All undocumented migrants that arrive on our planet are unwelcome by definition. We should assume they are rapists and drug pushers, and deport them immediately.
The tensor processing units Google developed seem also very capable compared to regular processors. Does anyone know how MIT's new chips stack up against what Google already has in operation?
I agree with Bannon that the June 2016 meeting was treasonous. Those present had evidence that Russians had hacked Americans, and were planning to use what they found for their own political purposes. This should have been immediately reported to the FBI. If, as I believe, Trump Senior was aware of what was going on, he was an accessory to treason. Whether this can be proven is another matter.
If both Mueller provides sufficient ammunition, and the 2018 midterms are a disaster for the Republicans, I think the Republican party will calculate that a break from Trump is in their best interests. After all, is replacing Trump with Pence so bad from a right wing perspective? They get the same opportunities to move the Supreme and Federal courts more conservative, while beginning the process of suggesting they are a credible governing party, and not a rabble beholden to a raving lunatic.
If Mueller does not provide overwhelming justification for impeachment, I think it is better if it does not occur. While Trump is a disaster, the long term consequences of a precedent of impeachment for less than clear cut reasons would be worse. It would doom the US to continuous attempts at impeachment without proper justification.
My company received an apology from a Japanese supplier because a shipment of parts were a week late after the 2011 earthquake.
No doubt, that is what Trump is referring to when he talks about other countries' unfair business practices. How can any US company be expected to compete with that kind of service?
Though it seems surreal looking back, I can remember when Internet Explorer 3.0 was actually the best browser around, demonstrably superior to Netscape 3.0 its only real competition! The world, fortunately, does sometimes progress.
This is an example of bad things happening to good people. It was an accident. It seems neither father nor daughter blame Apple. Indeed, I get the impression that Apple acted because they felt the credibility of their rules needed to be protected, not because they thought there was any malice involved in posting the video.
I feel sorry for them. They seem like a fine family.
This whole article depresses me, but the ignorance of suggesting complex, frequently changed passwords as the solution is the biggest indication of someone who does not know what he is talking about. We have known for many years that such an approach is counterproductive, because it virtually guarantees that passwords will be written down and saved in the browser.
In fact, there is a need to get rid of passwords. Approaches like U2F and biometric methods are superior, but far from perfect. Sadly, passwords are here to stay for quite a few years yet. Educating people on how to choose passwords they can remember, stressing the importance of keeping them secure, and not using the same password for multiple important accounts helps. However, passwords remain a weak form of authentication.
That's great. How much? The article claims $1500 for the "base model" whatever that is.
Indeed, I suspect the 15" model with a high end i7 with 16 GB RAM, discrete graphics and 1 TB SSD might cost a tad more. Just look at the difference in price between the cheapeat and most expensive current Surface Books.
The problem with suppressing conspiracy theories, and promoting "authoritative" sources, is that it makes real conspiracies even easier for the authorities to cover up.
It is a difficult problem. I think the suggestion in the article that top ranking posts be reviewed by a human (while that is no guarantee) is good. Theories that challenge the official narrative (usually false, but occasionally true) are nothing new, predating the Internet age. The main difference is the speed at which such theories can be disseminated (and challenged).
Scaramucci strikes me as awfully similar in many ways to Trump himself - a foul mouthed, self centered narcissist. I guess Trump does not like to have others like himself around. What surprises me is that he did not realize what Scaramucci was like before appointing him. I guess, as usual, Trump failed to listen to those around him.
Remember the occasion when a male passenger was allowed a full can of beer, but the female passenger sitting beside him denied a full can of cola for security reasons.
Surely in the systemd era, we should be deprecating setuid on executables, and replacing it with some kind of systemd api. This provides a much more modern "unified" approach then all that minimalist, modular rubbish that infected the system for so long.
Reducing deaths on the roads by 99% will not be sufficient for widespread short-term legalization and adoption of self driving vehicles. People will still point at unlikely complex situations where humans might avoid an accident where the AI would not. Humans, emotionally, do not want to admit that AIs are better drivers than they are, though the best self driving vehicle technologies undeniably are already much better then the average human driver.
What chess playing programs do can pretty much be described as brute force, not to an end of game solution, but choosing the best move based on examining all plausible lines, and using an evaluation function to determine how good each line is.
What is exciting about the (still narrow) AIs developed recently, based primarily on multi level neural networks, is that they can work in situations where no one knows how to create a hand crafted evaluation system. Basically, the system works out for itself what are promising actions, based on its learned knowledge of the probability that each action will lead to desired outcomes. The huge difference is that such an AI can be built without the creators knowing much about the problem domain, or being able to understand the basis upon which the AI is coming up with its solutions. For the most part, the AI is just presented with objectives and large amounts of historical data relevant to the target domain, and works out how to make good decisions in an unsupervised fashion. In some cases, the AI can then refine itself via reinforcement learning where it generates its own data and determines the best solutions. This is still narrow AI, but looks to the observer much more like an independent thinking entity.
Correct. Pretty much all the AI systems now in use are based on narrow AI. In almost any situation where you have sufficient training data, and a limited number of variables, you can develop narrow AIs that will out perform humans on specific tasks. In the specific domain, a modern narrow AI does feel like a super intelligent human, exhibiting intuition and creativity. The ultimate objective of DeepMind is to solve intelligence properly, building artificial general intelligence. To do this, we need to find ways where the available algorithms can apply what has been concluded about domains with lots of training data to novel situation where, by their nature, limited or even no training data is present. We really do not know how we are going to do that. It is possible that one or two breakthroughs could get us there quite quickly. Alternatively, it may require a large number of individual techniques that interact in complex ways . We are probably not going to know until the first successful AGI system is actually built.
I have a lot of sympathy with what you write. However, most large organizations these days do take disaster recover and business continuity somewhat seriously, and have budgets to ensure issues like this are supposed to be addressed.
As others have opined, while some kind of power outage might have precipitated the problems, surely there must have been a cascading series of events that prevented the system from coming back online quickly. Of course, a power outage should not bring down critical servers in the first place: UPS combined with backup generators should keep critical infrastructure online. However, I have seen cases where backup generators were not regularly tested, and failed when really needed. Once servers crash, especially if it is at peak transaction times, you may be faced with time consuming database restarts, with lengthy logs that need to be processed. Even assuming there is a redundant secondary data center with database replication, it only takes a software glitch to prevent that from saving you (especially as such recovery scenarios are rarely sufficiently tested). I doubt we will ever get the complete story, but multiple infrastructure, software and human deficiencies probably combined to generate this fiasco.
It was over 1000 CPUs and over 200 GPUs. That's rather beefy, mate.
Beefy, yes, but nothing extraordinary, and as I have mentioned since reduced by about 90%.
there was definitely an inflection point in Go progress around 2005
Between about 2005 and 2011, significant progress was made. The top programs got to the point where they were competitive with professional players when receiving a (still very big) 4 or 5 stone handicap. (Note, however, that players could exploit weaknesses in computer Go programs once they were studied which makes the achievement a bit less impressive than it seems at first glance.) Between 2011 and 2015, no further progress was made, as the techniques then being applied had reached their limits. The expectation by most in 2015 was that there would be a slow progression of beating average professional players at 3 stone handicap, 2 stone, 1 stone, and eventually level, followed by the ultimate achievement of beating the top professional at even. See Sensei's library page on computer Go which has been updated over time as progress was made, and is not a revisionist account. AlphaGo advanced the state of the art to an extent that shocked the Go community and 99% of the AI community. The leap was not just from beating a middle rank professional with a 4 or 5 stone handicap to beating the top professionals level, it was doing so while having no apparent weaknesses the human player could exploit.
As I tried to explain, the amount of hardware you throw at the problem of paying Go well does not really help. Even the older system that beat Lee Sidol was not running on a humongous supercomputer. Pure computer power is of very limited benefit. Indeed, while dramatically improving the capabilities of AlphaGo over the last year, DeepMind has succeeded in reducing the computing requirements of the system by 90%. It now runs on a single TensorFlow machine (albeit, this is hardware with an architecture tailored to the needs of AI).
As recently as 2010, AI textbooks were typically writing that the field was 20-30 years away from creating a machine that could beat professional Go players. By late 2015, some AI researchers were more optimistic, believing the milestone of beating a top professional might be reached within a decade. Whatever some might have claimed later, very few as late as 2015 were expecting a solution within 5 years. Meanwhile, Go players truly believed a solution was between decades and forever away.
It is instructive (and important in understanding the significance of AlphaGo in overall AI research) to know the important differences between the nature of chess and go that leads to a totally different challenge in playing it well. The most important differences are:
It is very difficult (still impossible and will probably remain so) to hand craft a set of rules to evaluate whether a particular board position is good at go. In chess, just counting up the value of the pieces on the board (counting queen 9, Rook 5 etc.) gives a good rough estimate, that can be refined by recognition of other factors such as passed pawns, king safety and inactive versus active pieces. At go, each stone has equal value (simplistically speaking) and a small change to the position of the single stone can often make a total difference to who is winning, only via effects that occur many moves later.
The branching factor at go is far greater than in chess, even without the challenge of knowing whether a position is good. This means that even examining all possible positions a few moves ahead becomes infeasible. At chess (especially given the previously cited relative ease of writing an evaluation function allowing pruning of obviously hopeless lines) very accurate selection of the most likely best line is possible by Monte Carlo techniques.
Chess programs can have an opening book that records known good early moves (the same in true at go to a lesser extent). However, after that a major difference happens. At chess the position is simplified as pieces are captured. Indeed, once down to about 7 or 8 major pieces, a chess program can use an endgame database to play perfectly without the need for any further calculation. Go, in contrast, is an additive game. The position continues to increase in complexity typically for at least the first 80 moves by each player.
A chess grandmaster can, indeed, explain why a particular move is good, usually by demonstration. Even where the benefits cannot be directly shown, there is established theory known to be sound, to justify it. Actually, a grandmaster cannot improve his knowledge of chess by examining the moves of a chess program that is only superior because of greater calculation and storage capacity
Top go professionals mostly cannot explain in a clearly irrefutable way why certain moves are good. Often, they just need to say they instinctively feel a move is right. There is a 3000 year-old repository of theory (which has been upended twice before in history, first via innovations about 300 years ago, and then again around 70 years ago) but this received wisdom is not known to be totally correct. In fact, the evidence from AlphaGo's play is that much of the existing theory is wrong. The top go professionals find this extremely exciting, as they begin to understand the logic behind AlphaGo's new moves, and the play of these professionals is already changing to incorporate the new knowledge it is allowing them to learn.
There were reasons why AI and go experts believed it would be 20 more years before a go program could best the top professionals. The AI techniques that made it possible are immensely exciting because they are definitely applicable in the area of artificial general intelligence. They are mostly not go specific.
Slashdot Mirror
I guess you are not familiar with the Chinese WeChat, or the TransferWise service.
All undocumented migrants that arrive on our planet are unwelcome by definition. We should assume they are rapists and drug pushers, and deport them immediately.
The tensor processing units Google developed seem also very capable compared to regular processors. Does anyone know how MIT's new chips stack up against what Google already has in operation?
I agree with Bannon that the June 2016 meeting was treasonous. Those present had evidence that Russians had hacked Americans, and were planning to use what they found for their own political purposes. This should have been immediately reported to the FBI. If, as I believe, Trump Senior was aware of what was going on, he was an accessory to treason. Whether this can be proven is another matter.
If both Mueller provides sufficient ammunition, and the 2018 midterms are a disaster for the Republicans, I think the Republican party will calculate that a break from Trump is in their best interests. After all, is replacing Trump with Pence so bad from a right wing perspective? They get the same opportunities to move the Supreme and Federal courts more conservative, while beginning the process of suggesting they are a credible governing party, and not a rabble beholden to a raving lunatic.
If Mueller does not provide overwhelming justification for impeachment, I think it is better if it does not occur. While Trump is a disaster, the long term consequences of a precedent of impeachment for less than clear cut reasons would be worse. It would doom the US to continuous attempts at impeachment without proper justification.
No doubt, that is what Trump is referring to when he talks about other countries' unfair business practices. How can any US company be expected to compete with that kind of service?
Though it seems surreal looking back, I can remember when Internet Explorer 3.0 was actually the best browser around, demonstrably superior to Netscape 3.0 its only real competition! The world, fortunately, does sometimes progress.
This is an example of bad things happening to good people. It was an accident. It seems neither father nor daughter blame Apple. Indeed, I get the impression that Apple acted because they felt the credibility of their rules needed to be protected, not because they thought there was any malice involved in posting the video.
I feel sorry for them. They seem like a fine family.
Just have the delivery people leave your package hidden outside under the doormat
How about this example. Being old is no guarantee that you will be responsible.
This whole article depresses me, but the ignorance of suggesting complex, frequently changed passwords as the solution is the biggest indication of someone who does not know what he is talking about. We have known for many years that such an approach is counterproductive, because it virtually guarantees that passwords will be written down and saved in the browser.
In fact, there is a need to get rid of passwords. Approaches like U2F and biometric methods are superior, but far from perfect. Sadly, passwords are here to stay for quite a few years yet. Educating people on how to choose passwords they can remember, stressing the importance of keeping them secure, and not using the same password for multiple important accounts helps. However, passwords remain a weak form of authentication.
Indeed, I suspect the 15" model with a high end i7 with 16 GB RAM, discrete graphics and 1 TB SSD might cost a tad more. Just look at the difference in price between the cheapeat and most expensive current Surface Books.
It is a difficult problem. I think the suggestion in the article that top ranking posts be reviewed by a human (while that is no guarantee) is good. Theories that challenge the official narrative (usually false, but occasionally true) are nothing new, predating the Internet age. The main difference is the speed at which such theories can be disseminated (and challenged).
He is here to post, not read anything he posts about.
Scaramucci strikes me as awfully similar in many ways to Trump himself - a foul mouthed, self centered narcissist. I guess Trump does not like to have others like himself around. What surprises me is that he did not realize what Scaramucci was like before appointing him. I guess, as usual, Trump failed to listen to those around him.
Remember the occasion when a male passenger was allowed a full can of beer, but the female passenger sitting beside him denied a full can of cola for security reasons.
Surely in the systemd era, we should be deprecating setuid on executables, and replacing it with some kind of systemd api. This provides a much more modern "unified" approach then all that minimalist, modular rubbish that infected the system for so long.
Reducing deaths on the roads by 99% will not be sufficient for widespread short-term legalization and adoption of self driving vehicles. People will still point at unlikely complex situations where humans might avoid an accident where the AI would not. Humans, emotionally, do not want to admit that AIs are better drivers than they are, though the best self driving vehicle technologies undeniably are already much better then the average human driver.
What chess playing programs do can pretty much be described as brute force, not to an end of game solution, but choosing the best move based on examining all plausible lines, and using an evaluation function to determine how good each line is.
What is exciting about the (still narrow) AIs developed recently, based primarily on multi level neural networks, is that they can work in situations where no one knows how to create a hand crafted evaluation system. Basically, the system works out for itself what are promising actions, based on its learned knowledge of the probability that each action will lead to desired outcomes. The huge difference is that such an AI can be built without the creators knowing much about the problem domain, or being able to understand the basis upon which the AI is coming up with its solutions. For the most part, the AI is just presented with objectives and large amounts of historical data relevant to the target domain, and works out how to make good decisions in an unsupervised fashion. In some cases, the AI can then refine itself via reinforcement learning where it generates its own data and determines the best solutions. This is still narrow AI, but looks to the observer much more like an independent thinking entity.
Correct. Pretty much all the AI systems now in use are based on narrow AI. In almost any situation where you have sufficient training data, and a limited number of variables, you can develop narrow AIs that will out perform humans on specific tasks. In the specific domain, a modern narrow AI does feel like a super intelligent human, exhibiting intuition and creativity. The ultimate objective of DeepMind is to solve intelligence properly, building artificial general intelligence. To do this, we need to find ways where the available algorithms can apply what has been concluded about domains with lots of training data to novel situation where, by their nature, limited or even no training data is present. We really do not know how we are going to do that. It is possible that one or two breakthroughs could get us there quite quickly. Alternatively, it may require a large number of individual techniques that interact in complex ways . We are probably not going to know until the first successful AGI system is actually built.
I have a lot of sympathy with what you write. However, most large organizations these days do take disaster recover and business continuity somewhat seriously, and have budgets to ensure issues like this are supposed to be addressed.
As others have opined, while some kind of power outage might have precipitated the problems, surely there must have been a cascading series of events that prevented the system from coming back online quickly. Of course, a power outage should not bring down critical servers in the first place: UPS combined with backup generators should keep critical infrastructure online. However, I have seen cases where backup generators were not regularly tested, and failed when really needed. Once servers crash, especially if it is at peak transaction times, you may be faced with time consuming database restarts, with lengthy logs that need to be processed. Even assuming there is a redundant secondary data center with database replication, it only takes a software glitch to prevent that from saving you (especially as such recovery scenarios are rarely sufficiently tested). I doubt we will ever get the complete story, but multiple infrastructure, software and human deficiencies probably combined to generate this fiasco.
Correction to the URL. Try here.
Beefy, yes, but nothing extraordinary, and as I have mentioned since reduced by about 90%.
Between about 2005 and 2011, significant progress was made. The top programs got to the point where they were competitive with professional players when receiving a (still very big) 4 or 5 stone handicap. (Note, however, that players could exploit weaknesses in computer Go programs once they were studied which makes the achievement a bit less impressive than it seems at first glance.) Between 2011 and 2015, no further progress was made, as the techniques then being applied had reached their limits. The expectation by most in 2015 was that there would be a slow progression of beating average professional players at 3 stone handicap, 2 stone, 1 stone, and eventually level, followed by the ultimate achievement of beating the top professional at even. See Sensei's library page on computer Go which has been updated over time as progress was made, and is not a revisionist account. AlphaGo advanced the state of the art to an extent that shocked the Go community and 99% of the AI community. The leap was not just from beating a middle rank professional with a 4 or 5 stone handicap to beating the top professionals level, it was doing so while having no apparent weaknesses the human player could exploit.
As I tried to explain, the amount of hardware you throw at the problem of paying Go well does not really help. Even the older system that beat Lee Sidol was not running on a humongous supercomputer. Pure computer power is of very limited benefit. Indeed, while dramatically improving the capabilities of AlphaGo over the last year, DeepMind has succeeded in reducing the computing requirements of the system by 90%. It now runs on a single TensorFlow machine (albeit, this is hardware with an architecture tailored to the needs of AI).
As recently as 2010, AI textbooks were typically writing that the field was 20-30 years away from creating a machine that could beat professional Go players. By late 2015, some AI researchers were more optimistic, believing the milestone of beating a top professional might be reached within a decade. Whatever some might have claimed later, very few as late as 2015 were expecting a solution within 5 years. Meanwhile, Go players truly believed a solution was between decades and forever away.
It is instructive (and important in understanding the significance of AlphaGo in overall AI research) to know the important differences between the nature of chess and go that leads to a totally different challenge in playing it well. The most important differences are:
A chess grandmaster can, indeed, explain why a particular move is good, usually by demonstration. Even where the benefits cannot be directly shown, there is established theory known to be sound, to justify it. Actually, a grandmaster cannot improve his knowledge of chess by examining the moves of a chess program that is only superior because of greater calculation and storage capacity
Top go professionals mostly cannot explain in a clearly irrefutable way why certain moves are good. Often, they just need to say they instinctively feel a move is right. There is a 3000 year-old repository of theory (which has been upended twice before in history, first via innovations about 300 years ago, and then again around 70 years ago) but this received wisdom is not known to be totally correct. In fact, the evidence from AlphaGo's play is that much of the existing theory is wrong. The top go professionals find this extremely exciting, as they begin to understand the logic behind AlphaGo's new moves, and the play of these professionals is already changing to incorporate the new knowledge it is allowing them to learn.
There were reasons why AI and go experts believed it would be 20 more years before a go program could best the top professionals. The AI techniques that made it possible are immensely exciting because they are definitely applicable in the area of artificial general intelligence. They are mostly not go specific.