the 24th century, nobody will care."
Ideally that's just not something parents will care to change. If they do care, it's probably better for the child to go ahead and make the change, instead of setting them up to grow up homosexual in a family that is non-accepting, and all the psychological issues that would come with that.
Exactly. The only thing I'd like to add is a slight comment on this one sentence:
If you think Peer review catches mistakes then you need to learn more about peer review because that's not what it does.
To be fair, it does. The problem here is the assumption that peer review refers only to the reviewers who look at unpublished manuscripts before accepting them for publication. That's the first level of it, but it's only going to catch the most obvious issues with methodology, lack of sufficient literature review, conclusions that aren't well supported by provided data, etc. It's what you get when somebody in your basic field spends an hour reading your paper.
The bulk of peer review happens after the fact. The paper is published, it's read by a much larger pool of scientists in the field, many with competing theories that can offer a different viewpoint and analysis. Other research groups attempt to reproduce any experiments and can publish confirmation or inability to reproduce, etc.
It doesn't always go this smoothly: A lot of papers end up not getting published in prestigious journals like Nature, so not enough eyes look at every paper. Journals have a bias against negative results, so it's harder to publish papers that just reproduces somebody else's work and confirms they got the same result (unless it's a hot topic and/or controversial result), so not as much work gets done in that area as there should be done. That said, the basic process is sound, and the thing to take away from stories like these isn't, "climate scientists are wrong." It's, "climate scientists are the ones that point out when they are wrong after further review, because that's what scientists do. They're not protecting an agenda, and they're constantly looking for errors in each others' work."
Pascal's Wager is trivially rebutted: not believing in a god is not any more dangerous than choosing the wrong religion assuming god or gods exist. Or choosing the right religion, but interpreting it incorrectly.
For instance, let's assume you're Christian. Also, let's assume your God is real. However, let's assume Judaism is the right religion. Well, both Judaism and Christianity share the Ten Commandments, so you're well aware that "thou shalt have no other gods before me." If Judaism is true, as a Christian your belief that Jesus is God violates that commandment, and now you're fucked.
Out of all the religions out there, and all the contradicting interpretations of the same religious texts, there's no "safe" choice. It's just as easy to be "wrong" being religious as it is being an atheist.
I have nothing against having faith, and if your basis for picking your religion is simply that you choose to believe it's true, more power to you. But do so out of that faith, not out of a belief that you have any evidence to draw a logical conclusion.
If morality is a human construct, then it is arbitrary.
I never understood that argument. First, there are certain biological imperatives we've evolved that are universal to humans because they benefited our ancestors. Morality is largely a result of humans having empathy, which is a useful thing in a social species that survives through collaboration.
Second, by your definition of arbitrary, it's also arbitrary if there is a God. Whatever He decided is good is, but it was essentially up to him. You can use the argument that God is incapable of evil, but that is either a circular definition (God's morality is good because everything from God is good), or an arbitrary decision, "if my personal morality disagrees with that of God, then I'm going to take God's morality to be the correct one instead."
In the end, I am a moral indeed a moral relativist, in the sense that I believe there is no universal good or evil. However, I do have a strong moral code, which I adhere to for biological reasons (the empathy I have as a result of human biology means I don't like to see others suffer), cultural reasons (I was raised in a society with specific moral beliefs and was taught them from a young age), and personal reasons (I've reasoned new morals as a result of moral axioms developed from the above biological and cultural, as well as to resolve inconsistencies between them).
he title is total clickbait.
From the article: "Then we tracked everyone’s actual performance on the tasks. Here we found that on average, having meditated neither benefited nor detracted from a participant’s quality of work."
That's also in the summary, but the point is the quality of the work didn't change, but the quantity of it did, because they lacked motivation.
Which shouldn't be surprising to anyone. There's a reason you're not in a relaxed state when you have a looming deadline: it's not beneficial to meeting the deadline. Also, there's a reason people abusing drugs to get shit done use amphetamines and other stimulants. You want to be focused and hyped, not calm and relaxed.
Perhaps a female director could clarify for me how a female character feels about a male character by directing the cameraman to do a long, slow scan up the male's body, which a male director would not think to do, since he doesn't classify ogling males as entertainment.
You know, that's a logical hypothesis, but interestingly, the evidence doesn't pan out, at least not in the Superhero genre.
All three Thor movies, Man of Steel, and Justice League were directed by men. They each had long, slow, panning shots of the titular muscular character, shirtless.
All the Jessica Jones episodes were directed by women. Most of the sex scenes in Jessica Jones are advance the story, but there were also plenty of scenes that were very much in the "sex sells" category. In this last season for instance, I remember a scene which obviously existed for the sole purpose of, "let's have Krysten Ritter in her underwear."
Basically, I don't think there's any evidence of directors of either gender being unable to successfully shoot from a point of view that is welcomed by the opposite gender. I think any capable director should, in fact, excel at seeing things from a different point of view than his or her own.
That said, that doesn't mean there is no prejudice in the business. There might very well be an unfair bias against female directors in the business. People often argue that women are naturally not as likely to navigate toward tech fields, but this is an artistic field. That's the field those same people argue women gravitate to. I don't like the idea of quotas per se, but I don't think anyone can argue that anytime we see a disproportionate bias toward or against a particular group, it's important to study it and understand why it exists. Maybe it's just a natural preference people have, and that's fine. Maybe it's discrimination, and that's not.
Well yeah clearly you do not understand the free market, supply and demand, PR=B$ and greed. You are not charged a reasonable price based upon reasonable costs on anything what so you. You are charged the highest possible price, where increasing it further would actually reduce profit.
Which, in a competitive market, approaches the production cost, because margins are a market inefficiency.
If I'm the only one selling you a product you need, the answer to your question, "how much?" is "how much you got?". Which is bad, and why we should avoid monopolies and have strong anti-trust legislation. However, if I'm one of multiple vendors for an equivalent product, the moment I do that, the others will start a price war. Barring other concerns such as collusion or loss leader products, none of us are going to sell it to you at a loss, so at some point we're going to compete on other things such as convenience, quality, aesthetics, whatever it is you value. In other words, you'll get the best possible product for the best possible price.
When that doesn't happen, it means some of the requirements aren't being fulfilled (such as no competition, or the consumers don't have perfect knowledge, or a variety of other things). In other words, we're operating at a mode that's not going to self-regulate to the point we want it to, so government regulation that corrects these issues and make it more like a free market is necessary. Sometimes that's just not feasible, I agree: people who want private roads don't really grasp the fact that you're not going to get ten different companies building parallel roads and allow you to pick the lowest toll for the best maintained road. For things like health care, it's feasible, but we need more insurance company competition, more hospital competition, and regulations that force hospitals to publish price information for procedure as well as their success rates to give consumers (and insurance companies negotiating for price) information they need to make decisions.
That said, tariffs are not what you're looking for in terms of government interference. They just don't work at all, because they're not taking you closer to that self-regulating market. The other countries aren't going to do what you want to reduce the tariffs, they're just going to implement tariffs in other goods. Not to mention the costs for what you produce locally is going to go up, which means your exports become less competitive, your buying power decreases, which decreases production, causes people to lose jobs...basically it screws with comparative advantage. They also increase diplomatic friction. Trade partners have an interest in being friendly, which is stronger the more their economies are interdependent. Tariffs decrease the volume of trade, which decrease that interdependence / force people to search for other trade partners. Which means they don't care about keeping you happy anymore.
Someone suggested that this is a brute-force attack (and TFA even hints at that). I don't buy that, because a brute-force attack involving opening up the phone would be nothing really new. I expect they are exploiting a vulnerability.
Apple does the maximum number of wrong attempts before deleting the contents of the phone thing. So I'm pretty sure the vulnerability is being able to stop that from happening (and likely adding the code to do to the brute forcing). The brute force code is to unlock. By default, iOS has a 4-digit pin to unlock. That's really easy to brute-force in no time at all if you can the software input the numbers and get around the maximum number of retries thing, so no reason to even try something other than brute-forcing.
The part that gets me thinking is that the firm is run by an ex-apple security researcher. If he gets to do the above by jailbreaking it through a private key he swiped from Apple and took with him, or a backdoor he coded in himself, then he's in serious legal trouble.
They do work in the same way? Then why can't it recognize sheep?
It is recognizing sheep. The problem is that it's recognizing sheep when it shouldn't.
This is a problem that is very familiar to humans. You see it every time numerologists find "codes" in their holy book that supposedly prophesy real events. Every time somebody finds a shape in a cloud or Jesus in burnt toast.
The neural network went, "I expect to see sheep flocks in pastures, this is a pasture, and there are whitish things here. I see a pattern, so I'm now classifying the whiteish things as sheep."
Then why can't it recognize sheep? A two year old can.
We come with a very specialized hardware to recognize certain shapes. We are AMAZING at recognizing human facial expressions, and a large part of our brain is dedicated just to that. This extends in part to animals, especially other mammals. Here are some things artificial neural networks will kick your ass in.
Wikipedia is an Encyclopedia. By definition it musn't have original content.
Precisely, which is why it shouldn't be cited. Everyone is making this about wikipedia, but the same rule has always been applied: you don't use an encyclopedia as references to cite. You use primary sources.
That doesn't mean encyclopedias aren't useful, or that they shouldn't be used. You want to use them as your first stop, so you can learn enough to know where to focus your research.
Slashdot Mirror
He is untethered, and his rage knows no bounds!
The C# (or Java or JavaScript) programmer who doesn't understand memory management has memory leaks all over his (or her) code (or they are not writing anything substantial).
True, but they'll have a lot less of those leaks. You have to admit that keeping references to objects you no longer need is less of a memory leak problem than forgetting to explicitly free and losing the reference.
I agree with your point, but I don't think you could even get any real work done in C without understanding part of the memory management.
As somebody who specifically works with undergraduate students with a Java background, the first few C++ programs I get from (almost) every new intern while training them is full of "Foo* bar = new Foo();", without the accompanying deletes. The way that happens is pretty clear: they're used to instantiating instances with new, they get a compiler error that indicates they should have an asterix next to they type when they do that, so they add an asterix.
Granted, the process with C would be a bit different because malloc and free don't look like a keyword they're familiar with, but a lot of the same problems apply. Usually allocating memory in a function and returning the pointer which never gets freed. Sometimes they delete a pointer but keep references to it, and the test programs works perfectly ok because the process hasn't overwritten the memory marked for deletion by that point yet, etc.
I love C, but there's no denying that it is a lot easier to create vulnerabilities from ignorance in that language than in a garbage collected language with actual array objects that know their size instead of a head address in contiguously allocated memory, etc. Understanding that helps us train better programmers, because we can hammer that early on.
I would actually be more interested in seeing them break apart into hardware / software / licensed entertainment.
I don't think the mobile computing vs. desktop computer is a big deal. The real issue is that the same people making the hardware are making the software. If you can buy yourself an iphone and decide whether to install android or ios, then it's all good.
If I was an owner of a business there is no way in heck that I would want to belong to a 'scheduling service' that results in a computerized voice calling my employees constantly. What a time waster for the person taking calls.
Which is why they need to not ask permission and stop identifying themselves as a computer, and give it enough voice variation that you can't clue in by that. The entire point of the duplex demo was that it sounded human.
You want me to not use such a service, have online services. If the only way to make an appointment is to make a phone call, and I have a system that will do that for me, I'll be more than glad to shift the burden. If there's on website where I can do it, that's better, but if you want to remain in the 20th century, that's your problem.
How does anyone know it's a joke until they arrest the kid and question him and investigate?
Yeah, question the knothead. And investigate as needed
Although I do agree with you that when that questioning and investigating is the first step is you believe there's a possible danger there, I do fear that as a whole our society is getting very bad at understanding the nuances of context. Well, there have always been people with trouble doing that, but it's a bigger issue when police has that problem.
It's incredibly clear the post was making fun of the fact Siri's lack of understanding of the question and its implications led it to give it a response that was apparently a willingness to help a potential school shooter. Ironically, the kid found out human beings can be just as stupid as Siri regarding natural language comprehension. Siri understood his school shooting request as just someone looking for a school. People understood his humorous post as a potential threat.
I know we have a path for improving natural language understanding in AI software. How do we tackle improving humans (or at least governement organizations made up of humans)? Because that's the real problem here.
Yes, they started it in 2014. It was big in the news last year, as people started being unable to travel due to their poor social rating.
Well, The Orville episodes are deliberately social commentary. That episode was clearly inspired by China's new social credit system as well as the US latest propensity to treat popular opinion as facts.
Exactly Musk (and he is smart enough to know better) is conflating intelligence with access to information. They two are not the same.
Although you are correct, it's more difficult to draw the line than you'd think. You show up to your grandparent's place, they ask you to fix the computer. Maybe they have a Mac, and you don't use a Mac. But you have experience, you know your away around computers, you know how to Google. And you have their problem fixed up in a jiffy. They immediately say, "our grandson, he's so smart..."
Turns out knowledge, experience, and access to information allow you to accomplish things that people without those things can't accomplish. Similarly to how greater intelligence might let you accomplish things others can't. I say might, because a "dumber" person who applied themselves to acquire a lot of knowledge and experience will accomplish more than a highly intelligent person who is lazy and lacks ambition. And now you ask, "is the other person more intelligent, because they planned their life better?"
In some ways, the answer is yes, and now we're getting into different kinds of intelligence. You take a Rain Man style idiot-savant who can do very complex math in their heads, but can't take care of themselves. Are they more intelligent than you? In mathematics, sure. But now you have a chip in your head, and that hardware can do the complex math for you. Did you become more intelligent? Chess programs are unbeatable by our best chess Grandmasters now, are they intelligent? We used to claim that would be a big milestone in hard AI, but it turned we can make a really dumb, but specialized machine for that task. So we moved the goal posts. I had the opportunity to play Lubomir Ftacnik in a group game when he came to my University's chess club, and the most impressive part wasn't that he beat everyone in the simultaneous game. It's that he sat down with each person later and went over their game with them, talking abut different things they could have done. Every move. From memory. I've later come to find out this is common among chess Grandmasters. Is that type of chess eidetic memory sufficient to make someone into a grandmaster? No. Is it beneficial? Apparently so, since so many of them have it. And that's easy to supplement you with.
You can't give just anyone access to information and suddenly expect them to be smart. Stupid is as Stupid does. There are lot of smart phone running around this country and all I have to do is flip on the news for 5min to confirm its NOT making people wiser, if anything they are just letting people do more stupid faster.
Yeah, but now it's not a smartphone, it's in your brain. What if you don't see it as data, but see it as thought. When you're playing chess, it computes a bunch of possible moves, but you're not consciously aware of all those choices, just the ones it deemed best and fed them back to you. When you make a decision, it computes which has the best expected value, and now you get a "good feeling" about the option with the best expected value. This is the best case scenario, imagine the dystopian one. Instead of getting a good feeling about the option with the best expected value, you get a good feeling about the option the chip manufacturer gets paid to promote.
Basically, it's not the question of what is intelligence and what is not that bothers me. I think the chip can be undoubtedly be used to make us all either more intelligent or sufficiently emulating higher intelligence, whatever that means. What bothers me is that this is root access to my personality. I can be subtly controlled, not know it, and not care once I do find out.
That may be the case in some places, but not all...This has driven up the going rate for farm labor, which is now pushing $12/hr, sometimes going as high as $15/hr. The demand for labor and decent pay has brought in an influx of Mexican workers, and definitely not all of it them are legal.
When the demand for labor is outpacing the supply of labor, that's a good argument for increasing the number of visas we grant. I still think we have the same problem with hiring illegal immigrants. Even if they're earning the same hourly rate as legal workers, these employers will still prefer them for the lack of payroll taxes and other employment costs. The employees are still not paying into social security and have incentives to avoid reporting workplace abuse for fear of deportation.
But you're right in that I didn't touch upon the case where illegal immigrant labor is *not* competing against full benefit job position because the demand for it as high. Congress should certainly consider whether our current immigration limits might be too low and whether we need to raise them.
Why? Because they're hard working, they stay out of trouble, and they don't leave for greener pastures as soon as they see a potentially better option.
You don't see this is a problem? Have you asked yourself why someone wouldn't leave for a better opportunity when one presents itself (statistically speaking, that is. Of course individuals may be satisfied with what they have and lack ambition, but when you talk about an entire group of people as having that trait, you should look closer). Once you answer that question, you might ask yourself what those employers mean by hard working.
When I was an undergraduate at university, there was a group of graduate students under a South American professor known to work nights weekends, and basically not having a life outside of school. It was later found out he had a penchant for hiring people on student visas, and threaten them with revoking that visa if they didn't work hard enough. Could he do that? Of course not. They didn't know that. For the longest time no one reported him, because hey feared repercussions. They couldn't easily change advisors, like a US or permanent resident could. And their apparently hard-working nature wasn't a case of better work ethic than Americans, as it was a case of them thinking they didn't have the same labor protection as Americans.
Immigration, legal and illegal, is benefiting both our countries in this regard
I welcome hard working immigrants with open arms, and I see the benefits of even illegal immigration to our society just as you do. I agree with everything you've said. My problem is that I also see the danger of looking the other way to reap those benefits by allowing a labor black market. But sure, if we think we need to increase current immigration limits, let's do it.
Sure, you can take the free market approach of "if they're not profitable, let them die", but that's the same as saying, "I hate Mexicans so much I want to see both our countries poorer."
Well, I think if a business is not profitable because there's no market for it, it should die (or really, we should direct a pivot into a different field: I don't want the people owning those business to go bankrupt and for people to lose jobs, I want the market to shift into something different that works, and we can structure a system to help them do that). If a business is not profitable because the supply of labor is too low and current government regulations are keeping us from importing labor, that's a problem with the current regulations. We can fix that cheaply and easily.
Would you be supportive of immigration controls that are effective, such as random ID checks and fines for employers of illegal immigrants?
I'm very much in favor of cripplingly high fines for employers of illegal immigrants. The way I see it, they are the cause of the biggest problems with said illegal immigration. If employers are hiring illegals instead of Americans, they're doing so because they can hire them for less than minimum wage while not paying for required benefits and employment taxes. This creates a second-class citizen situation: yes, we get cheaper products, but we do so because we're supporting a type of slave labor where illegal immigrants are forced to earn significantly below the cost of living for their region, which is why you see them having to group up several families in a one-family house. They don't complain about any abuse or safety violations at their work place because they fear deportation is found out.
Random ID checks, not so much. It's unconstitutional to perform a warrant-less search, and this is what it amounts to. If you have cause to perform a check on someone's resident / citizenship status, then you perform it, such as when hiring a new employee.
That said, I'm not a Democrat. So gauging my opinion on the above isn't a representative sample of that if it's what you're looking for.
Besides, even if it's not perfect, a one-time $5 billion is peanuts compared to the cost of hosting illegal immigrants. Even the liberal politifact says the costs is between $43 to $279 billion per year [politifact.com]. Over the lifetime of the wall, which is probably 20 years or more, that's 0.0008% to 0.005%. So the wall only has to be 0.005% effective to save us money, which it certainly will be. Heck, even Trump's rhetoric about building the wall is more than 0.005% effective.
The $5 billion isn't for a complete border wall. It's what he's asking to build a section of it right now. The estimates are at $25 billion. And it's not a sunk cost. It's not like you build the wall then don't do anything for 20 years. You have maintenance, you have patrolling. Smugglers build tunnels to get past existing patrols. People vandalize existing barriers to get through right now.
Most importantly, even if you're right, and it would save us money, as I've stated, it's not the hot path for illegal immigration. If you apply those funds elsewhere, you can save more money. Trump talked about the cost of drug trafficking, but the majority of that cost would go away for free if we ended the drug war and just legalized all drugs. That would do away with enforcement costs, leaving only the societal costs. Taxation of those sales can be used to offset those societal costs.
Basically, it's not just a question of whether there are positive gains from investing the money on a border wall. Even if we have the money, there's an opportunity cost to not investing it someplace else with a higher return on investment. You'd think a businessman like Trump would understand that concept.
Which means we should be spending less and taxing more, not spending $5 billion on something that's not even effective at preventing what it's designed to prevent. Considering the majority of illegal immigration is a result of overstayed visas, the majority of the drug traffic happens at legal ports of entry, and the known or suspected terrorists that have attempted to enter the US have been caught at airports, it seems to me the return on investment on building a border wall is nowhere near the $5 billion he's asking for.
This is slashdot, so in programmer terms, this is premature optimization. Sure, there are thousands of people illegally crossing the border, but if you actually run a profiler you'll see that's not the even the hot path, so why are we proposing to spend so much of resources on that instead of elsewhere?
Escalators are stairs, not elevators. They're moving stairs. Everybody should be climbing it up.
It's purpose is the exact same as the flat moving walkways in airports. They get you where you want to go quicker.
You're not wrong that they're pushing for it due to business opportunities, but telemetry and advertising aren't the only business opportunities that come with fast internet. Everybody is hopping on the streaming media bandwagon. And if you're Microsoft instead of Netflix or Hulu, they provide services like azure and skype, office 365, all of which benefit from broadband.
You're not wrong that increased access to broadband is good for megacorporations. You're way off base to imply it's one-sided, and especially in the implication that their benefit is entirely for services that go against the end-user interest.
"by the 24th century, nobody will care." -- bwhaha how naive can you be to consider such a thing.
It's a fictional show set in an utopian future (at least during the Roddenberry era, before DS9). Naive optimism was the point.
Nobody thinks prejudice will be completely eradicated from human society, but those of us who wish for that future work toward it. Roddenberry did it by showing his vision of what such a society could be like. To be fair, I'd say we have been getting better, and one would hope the trend will continue. The world is certainly better for ethnic minorities, homosexuals, and even bald men today than it was in the 1950s, for instance.
Not even that is going to be simple. There are some deaf people that don't want their own deaf children to get cochlear implants [theatlantic.com] or to have other types of procedures that could restore their hearing. If they don't want that, odds are they won't accept a genetic fix to prevent the problem from developing in the first place.
That's an individual decision, and it's up to the parents. If they don't want to consider a genetic fix because they don't consider it a problem, that's fine, but it shouldn't stop the research from helping others who do want to correct the issue in their children.
I personally see the decision to not add a genetic fix as more ethically ambiguous than the research for a fix. You could remove one stumbling block from your child's life and chose not to. That said, if you're confident you can look your adult child in the eye and explain your reasoning with conviction if they ever come asking you chose not to, then, again, it's your choice. All you can do as a parent is make the choices you think are right.
Suppose for sake of argument that sexual preference has a genetic control (I don't believe that this is the case, but this is for the sake of argument) and some parent doesn't want (or does, as some people today may well do) their child to be a homosexual. Is that something that's permissible to "fix"?
I give the same answer here as the rumored answer to why Jean-Luc Picard is bald in the 24th century: surely they have come up with a cure by then. The answer is, "by the 24th century, nobody will care."
Ideally that's just not something parents will care to change. If they do care, it's probably better for the child to go ahead and make the change, instead of setting them up to grow up homosexual in a family that is non-accepting, and all the psychological issues that would come with that.
Exactly. The only thing I'd like to add is a slight comment on this one sentence:
If you think Peer review catches mistakes then you need to learn more about peer review because that's not what it does.
To be fair, it does. The problem here is the assumption that peer review refers only to the reviewers who look at unpublished manuscripts before accepting them for publication. That's the first level of it, but it's only going to catch the most obvious issues with methodology, lack of sufficient literature review, conclusions that aren't well supported by provided data, etc. It's what you get when somebody in your basic field spends an hour reading your paper.
The bulk of peer review happens after the fact. The paper is published, it's read by a much larger pool of scientists in the field, many with competing theories that can offer a different viewpoint and analysis. Other research groups attempt to reproduce any experiments and can publish confirmation or inability to reproduce, etc.
It doesn't always go this smoothly: A lot of papers end up not getting published in prestigious journals like Nature, so not enough eyes look at every paper. Journals have a bias against negative results, so it's harder to publish papers that just reproduces somebody else's work and confirms they got the same result (unless it's a hot topic and/or controversial result), so not as much work gets done in that area as there should be done. That said, the basic process is sound, and the thing to take away from stories like these isn't, "climate scientists are wrong." It's, "climate scientists are the ones that point out when they are wrong after further review, because that's what scientists do. They're not protecting an agenda, and they're constantly looking for errors in each others' work."
The safest course is what?
Pascal's Wager is trivially rebutted: not believing in a god is not any more dangerous than choosing the wrong religion assuming god or gods exist. Or choosing the right religion, but interpreting it incorrectly.
For instance, let's assume you're Christian. Also, let's assume your God is real. However, let's assume Judaism is the right religion. Well, both Judaism and Christianity share the Ten Commandments, so you're well aware that "thou shalt have no other gods before me." If Judaism is true, as a Christian your belief that Jesus is God violates that commandment, and now you're fucked.
Out of all the religions out there, and all the contradicting interpretations of the same religious texts, there's no "safe" choice. It's just as easy to be "wrong" being religious as it is being an atheist.
I have nothing against having faith, and if your basis for picking your religion is simply that you choose to believe it's true, more power to you. But do so out of that faith, not out of a belief that you have any evidence to draw a logical conclusion.
If morality is a human construct, then it is arbitrary.
I never understood that argument. First, there are certain biological imperatives we've evolved that are universal to humans because they benefited our ancestors. Morality is largely a result of humans having empathy, which is a useful thing in a social species that survives through collaboration.
Second, by your definition of arbitrary, it's also arbitrary if there is a God. Whatever He decided is good is, but it was essentially up to him. You can use the argument that God is incapable of evil, but that is either a circular definition (God's morality is good because everything from God is good), or an arbitrary decision, "if my personal morality disagrees with that of God, then I'm going to take God's morality to be the correct one instead."
In the end, I am a moral indeed a moral relativist, in the sense that I believe there is no universal good or evil. However, I do have a strong moral code, which I adhere to for biological reasons (the empathy I have as a result of human biology means I don't like to see others suffer), cultural reasons (I was raised in a society with specific moral beliefs and was taught them from a young age), and personal reasons (I've reasoned new morals as a result of moral axioms developed from the above biological and cultural, as well as to resolve inconsistencies between them).
he title is total clickbait. From the article: "Then we tracked everyone’s actual performance on the tasks. Here we found that on average, having meditated neither benefited nor detracted from a participant’s quality of work."
That's also in the summary, but the point is the quality of the work didn't change, but the quantity of it did, because they lacked motivation.
Which shouldn't be surprising to anyone. There's a reason you're not in a relaxed state when you have a looming deadline: it's not beneficial to meeting the deadline. Also, there's a reason people abusing drugs to get shit done use amphetamines and other stimulants. You want to be focused and hyped, not calm and relaxed.
Perhaps a female director could clarify for me how a female character feels about a male character by directing the cameraman to do a long, slow scan up the male's body, which a male director would not think to do, since he doesn't classify ogling males as entertainment.
You know, that's a logical hypothesis, but interestingly, the evidence doesn't pan out, at least not in the Superhero genre.
All three Thor movies, Man of Steel, and Justice League were directed by men. They each had long, slow, panning shots of the titular muscular character, shirtless.
All the Jessica Jones episodes were directed by women. Most of the sex scenes in Jessica Jones are advance the story, but there were also plenty of scenes that were very much in the "sex sells" category. In this last season for instance, I remember a scene which obviously existed for the sole purpose of, "let's have Krysten Ritter in her underwear."
Basically, I don't think there's any evidence of directors of either gender being unable to successfully shoot from a point of view that is welcomed by the opposite gender. I think any capable director should, in fact, excel at seeing things from a different point of view than his or her own.
That said, that doesn't mean there is no prejudice in the business. There might very well be an unfair bias against female directors in the business. People often argue that women are naturally not as likely to navigate toward tech fields, but this is an artistic field. That's the field those same people argue women gravitate to. I don't like the idea of quotas per se, but I don't think anyone can argue that anytime we see a disproportionate bias toward or against a particular group, it's important to study it and understand why it exists. Maybe it's just a natural preference people have, and that's fine. Maybe it's discrimination, and that's not.
Well yeah clearly you do not understand the free market, supply and demand, PR=B$ and greed. You are not charged a reasonable price based upon reasonable costs on anything what so you. You are charged the highest possible price, where increasing it further would actually reduce profit.
Which, in a competitive market, approaches the production cost, because margins are a market inefficiency.
If I'm the only one selling you a product you need, the answer to your question, "how much?" is "how much you got?". Which is bad, and why we should avoid monopolies and have strong anti-trust legislation. However, if I'm one of multiple vendors for an equivalent product, the moment I do that, the others will start a price war. Barring other concerns such as collusion or loss leader products, none of us are going to sell it to you at a loss, so at some point we're going to compete on other things such as convenience, quality, aesthetics, whatever it is you value. In other words, you'll get the best possible product for the best possible price.
When that doesn't happen, it means some of the requirements aren't being fulfilled (such as no competition, or the consumers don't have perfect knowledge, or a variety of other things). In other words, we're operating at a mode that's not going to self-regulate to the point we want it to, so government regulation that corrects these issues and make it more like a free market is necessary. Sometimes that's just not feasible, I agree: people who want private roads don't really grasp the fact that you're not going to get ten different companies building parallel roads and allow you to pick the lowest toll for the best maintained road. For things like health care, it's feasible, but we need more insurance company competition, more hospital competition, and regulations that force hospitals to publish price information for procedure as well as their success rates to give consumers (and insurance companies negotiating for price) information they need to make decisions.
That said, tariffs are not what you're looking for in terms of government interference. They just don't work at all, because they're not taking you closer to that self-regulating market. The other countries aren't going to do what you want to reduce the tariffs, they're just going to implement tariffs in other goods. Not to mention the costs for what you produce locally is going to go up, which means your exports become less competitive, your buying power decreases, which decreases production, causes people to lose jobs...basically it screws with comparative advantage. They also increase diplomatic friction. Trade partners have an interest in being friendly, which is stronger the more their economies are interdependent. Tariffs decrease the volume of trade, which decrease that interdependence / force people to search for other trade partners. Which means they don't care about keeping you happy anymore.
Someone suggested that this is a brute-force attack (and TFA even hints at that). I don't buy that, because a brute-force attack involving opening up the phone would be nothing really new. I expect they are exploiting a vulnerability.
Apple does the maximum number of wrong attempts before deleting the contents of the phone thing. So I'm pretty sure the vulnerability is being able to stop that from happening (and likely adding the code to do to the brute forcing). The brute force code is to unlock. By default, iOS has a 4-digit pin to unlock. That's really easy to brute-force in no time at all if you can the software input the numbers and get around the maximum number of retries thing, so no reason to even try something other than brute-forcing.
The part that gets me thinking is that the firm is run by an ex-apple security researcher. If he gets to do the above by jailbreaking it through a private key he swiped from Apple and took with him, or a backdoor he coded in himself, then he's in serious legal trouble.
They do work in the same way? Then why can't it recognize sheep?
It is recognizing sheep. The problem is that it's recognizing sheep when it shouldn't.
This is a problem that is very familiar to humans. You see it every time numerologists find "codes" in their holy book that supposedly prophesy real events. Every time somebody finds a shape in a cloud or Jesus in burnt toast.
The neural network went, "I expect to see sheep flocks in pastures, this is a pasture, and there are whitish things here. I see a pattern, so I'm now classifying the whiteish things as sheep."
Then why can't it recognize sheep? A two year old can.
We come with a very specialized hardware to recognize certain shapes. We are AMAZING at recognizing human facial expressions, and a large part of our brain is dedicated just to that. This extends in part to animals, especially other mammals. Here are some things artificial neural networks will kick your ass in.
Wikipedia is an Encyclopedia. By definition it musn't have original content.
Precisely, which is why it shouldn't be cited. Everyone is making this about wikipedia, but the same rule has always been applied: you don't use an encyclopedia as references to cite. You use primary sources.
That doesn't mean encyclopedias aren't useful, or that they shouldn't be used. You want to use them as your first stop, so you can learn enough to know where to focus your research.