I know insecure people like to imagine the human race as being so technologically advanced that we could affect the entire planet, but we aren't.
Here's the catch: Destruction is easier than construction.
Small insects can seriously damage trees or buildings many orders of magnitude larger than themselves. Their ability to do so has nothing to do with technological advancement. Humans could affect the entire planet even with medieval technology if only there were enough of them. Plastic and fossil fuels simply greatly accelerate the process by having immediate effects.
And let's not forget that there is zero, zilch, no advantage to using cryptocurrencies for anything but crimes
This is a false statement.
Any community that trusts each other just enough to agree on a standard, but trusts none of their members enough to play the role of a central clearing house, or would for other reasons prefer not to have a central control entity, could solve their problem using a blockchain. Currency is only one application.
Note that blockchains actually are completely traceable and visible. The reason why bitcoin et al are considered anonymous is that you cannot automatically link an account to a person, but this fact is independent of the blockchain as a technology.
Even in an oh-shit scenario where it knows it's going to crash, it's all about MINIMIZING DAMAGES.
Completely agreed and agreed as well that reducing your speed as much as possible is the #1 priority since it goes into the damage equation squared.
They desperately want a philosophical debate.
There are two kinds of people involved there. One whom you are probably right about, and the other who actually genuinely wonder about the question. I agree with them that it's an interesting question to ponder. But I agree with you that it's not going to impact the actual making of the AI much.
Oh please, Plato ? You kidding me? I actually studied ancient greek, and I've read Plato in original, so go out.
Consider that "what works" will be dependent on intrinsic values, what that means, and what people ought to value.
No. "what works" is very simple. One thing survives, and the other doesn't. The one that survives more often works, the other dies out. That's how evolution works and it doesn't care at all if you like it or not.
If "evolution wins," why is rape bad?
Because it creates distrust and uncertainty among social animals, which hinder the survival of the group. It also has a positive side (not a moral judgement, a statement about survival) - the fact that it has not been selected out of the gene pool proves that there must be some advantage to it, and that it also exists as a behaviour in the animal kind adds more support.
Why, when striking a pedestrian, a person should not kill all witnesses?
In some countries in South America, drivers have been known in considerable numbers to drive back over a victim a few times more if they hit someone and injured them badly - because the criminal penalty is almost the same, but the civil costs of paying for someone's death is much lower than paying for someone who is crippled for life.
Not judging, just saying. People can be very practical, like it or not.
Isn't that a practical solution to the problem of potential lawsuits?
Witness protection programs exist for a reason. Depending on the crime, offing all the witnesses is, in fact, a strategy that is sometimes employed.
Morality matters. Morality isn't part of engineering. This is the same problem as climate change denial. Scoffing at experts.
No, it is scoffing at one set of experts and pointing to another set of experts who are already working on the solution while the first set is still discussing the implications.
Climate change is just scoffing at experts and praying to idiots.
I do understand where they come from. If you already calculated that you won't stop in time, and will crash into something, but you could steer away from the concrete wall you are currently going to hit, then there is a decision to be made.
Also the reason why engineers are vastly over represented in mass murder incidents of terrorism.
That is simply because they actually know how to build a bomb or make other things work, while most other people who'd like to do it simply can't put their thoughts into actions.
That is the problem.
Evolution would like to have a word with you. In the end, what works, wins. And don't get me wrong, morality is a part of the working set, as evidenced by the fact that societies with a moral system have a good track record on surviving.
But every society, with absolutely no exception, ever is also practical about it. The holy book, or the law book, or the songs of the elders or whatever may proscribe this and that behaviour. In the real world, those rules are interpreted and quite often bended.
Every major religion (and almost every minor one) prohibits killing. And every single one of them has examples in its history where murder was (or still is) accepted by both the believers in general and the spiritual leaders in specific. Or take any less controversial example. I bet you that there is not a single rule in whatever religion you offer me, that was not at least once violated and the violation was generally accepted.
This is the same with AI and morality. Whatever rules you program into the thing, there will always be a case where it follows the rules but still we feel it did wrong.
Practicality will win. That doesn't mean to ignore morality - the engineers who will in the end write the solution are not amoral people. They just understand that it isn't the mathematics of aerodynamics that makes airplanes fly, it is engineers taking the math and manifesting them in a useable form in the real world.
Wow, that's a world record long jump distance in jumping to conclusions.
Nobody said that morality is overrated. But the simple technical facts make it likely that an imperfect system is going to have to make instant decisions based on incomplete information. Practicality will win out over impractical philosophical perspectives.
That is why I said that yes, there will be simple heuristics, but the AI will not go and take age into account, for example (how, exactly it is going to spot the difference between a group of midgets and a group of school children?). It will simply try to not kill anyone, as good as possible.
If system A can make an instant decision to minimize human casualties, while system B can take into account age, gender and positioning in relation to the currently active road rules (i.e. jaywalking), but system B takes a bit more processing time and as a result its overal accident avoidance rate is 0.5% weaker, then system A will win out.
It's a cute experiment with not exactly surprising results (humans prefer humans over animals - who'd have thought?).
But in the end, like the trolley experiment, it is informative and insightful and a bunch of other +5 mod points buzzwords, but the actual solution for the real world will be made by engineers, not by philosophers, and it will almost certainly not involve a "moral decision" subsystem. The primary effort of a practical AI is in making a decision so quickly that it can still minimize damage. Every CPU cycle wasted on evaluating the data in other ways is silly. It will rely for its decision on whatever data its sensors have already provided, and that data will not be in the shape or form of "there are 3 black people with this age range and these fitness indicators in the car, here are their yearly incomes, family relations and social responsibilities. Outside the car we can choose between the river, average temperature 2 degrees, giving the passengers this table of survival probabilities. Or crowd A, here is a data set of their apparent age, social status and survival probabilities. Or crowd B, here is their data set."
This is how the philosopher imagines the problem would be stated to the AI - or to a human in a survey.
But in reality, the question will be more likely something like: "Collision avoidance subsystem. Here's some noisy sensor data that looks like the road ends over there. A bunch of pixels to the left could be people, number unclear. A bunch of pixels to the right also seem to be people, trajectory prediction subsystem has just given up on them because they're running fuck knows where. Estimated time to impact: 0.5 seconds. You have 1 ms to plot a course somewhere or it doesn't make a difference anymore. Figure something out, I need to adjust the volume on the infotainment system and make the crash warning icon blink."
What we will end up with is some general heuristics, like "don't crash into people if you can avoid it" and then the AI will somehow come up with some result, and it will work ok in most cases in the simulator, and then it will be installed in cars.
Note that this is some anonymous source, not an official Apple statement.
I doubt anyone inside Apple would position the Mac Mini for professional use. From a marketing perspective alone, that's ridiculous. ("Mini" and "Pro", which part you don't understand?)
The Mini is still a wonderful device and if they update it, I'll think about getting one again. My last one was the best media station I ever had. I know the Apple TV is meant to be that, but I don't need TV or radio, I just need something that stores and plays my own media, runs VLC, has a small remote control just for start/stop (volume is on the 7.1 stereo anyway) and has HDMI and maybe seperate digital audio output. I don't need or want integration into some ecosystem, I just want a small, inexpensive media player device. It could be that device. We'll see. (right now I'm bringing my notebook into the home cinema. Not a perfect solution because sometimes my wife wants to watch a movie while I'm away).
I didn't know about "Snaps" and read the comments to figure out what it is. Then I laughed.
There used to be a dependencies problem in Linux. Then package managers solved it, and quite well actually. Very rarely some obscure software needs some specific version of some unusual library and you need to compile it yourself. Ok, not a big problem, thanks to proper versioning Linux can have several versions of the same library installed at the same time.
Then I went back into we development a little and I understood. Everything there, especially the abomination called node.js, was dependeny hell. There are kind-of-but-not-really package managers, well basically just reinventions of the 20 year old Makefile concept, and they just-about-most-of-the-time work. But everything comes with its own little princess attitude, nothing is shared globally, no proper versioning...
Of course, if you live in that environment, you come up with an idea like "Snaps" and think you just invented sliced bread.
I might have been dealing with computers for too long, but I fill like some of the old guys who sometimes write articles where they look at todays computing world and essentially say "come on kids, that is what you got? We had that in the 80s."
Next thing, someone will write "Snap Manager" and "Snap Global Library Storage" (probably with snappier names, pun not intended, and ending in consonant-r something) and we're finally full circle.
They are different approaches, but you can find traces of each in the other. For example, once CyC could read by itself, they gave it a lot of input and let it ask questions. This is not unlike a backpropagation network training.
Their "products" consist of the knowledge database and inference engine. Making it actually do something useful is up to the customer.
This.
I've been following the CyC program for almost 20 years. Well, for sufficiently lenient definitions of "following". But anyways... I was thrilled when they finally released a product. I was deeply disappointed when I saw what. And playing around with the free version a bit was even more disappointing.
For the 30+ years that they've put into that, the results are ridiculous.
Deep Learning (the opposite approach to AI) has many applications including,
I disagree that it's opposite to AI.
Deep learning ditches the "meaning" part and simply does large numbers statistics. That's a brute-force attempt substituting correlation for causation. It works unsurprisingly well because a causal relationship will also show a very high correlation, but as the large number of nonsense results show (you know, the white noise pixel fields that get misidentified as giraffes with 99% confidence) the substitution isn't perfect.
and having people invest in factories, fabs, etc is good for your country
That's why the Cayman Islands are full of factories, fabs, etc...
Tax havens don't work by attracting actual companies with actual headquarters and actual production. Many years ago a journalist went to the Cayman Islands to find all those corporate headquarters. He found one building where a hundred or so international corporations share one office. The kind of corporations that have their own streets named after them in their actual corporate locations.
This is all about money and nothing else. Pure money. Not money tied to any productivity, but the same kind of money you use in speculative derivate finance products. Money completely removed from any economic effect.
The correct solution isn't a global tax but rather to charge taxes based on sales in that country.
This. The whole problem only exists because tricky account tricks can move profits around. The more complicated the tax system, the easier it is to find loopholes.
Which is why any additional complication, like the ones proposed in the past decade, will only make the problem worse. Of course, politicians not being idiots (simply corrupt criminals) know this very well. Proposals like this one (of a failed prime minister candidate, i.e. in political terms: The right person to propose something that just might get you a lot of flak, because he has nothing to lose anymore) make the public believe more fairnes is being added, while tax lawyers and corporate heads know that new loopholes are being added to their benefit. Everyone is happy.
If Germany wants to tax the iphone or facebook it should tax the company based on the amount of revenue that company is receiving from its citizens.
It is such a simple solution - if you make your business here, pay your taxes here - that you really wonder how much money changes hands to keep politicians from spotting it.
Now, you can apply some circular logic to show that somehow, in the end, it is always people who pay taxes. Not that I couldn't re-apply the same kind of logic to show that people, in turn, will respond by a) buying less, b) buying cheaper things (which usually have a lower profit margin) or c) turning to other alternatives such as self-made, neighbour-help, etc. -- and voila, I can now prove that any tax on people in the end is paid by corporate profits.
So it's easy to eliminate this problem - eliminate corporate taxes and shift them to
Now that is a kind of logic people need to be paid for having or brainwashed into having. Yes, certainly the solution to tax evasion is to eliminate taxes. How could anyone think that eliminating the intentionally placed loopholes in the tax laws would somehow accomplish the same?
There is no difference - no matter what you tax, in the end a person somewhere pays for it.
Only if you follow the money and stop at the point where a person is involved. You could, of course, follow the money further and end up somewhere else, if you wanted.
The economy is circular. Money moves around. That is why your logic works - you will always sooner or later find a human person somewhere in the loop. But that is also why your logic fails - you are never at the end, because there is no end, so the "in the end" part is the wrong assumption that makes your whole argument false.
There's a shortage of CISOs. I've had several offers from headhunters, and my company has been asked more than once if we can at least temporarily provide an external CISO. I've developed a CISO training program to coach potential candidates inside the companies because it's difficult and costly to find candidates on the job market.
If you are in this sphere and are in the area or want to move, we should talk - tom@lemuria.org
How many russian trolls are active at all? I've heard numbers in the double and tripple digits.
But the Pentagon employs at least 27,000 people for PR purposes. Some of them certainly do advertisement and recruitment, but a lot of them are active on social media and to influence journalists, all to make sure the US military appears in the right light. Budget 2009: 4.7 billion
That begs the question who to trust at all. Certainly the mainstream media image of US politics is no more trustworthy than any troll posting or any fringe lunatic conspiracy theorists - just influenced by different motives and money.
In the end, it is all a show anyways. It just became even less clear who the performers are.
If you are a top-level executive and you are paid a six-digit salary (not barely six-digit, starting with a 1, you know what I mean), then part of that salary is the expectation that you will sacrifice your private life for the sake of the company if needed.
Bad managers (i.e. statistically speaking half of them) believe they can have the same expectation towards people who earn a fraction of their salary.
Good managers understand that one reason they get this salary is that it is their job to make things work with the resources available.
"Crunch" time is almost always an excuse for bad planning, over-eager resource and deadline estimates made but not owed up to by management and, frankly speaking, what the guys really mean when they start the talk is "I need you to work additional to save my ass, because I promised something I couldn't deliver".
The worst is when crunch time is a fixed part of the plan from the start.
---
All that said, there can be real need for crunch time. If not mismanagement but an externally caused crisis happened. If circumstances changed. If a serious problem with everything is discovered too late.
My profession is Information Security. If there is a serious incident, I would expect that certain key people drop everything and come in. And I would strongly recommend the managers above them to give these people massive rewards for doing so. Not just monetary. Pulling someone away from their family on the weekend can only be compensated for by giving them extra free days (paid, for you Americans as that is apparently not self-evident over there).
Because when you buy ads from a company whose business model is ads, you can absolutely certainly trust the measurements and statistics that they are making themselves about the effectiveness of theirs ads delivered through their platform to their users.
Ironically, a famous recent saint, the so-called "Mother Theresa" was strong on theology and much less so in medical care and her houses of the dying were exactly that and nothing else. A few documentaries unveiled that people in there didn't even get basic care. Didn't stop the woman from posing with the stars and gathering millions (much of which nobody has any idea of where it went).
I would say... on the level of boasting about how you got an SCO license to use Linux cheaply and now you're the only person in the room with a legal copy of Linux...
There's a simple thought here that is prominent among thinkers but not easily digestible for most people:
Humans are just a step in the ongoing evolution, and will go the way of the Neanderthal sooner or later. AI or intelligence not based on biology, is the next step. Its main advantage is that it is so much easier for machine-based life to travel the cosmos. You can just power down during the thousand years between solar systems. That is, of course, a very short version of the story, but you get the idea.
Yes, AI will replace us. The primary question is whether it will happen in evolutionary speed, which means we wouldn't notice, or at Moore's Law speed, which means within a few generations?
It is cute when great minds cling to the idea that AI is a threat to humanity and we need to preserve the human species, but evolution is a force based on principles of nature and like gravity doesn't much care what you think of it.
"I can think of no industry in the past 100 years that has improved its safety and security without being compelled to do so by government."
Not in the US, that's for sure.
There is a big culture issue here. In Europe, we like to play it safe and slow. Companies are founded by people not looking for an IPO and early retirement, but those hoping to create a legacy that future generations can continue. Many of our companies, including some of the biggest, are still earned by the family that founded them.
This creates a relatively risk-averse business culture in which opportunities are sometimes not taken. You americans call it "socialism".
The US has a "go big or go bust" attitude. The culture is risk-seeking and failure is considered to be just a detour to eventual success. This leads to every opportunity being exploited, sometimes at considerable risk. We Europeans call it "the next financial crisis just around the corner".
While this manifests in laws and government regulations, it also expresses itself more directly in customer and investor expectations. In information security (my professional field), for example, my mostly european customers expect the reputation costs of a data breach to be much higher than the data justifies it would probably be. But the data is mostly from the US, so... maybe they are right in the end.
Nobody said that engineers are perfect. But at least they have a clear definition of what works and what doesn't.
There's lots of real psychology like confirmation bias, sunken cost fallacy etc.
Yes, there is. There is also all the intro-level psychology about perception and cognition. Very interesting stuff.
If you want to study how and stop them from doing it again, that's economics.
No, that is politics. You don't need economics to understand that the financial crisis was a giant theft / scam. And you don't stop it with economics. You prevent a repetition with politics, police and prison.
I'd actually like to defend gender studies as a subset of psychology too,
You are thinking of a different thing than what is done under this header these days.
Slashdot Mirror
I know insecure people like to imagine the human race as being so technologically advanced that we could affect the entire planet, but we aren't.
Here's the catch: Destruction is easier than construction.
Small insects can seriously damage trees or buildings many orders of magnitude larger than themselves. Their ability to do so has nothing to do with technological advancement. Humans could affect the entire planet even with medieval technology if only there were enough of them. Plastic and fossil fuels simply greatly accelerate the process by having immediate effects.
And let's not forget that there is zero, zilch, no advantage to using cryptocurrencies for anything but crimes
This is a false statement.
Any community that trusts each other just enough to agree on a standard, but trusts none of their members enough to play the role of a central clearing house, or would for other reasons prefer not to have a central control entity, could solve their problem using a blockchain. Currency is only one application.
Note that blockchains actually are completely traceable and visible. The reason why bitcoin et al are considered anonymous is that you cannot automatically link an account to a person, but this fact is independent of the blockchain as a technology.
Even in an oh-shit scenario where it knows it's going to crash, it's all about MINIMIZING DAMAGES.
Completely agreed and agreed as well that reducing your speed as much as possible is the #1 priority since it goes into the damage equation squared.
They desperately want a philosophical debate.
There are two kinds of people involved there. One whom you are probably right about, and the other who actually genuinely wonder about the question. I agree with them that it's an interesting question to ponder. But I agree with you that it's not going to impact the actual making of the AI much.
Oh please, Plato ? You kidding me? I actually studied ancient greek, and I've read Plato in original, so go out.
Consider that "what works" will be dependent on intrinsic values, what that means, and what people ought to value.
No. "what works" is very simple. One thing survives, and the other doesn't. The one that survives more often works, the other dies out. That's how evolution works and it doesn't care at all if you like it or not.
If "evolution wins," why is rape bad?
Because it creates distrust and uncertainty among social animals, which hinder the survival of the group. It also has a positive side (not a moral judgement, a statement about survival) - the fact that it has not been selected out of the gene pool proves that there must be some advantage to it, and that it also exists as a behaviour in the animal kind adds more support.
Why, when striking a pedestrian, a person should not kill all witnesses?
In some countries in South America, drivers have been known in considerable numbers to drive back over a victim a few times more if they hit someone and injured them badly - because the criminal penalty is almost the same, but the civil costs of paying for someone's death is much lower than paying for someone who is crippled for life.
Not judging, just saying. People can be very practical, like it or not.
Isn't that a practical solution to the problem of potential lawsuits?
Witness protection programs exist for a reason. Depending on the crime, offing all the witnesses is, in fact, a strategy that is sometimes employed.
Morality matters. Morality isn't part of engineering. This is the same problem as climate change denial. Scoffing at experts.
No, it is scoffing at one set of experts and pointing to another set of experts who are already working on the solution while the first set is still discussing the implications.
Climate change is just scoffing at experts and praying to idiots.
Well, that would be obvious heuristic #1.
I do understand where they come from. If you already calculated that you won't stop in time, and will crash into something, but you could steer away from the concrete wall you are currently going to hit, then there is a decision to be made.
And hey, soft bodies do soften the impact.
Also the reason why engineers are vastly over represented in mass murder incidents of terrorism.
That is simply because they actually know how to build a bomb or make other things work, while most other people who'd like to do it simply can't put their thoughts into actions.
That is the problem.
Evolution would like to have a word with you. In the end, what works, wins. And don't get me wrong, morality is a part of the working set, as evidenced by the fact that societies with a moral system have a good track record on surviving.
But every society, with absolutely no exception, ever is also practical about it. The holy book, or the law book, or the songs of the elders or whatever may proscribe this and that behaviour. In the real world, those rules are interpreted and quite often bended.
Every major religion (and almost every minor one) prohibits killing. And every single one of them has examples in its history where murder was (or still is) accepted by both the believers in general and the spiritual leaders in specific. Or take any less controversial example. I bet you that there is not a single rule in whatever religion you offer me, that was not at least once violated and the violation was generally accepted.
This is the same with AI and morality. Whatever rules you program into the thing, there will always be a case where it follows the rules but still we feel it did wrong.
Practicality will win. That doesn't mean to ignore morality - the engineers who will in the end write the solution are not amoral people. They just understand that it isn't the mathematics of aerodynamics that makes airplanes fly, it is engineers taking the math and manifesting them in a useable form in the real world.
Wow, that's a world record long jump distance in jumping to conclusions.
Nobody said that morality is overrated. But the simple technical facts make it likely that an imperfect system is going to have to make instant decisions based on incomplete information. Practicality will win out over impractical philosophical perspectives.
That is why I said that yes, there will be simple heuristics, but the AI will not go and take age into account, for example (how, exactly it is going to spot the difference between a group of midgets and a group of school children?). It will simply try to not kill anyone, as good as possible.
If system A can make an instant decision to minimize human casualties, while system B can take into account age, gender and positioning in relation to the currently active road rules (i.e. jaywalking), but system B takes a bit more processing time and as a result its overal accident avoidance rate is 0.5% weaker, then system A will win out.
It's a cute experiment with not exactly surprising results (humans prefer humans over animals - who'd have thought?).
But in the end, like the trolley experiment, it is informative and insightful and a bunch of other +5 mod points buzzwords, but the actual solution for the real world will be made by engineers, not by philosophers, and it will almost certainly not involve a "moral decision" subsystem. The primary effort of a practical AI is in making a decision so quickly that it can still minimize damage. Every CPU cycle wasted on evaluating the data in other ways is silly. It will rely for its decision on whatever data its sensors have already provided, and that data will not be in the shape or form of "there are 3 black people with this age range and these fitness indicators in the car, here are their yearly incomes, family relations and social responsibilities. Outside the car we can choose between the river, average temperature 2 degrees, giving the passengers this table of survival probabilities. Or crowd A, here is a data set of their apparent age, social status and survival probabilities. Or crowd B, here is their data set."
This is how the philosopher imagines the problem would be stated to the AI - or to a human in a survey.
But in reality, the question will be more likely something like: "Collision avoidance subsystem. Here's some noisy sensor data that looks like the road ends over there. A bunch of pixels to the left could be people, number unclear. A bunch of pixels to the right also seem to be people, trajectory prediction subsystem has just given up on them because they're running fuck knows where. Estimated time to impact: 0.5 seconds. You have 1 ms to plot a course somewhere or it doesn't make a difference anymore. Figure something out, I need to adjust the volume on the infotainment system and make the crash warning icon blink."
What we will end up with is some general heuristics, like "don't crash into people if you can avoid it" and then the AI will somehow come up with some result, and it will work ok in most cases in the simulator, and then it will be installed in cars.
Note that this is some anonymous source, not an official Apple statement.
I doubt anyone inside Apple would position the Mac Mini for professional use. From a marketing perspective alone, that's ridiculous. ("Mini" and "Pro", which part you don't understand?)
The Mini is still a wonderful device and if they update it, I'll think about getting one again. My last one was the best media station I ever had. I know the Apple TV is meant to be that, but I don't need TV or radio, I just need something that stores and plays my own media, runs VLC, has a small remote control just for start/stop (volume is on the 7.1 stereo anyway) and has HDMI and maybe seperate digital audio output. I don't need or want integration into some ecosystem, I just want a small, inexpensive media player device. It could be that device. We'll see. (right now I'm bringing my notebook into the home cinema. Not a perfect solution because sometimes my wife wants to watch a movie while I'm away).
I didn't know about "Snaps" and read the comments to figure out what it is. Then I laughed.
There used to be a dependencies problem in Linux. Then package managers solved it, and quite well actually. Very rarely some obscure software needs some specific version of some unusual library and you need to compile it yourself. Ok, not a big problem, thanks to proper versioning Linux can have several versions of the same library installed at the same time.
Then I went back into we development a little and I understood. Everything there, especially the abomination called node.js, was dependeny hell. There are kind-of-but-not-really package managers, well basically just reinventions of the 20 year old Makefile concept, and they just-about-most-of-the-time work. But everything comes with its own little princess attitude, nothing is shared globally, no proper versioning...
Of course, if you live in that environment, you come up with an idea like "Snaps" and think you just invented sliced bread.
I might have been dealing with computers for too long, but I fill like some of the old guys who sometimes write articles where they look at todays computing world and essentially say "come on kids, that is what you got? We had that in the 80s."
Next thing, someone will write "Snap Manager" and "Snap Global Library Storage" (probably with snappier names, pun not intended, and ending in consonant-r something) and we're finally full circle.
They are different approaches, but you can find traces of each in the other. For example, once CyC could read by itself, they gave it a lot of input and let it ask questions. This is not unlike a backpropagation network training.
Their "products" consist of the knowledge database and inference engine. Making it actually do something useful is up to the customer.
This.
I've been following the CyC program for almost 20 years. Well, for sufficiently lenient definitions of "following". But anyways... I was thrilled when they finally released a product. I was deeply disappointed when I saw what. And playing around with the free version a bit was even more disappointing.
For the 30+ years that they've put into that, the results are ridiculous.
Deep Learning (the opposite approach to AI) has many applications including,
I disagree that it's opposite to AI.
Deep learning ditches the "meaning" part and simply does large numbers statistics. That's a brute-force attempt substituting correlation for causation. It works unsurprisingly well because a causal relationship will also show a very high correlation, but as the large number of nonsense results show (you know, the white noise pixel fields that get misidentified as giraffes with 99% confidence) the substitution isn't perfect.
and having people invest in factories, fabs, etc is good for your country
That's why the Cayman Islands are full of factories, fabs, etc...
Tax havens don't work by attracting actual companies with actual headquarters and actual production. Many years ago a journalist went to the Cayman Islands to find all those corporate headquarters. He found one building where a hundred or so international corporations share one office. The kind of corporations that have their own streets named after them in their actual corporate locations.
This is all about money and nothing else. Pure money. Not money tied to any productivity, but the same kind of money you use in speculative derivate finance products. Money completely removed from any economic effect.
The correct solution isn't a global tax but rather to charge taxes based on sales in that country.
This. The whole problem only exists because tricky account tricks can move profits around. The more complicated the tax system, the easier it is to find loopholes.
Which is why any additional complication, like the ones proposed in the past decade, will only make the problem worse. Of course, politicians not being idiots (simply corrupt criminals) know this very well. Proposals like this one (of a failed prime minister candidate, i.e. in political terms: The right person to propose something that just might get you a lot of flak, because he has nothing to lose anymore) make the public believe more fairnes is being added, while tax lawyers and corporate heads know that new loopholes are being added to their benefit. Everyone is happy.
If Germany wants to tax the iphone or facebook it should tax the company based on the amount of revenue that company is receiving from its citizens.
It is such a simple solution - if you make your business here, pay your taxes here - that you really wonder how much money changes hands to keep politicians from spotting it.
Who pays corporate taxes?
Corporations.
Now, you can apply some circular logic to show that somehow, in the end, it is always people who pay taxes. Not that I couldn't re-apply the same kind of logic to show that people, in turn, will respond by a) buying less, b) buying cheaper things (which usually have a lower profit margin) or c) turning to other alternatives such as self-made, neighbour-help, etc. -- and voila, I can now prove that any tax on people in the end is paid by corporate profits.
So it's easy to eliminate this problem - eliminate corporate taxes and shift them to
Now that is a kind of logic people need to be paid for having or brainwashed into having. Yes, certainly the solution to tax evasion is to eliminate taxes. How could anyone think that eliminating the intentionally placed loopholes in the tax laws would somehow accomplish the same?
There is no difference - no matter what you tax, in the end a person somewhere pays for it.
Only if you follow the money and stop at the point where a person is involved. You could, of course, follow the money further and end up somewhere else, if you wanted.
The economy is circular. Money moves around. That is why your logic works - you will always sooner or later find a human person somewhere in the loop. But that is also why your logic fails - you are never at the end, because there is no end, so the "in the end" part is the wrong assumption that makes your whole argument false.
I work in Germany and Austria.
There's a shortage of CISOs. I've had several offers from headhunters, and my company has been asked more than once if we can at least temporarily provide an external CISO. I've developed a CISO training program to coach potential candidates inside the companies because it's difficult and costly to find candidates on the job market.
If you are in this sphere and are in the area or want to move, we should talk - tom@lemuria.org
How many russian trolls are active at all? I've heard numbers in the double and tripple digits.
But the Pentagon employs at least 27,000 people for PR purposes. Some of them certainly do advertisement and recruitment, but a lot of them are active on social media and to influence journalists, all to make sure the US military appears in the right light. Budget 2009: 4.7 billion
That begs the question who to trust at all. Certainly the mainstream media image of US politics is no more trustworthy than any troll posting or any fringe lunatic conspiracy theorists - just influenced by different motives and money.
In the end, it is all a show anyways. It just became even less clear who the performers are.
If you are a top-level executive and you are paid a six-digit salary (not barely six-digit, starting with a 1, you know what I mean), then part of that salary is the expectation that you will sacrifice your private life for the sake of the company if needed.
Bad managers (i.e. statistically speaking half of them) believe they can have the same expectation towards people who earn a fraction of their salary.
Good managers understand that one reason they get this salary is that it is their job to make things work with the resources available.
"Crunch" time is almost always an excuse for bad planning, over-eager resource and deadline estimates made but not owed up to by management and, frankly speaking, what the guys really mean when they start the talk is "I need you to work additional to save my ass, because I promised something I couldn't deliver".
The worst is when crunch time is a fixed part of the plan from the start.
---
All that said, there can be real need for crunch time. If not mismanagement but an externally caused crisis happened. If circumstances changed. If a serious problem with everything is discovered too late.
My profession is Information Security. If there is a serious incident, I would expect that certain key people drop everything and come in. And I would strongly recommend the managers above them to give these people massive rewards for doing so. Not just monetary. Pulling someone away from their family on the weekend can only be compensated for by giving them extra free days (paid, for you Americans as that is apparently not self-evident over there).
Because when you buy ads from a company whose business model is ads, you can absolutely certainly trust the measurements and statistics that they are making themselves about the effectiveness of theirs ads delivered through their platform to their users.
That requires an extraordinary level of stupid.
Ironically, from theology came medical care.
Ironically, a famous recent saint, the so-called "Mother Theresa" was strong on theology and much less so in medical care and her houses of the dying were exactly that and nothing else. A few documentaries unveiled that people in there didn't even get basic care. Didn't stop the woman from posing with the stars and gathering millions (much of which nobody has any idea of where it went).
Quoting the bible. On /.
Now that is courage.
I would say... on the level of boasting about how you got an SCO license to use Linux cheaply and now you're the only person in the room with a legal copy of Linux...
Everybody was right about DNT. News at 11.
Seriously, probably half of us here could dig up some old comment they made where they said exactly this would happen.
There's a simple thought here that is prominent among thinkers but not easily digestible for most people:
Humans are just a step in the ongoing evolution, and will go the way of the Neanderthal sooner or later. AI or intelligence not based on biology, is the next step. Its main advantage is that it is so much easier for machine-based life to travel the cosmos. You can just power down during the thousand years between solar systems. That is, of course, a very short version of the story, but you get the idea.
Yes, AI will replace us. The primary question is whether it will happen in evolutionary speed, which means we wouldn't notice, or at Moore's Law speed, which means within a few generations?
It is cute when great minds cling to the idea that AI is a threat to humanity and we need to preserve the human species, but evolution is a force based on principles of nature and like gravity doesn't much care what you think of it.
"I can think of no industry in the past 100 years that has improved its safety and security without being compelled to do so by government."
Not in the US, that's for sure.
There is a big culture issue here. In Europe, we like to play it safe and slow. Companies are founded by people not looking for an IPO and early retirement, but those hoping to create a legacy that future generations can continue. Many of our companies, including some of the biggest, are still earned by the family that founded them.
This creates a relatively risk-averse business culture in which opportunities are sometimes not taken. You americans call it "socialism".
The US has a "go big or go bust" attitude. The culture is risk-seeking and failure is considered to be just a detour to eventual success. This leads to every opportunity being exploited, sometimes at considerable risk. We Europeans call it "the next financial crisis just around the corner".
While this manifests in laws and government regulations, it also expresses itself more directly in customer and investor expectations. In information security (my professional field), for example, my mostly european customers expect the reputation costs of a data breach to be much higher than the data justifies it would probably be. But the data is mostly from the US, so... maybe they are right in the end.
Nobody said that engineers are perfect. But at least they have a clear definition of what works and what doesn't.
There's lots of real psychology like confirmation bias, sunken cost fallacy etc.
Yes, there is. There is also all the intro-level psychology about perception and cognition. Very interesting stuff.
If you want to study how and stop them from doing it again, that's economics.
No, that is politics. You don't need economics to understand that the financial crisis was a giant theft / scam. And you don't stop it with economics. You prevent a repetition with politics, police and prison.
I'd actually like to defend gender studies as a subset of psychology too,
You are thinking of a different thing than what is done under this header these days.