Sounds like it should have been titled "The man who didn't keep track of when his contract was up." Thank goodness we have computers to help us keep track of that stuff.
CANbus was developed for the automotive environment and it has some worthwhile features there, but for a flexible network where you want some amount of bandwidth? No way.
In the automation world we went through a decade or more of using DeviceNET, which is just a proprietary version of CANbus. Yes it's nice that power goes through the same cable, but you have to set your throughput at 125, 250, or 500 kbit/sec based on cable length and you have to add terminating resistors at both ends. There's only one master and all the rest are slaves, so no peer-to-peer. Drop cables are very limited in length (6'?). The power is limited to 4A, and the grounding/shielding issues... oh my! You could follow all the whitepaper instructions perfectly (checking for ground loops, etc.) you'd have random communication troubles until you did something perverse like disconnect the shield from ground and let it float and suddenly it'd work fine. DeviceNET was sold as perfect for long runs, such as along a conveyor system, but go too far and you're into weird power or grounding issues.
If you wanted a network with the same topology with more off-the-shelf hardware and more flexibility for bandwidth and peer-to-peer, just go with RS-485. That would work well with Arduino hardware because it already has the serial port chip and you can level shift to RS-485 levels with an off-the-shelf adapter. Buy some two-pair twisted-pair cable and run power in the other pair. Pick a wire gauge to suit.
Of course, why not just go with Ethernet, optionally PoE? You will pay more for the device with an Ethernet chip, but they're getting cheaper, and wouldn't you rather run CAT5/6 through your walls than some proprietary thing nothing else uses?
Or they were trained to avoid the element itself. What's more likely, that an insect can "understand the concept of zero" or that it can combine two stimuli (I want to go towards this thing for the sugar, but I want to stay away from the element)?
We're in a ridiculous bubble market right now. Some examples: at our company we're struggling to hire, more than we ever have, and new people are leaving 2 days in because they got a position somewhere else. Yep, it's great for employees, but please realize we're talking completely unskilled employees here. I've talked with people from other places and it seems to be similar everywhere. Also on the skilled side, I just went hunting for a plumber to sell me a new water heater for my house, and one plumber I used before just ignored my request, and then another one gave me a quote and then isn't getting back to me, even though I'm eager to get the job done and pay him. He claims they're very busy. Where my wife works they'd announced they want to increase the size of her department but they've had several unfilled positions for over a year and can't fill them.
This is what happens when the economy starts doing well - it goes into an uncontrolled upward spiral. That's why the government is increasing interest rates, to try to keep inflation from growing. What happens is, since everyone's eager to hire and buy stuff to fulfill demand, they're all willing to pay more and that's why inflation grows quickly. In reality there's not much flexibility in labour so once we get the unemployment rate down low, inflation starts to rise.
Unfortunately efficiency starts to drop. Training new people who are job hopping costs a lot more. Also, companies start to put off routine maintenance because they don't have enough maintenance people and there's a big push to produce more product to fulfill the demand. New capacity can't be brought online this fast because it requires large capital investment, and labour is already scarce.
On the radio I'm starting to hear a lot more advertisements for big loans "even if you have poor credit" and lots more ads for cheque cashing and payday loans. The wording is reminiscent of the ads that were on the radio leading up to the 2008 crash (at that time they were pushing interest-only mortgages). Giving people with bad credit more access to credit is a big red flag. You're dumping more demand into the marketplace (those people immediately spend that money), but the risks of default go way up.
It's also been a relatively long time since the last recession - longer than usual anyway. We're due.
I can't tell you when this is going to burst, but we've been through times like this before and they generally don't last very long. I suggest saving what you can now while times are good, because jobs are likely going to be scarce a few years from now. If you're looking for a job, find one at a company that's been around for a few economic cycles. Then hang on tight.
That movie was released in 2000. It seems the term "battles royal" or "battle royale" was first used in this context as early as 1671. 1.a.a fight participated in by more than two combatants; especially : one in which the last fighter in the ring or the last fighter standing is declared the winner.
One of the biggest sources of ridiculous plastic is K-cups. I bought a re-usable stainless steel insert that works quite well, and I can put bulk coffee grounds in it, but the number of those K-cups society must be putting in the landfill every day is staggering.
If I disable "Ok Google" I have zero confidence that it disables any other always-listening spying function meant to capture keywords for advertising. Remember that Xbox-One cameras were shown to be on and sending info to the internet even while it was just there not doing anything. These things are all designed for spying.
I agree, but what's the difference between that and a cell phone? It's always on, can be always listening (must be since I can say "Ok Google" to wake it up) and has an internet connection and a list of contacts. No doubt it's also looking for keywords in the conversations it hears and logs that for targeting advertising.
You can't ignore the psychological aspect of it. People are willing to use a kitchen knife even if they know they might cut themselves. If I tried to sell you a machine that sliced vegetables for you (faster) and statistically it only cut your fingers the same amount as doing it by hand, I think you'd probably feel uneasy about it. At least you feel in control of the knife. A machine that might just randomly cut you is different, if only inside your head. You can't market a device like that.
Don't get me wrong - there are assembly lines with systems that are very close to 100%. Most manufacturers measure defects out-the-door in the parts-per-million range. They do that by breaking down the problem into small easily understandable problems that can be solved by very reliable processes. In manufacturing we have the ability to amplify the effect of a defect so it's easily detectable, such as by making two components that will only fit together the right way and can't be assembled backwards, or keeping a product fixtured in a nest so we don't have to locate it with a camera or vision system. Unfortunately we don't have that luxury with self-driving cars. They have to work in the ugly real world and so far nobody has accepted the fact that we may have to modify that world to make it work. Personally I see a future where roads and street signs are modified to make it easier for self-driving cars to operate. Roads with more clearly defined features. Signs with unique identifiers on them that are machine readable. Cars with special indicators on them. Self-heating road indicators that work even in a snow-storm.
Yes, that's called the Pareto Principle. It's typically phrased as the "top 80% of your downtime is caused by the top 20% of your problems." I'm not sure that applies here though, because it leaves out the cost of fixing those problems (some might just be fundamental), or just be really, really impractical to solve.
I do industrial automation for a living, since about 2000. There's a certain class of automation problem where getting to a 90% solution is easy, getting to 95% takes a lot of work, and getting to 97% is extremely hard. That is, 90% of the parts coming down the assembly line are easy to categorize correctly, the next 5% you can do with a lot of effort, and so on. Unfortunately that last 2 or 3% are damn near impossible due to problems with how good our sensors are, or how good our algorithms are, or how good our mechanical sorting solutions are.
These problems are notorious for causing run-on projects that slurp up money but never end. That's because your initial effort appears to produce amazing results - 90% with almost no effort. How hard can the remaining 10% be? My first encounter with one of these problems was a barcode-reading system at an industrial facility reading barcoded tags with a camera instead of a barcode reader. The problem was that the barcodes were becoming more worn and faded over time, and management believed that if we used a camera instead of a barcode reader we'd be able to enhance the image, etc., and get a good read because clearly a human looking at the picture can clearly see the bars and the human-readable text below it. This project went on for months, and then years, always creeping closer to 100%, but never making that leap to 100%, having thrown several different engineers at the problem and bringing in outside machine vision specialists.
In most cases these problems come from over-estimating the capability of your sensors. A sensor with a little dirt on it suddenly gives the wrong result, or temperature fluctuations mess up the calibration, or the dreaded, "sensor seems to be giving valid values, but they're just wrong for no reason." Even if your sensor values are reliable, in many cases you'll end up with a measurement that doesn't fall clearly into the known-A or known-B range.
That's where "AI" is supposed to save us, but my limited experience with AI shows it falls into the same class of engineering problem: you can quickly build an AI that correctly categorizes 90% of your input correctly, and then with effort you can improve it and improve it some more, but you'll never reach that always-correct answer.
This is where engineering projects fail, because you can always find a manager or an optimistic engineer who can hand-wave away the ambiguity and say, "humans aren't perfect either" and "we can just keep making the AI better and better." That's convenient when you don't put a physical number on it. How good can you make the AI with the available sensors? We know the sensors are in some ways better than human perception, but in other ways they're worse. In what quantitative ways are they worse, and how are you compensating for that?
If I were going to tackle some problem like this, I'd start with a standardized sensor suite and data format. You can't have everyone developing AI based on proprietary sensor data because it's too opaque. You also need to standardize the system output format (accelerator percent, braking percent, steering value, etc.) Plus you need to standardize the parameters of the vehicle. Once you've got that you need to start collecting and publishing this data in this standard format - hundreds of thousands or millions of test case scenarios available for every researcher to use, and in each case you need to have an expert specify what the correct set of outputs should be (or correct range at least) for each scenario. Then you can develop your AI or algorithms and you can then run these through a test suite so your AI has to pass all of these scenarios before it can be certified. As we have crashes then we add to the list of scenarios, and if you make changes to the AI, it has to pass that new scenario and still pass all the old ones.
I get the sense this is what the companies doing research are trying to do, but how do we validate their product? If their databases are proprietary, and their sensor format and data isn't in a standard format, and we can't run the tests ourselves, then how can we trust their systems? Of course we can't.
1984 was about absolute and total control through fear, whilst Brave New World was all about social engineering. In 1984 there's also some controlling of what people thought too, but Brave New World is much closer.
Campaign finance isn't about the first amendment/free speech. It's about money. In a democracy everyone's supposed to have an equal vote and an equal voice. Money changes that because people with money can buy more influence and drown out their opposition which gives a minority (people with money) more than their equal voice. The first amendment is about making sure the government can't silence you. Campaign finance law is to regulate the power money has to subvert democracy. That's exactly what this story is about.
But a neural net is ultimately just a (complicated) nonlinear function that produces a deterministic output depending on it's input. It's completely algorithmic. The "learning" is tweaking of the function to get results closer to a desired output. I'm not saying AI is impossible, just that we're not "on the cusp."
"we are on the cusp of introducing a new life form; a self-aware AI." - citation needed!
Just because the media and a bunch of silicon valley types are throwing around the acronym AI suddenly doesn't mean we're close to solving any of the fundamental problems of AI research that we've been grappling with over the last half century or more. Artificial neural nets are just algorithmic ways to generate a nonlinear function for classifying things. We've had artificial neural nets for many years, and yes, now we have more computing power than ever, and neural nets do benefit from the increasing scale of parallel computing. We're not going to get to self-awareness anytime soon, unless you use an almost trivial definition of self-awareness in which case computers have already been self-aware for a very long time. Maybe when you say self-awareness you mean consciousness. Nobody in AI research is suggesting artificial neural networks are going to achieve consciousness.
Everyone here seems to think this is no big deal, and that's a bit odd. The main reason it's a big deal is because it skirts campaign finance laws, which encourages American politicians to court relationships with foreign states who want them elected. If you have a foreign state financing your campaign, that's something the American people really should know. More information and more transparency is better.
The next reason why it's a big deal is that Russia's reason for doing this wasn't about getting Trump elected (they reportedly thought it was impossible) -- it was to weaken the US by stoking more hatred amongst Americans. They're encouraging the far left & right and suppressing the middle group. They've actually succeeded in making you all believe that your "side" winning is more important than working together. Divided you fall.
Technically, if I were the programmer, I'd much rather write the software for controlling a flying car than one that drives on roads. Drone software's pretty much a solved problem since it's up, over, down and you have far fewer things that you need to actually detect. Not that I'd want the liability either way.
It's run by Joel Spolsky, and I've met him and as entrepreneurs go he isn't much of an asshole. Also, if it's closed as a Dupe then it'll point you to the original canonical question. The rest of your post is pointless and appears to be incorrect as well. Sorry if that makes you feel unwelcome here.
I don't think you remember what it was like before StackOverflow. Remember ExpertS-exChange.com? Or all the forum threads that end with the person who asked the question saying, "Thanks, I figured it out!"
Slashdot Mirror
Sounds like it should have been titled "The man who didn't keep track of when his contract was up." Thank goodness we have computers to help us keep track of that stuff.
CANbus was developed for the automotive environment and it has some worthwhile features there, but for a flexible network where you want some amount of bandwidth? No way.
In the automation world we went through a decade or more of using DeviceNET, which is just a proprietary version of CANbus. Yes it's nice that power goes through the same cable, but you have to set your throughput at 125, 250, or 500 kbit/sec based on cable length and you have to add terminating resistors at both ends. There's only one master and all the rest are slaves, so no peer-to-peer. Drop cables are very limited in length (6'?). The power is limited to 4A, and the grounding/shielding issues... oh my! You could follow all the whitepaper instructions perfectly (checking for ground loops, etc.) you'd have random communication troubles until you did something perverse like disconnect the shield from ground and let it float and suddenly it'd work fine. DeviceNET was sold as perfect for long runs, such as along a conveyor system, but go too far and you're into weird power or grounding issues.
If you wanted a network with the same topology with more off-the-shelf hardware and more flexibility for bandwidth and peer-to-peer, just go with RS-485. That would work well with Arduino hardware because it already has the serial port chip and you can level shift to RS-485 levels with an off-the-shelf adapter. Buy some two-pair twisted-pair cable and run power in the other pair. Pick a wire gauge to suit.
Of course, why not just go with Ethernet, optionally PoE? You will pay more for the device with an Ethernet chip, but they're getting cheaper, and wouldn't you rather run CAT5/6 through your walls than some proprietary thing nothing else uses?
Or they were trained to avoid the element itself. What's more likely, that an insect can "understand the concept of zero" or that it can combine two stimuli (I want to go towards this thing for the sugar, but I want to stay away from the element)?
We're in a ridiculous bubble market right now. Some examples: at our company we're struggling to hire, more than we ever have, and new people are leaving 2 days in because they got a position somewhere else. Yep, it's great for employees, but please realize we're talking completely unskilled employees here. I've talked with people from other places and it seems to be similar everywhere. Also on the skilled side, I just went hunting for a plumber to sell me a new water heater for my house, and one plumber I used before just ignored my request, and then another one gave me a quote and then isn't getting back to me, even though I'm eager to get the job done and pay him. He claims they're very busy. Where my wife works they'd announced they want to increase the size of her department but they've had several unfilled positions for over a year and can't fill them.
This is what happens when the economy starts doing well - it goes into an uncontrolled upward spiral. That's why the government is increasing interest rates, to try to keep inflation from growing. What happens is, since everyone's eager to hire and buy stuff to fulfill demand, they're all willing to pay more and that's why inflation grows quickly. In reality there's not much flexibility in labour so once we get the unemployment rate down low, inflation starts to rise.
Unfortunately efficiency starts to drop. Training new people who are job hopping costs a lot more. Also, companies start to put off routine maintenance because they don't have enough maintenance people and there's a big push to produce more product to fulfill the demand. New capacity can't be brought online this fast because it requires large capital investment, and labour is already scarce.
On the radio I'm starting to hear a lot more advertisements for big loans "even if you have poor credit" and lots more ads for cheque cashing and payday loans. The wording is reminiscent of the ads that were on the radio leading up to the 2008 crash (at that time they were pushing interest-only mortgages). Giving people with bad credit more access to credit is a big red flag. You're dumping more demand into the marketplace (those people immediately spend that money), but the risks of default go way up.
It's also been a relatively long time since the last recession - longer than usual anyway. We're due.
I can't tell you when this is going to burst, but we've been through times like this before and they generally don't last very long. I suggest saving what you can now while times are good, because jobs are likely going to be scarce a few years from now. If you're looking for a job, find one at a company that's been around for a few economic cycles. Then hang on tight.
Who wants to watch a stream of Farmville, the Sims, or Bejewelled? (I kid, I kid!)
That movie was released in 2000. It seems the term "battles royal" or "battle royale" was first used in this context as early as 1671. 1.a.a fight participated in by more than two combatants; especially : one in which the last fighter in the ring or the last fighter standing is declared the winner.
Battle Royal in pro wrestling goes back at least to Wrestlemania 2 which took place on April 7th, 1986.
Isn't Battle Royale from pro wrestling? Maybe I'm showing my age...
One of the biggest sources of ridiculous plastic is K-cups. I bought a re-usable stainless steel insert that works quite well, and I can put bulk coffee grounds in it, but the number of those K-cups society must be putting in the landfill every day is staggering.
If I disable "Ok Google" I have zero confidence that it disables any other always-listening spying function meant to capture keywords for advertising. Remember that Xbox-One cameras were shown to be on and sending info to the internet even while it was just there not doing anything. These things are all designed for spying.
I agree, but what's the difference between that and a cell phone? It's always on, can be always listening (must be since I can say "Ok Google" to wake it up) and has an internet connection and a list of contacts. No doubt it's also looking for keywords in the conversations it hears and logs that for targeting advertising.
You can't ignore the psychological aspect of it. People are willing to use a kitchen knife even if they know they might cut themselves. If I tried to sell you a machine that sliced vegetables for you (faster) and statistically it only cut your fingers the same amount as doing it by hand, I think you'd probably feel uneasy about it. At least you feel in control of the knife. A machine that might just randomly cut you is different, if only inside your head. You can't market a device like that.
Don't get me wrong - there are assembly lines with systems that are very close to 100%. Most manufacturers measure defects out-the-door in the parts-per-million range. They do that by breaking down the problem into small easily understandable problems that can be solved by very reliable processes. In manufacturing we have the ability to amplify the effect of a defect so it's easily detectable, such as by making two components that will only fit together the right way and can't be assembled backwards, or keeping a product fixtured in a nest so we don't have to locate it with a camera or vision system. Unfortunately we don't have that luxury with self-driving cars. They have to work in the ugly real world and so far nobody has accepted the fact that we may have to modify that world to make it work. Personally I see a future where roads and street signs are modified to make it easier for self-driving cars to operate. Roads with more clearly defined features. Signs with unique identifiers on them that are machine readable. Cars with special indicators on them. Self-heating road indicators that work even in a snow-storm.
Yes, that's called the Pareto Principle. It's typically phrased as the "top 80% of your downtime is caused by the top 20% of your problems." I'm not sure that applies here though, because it leaves out the cost of fixing those problems (some might just be fundamental), or just be really, really impractical to solve.
I do industrial automation for a living, since about 2000. There's a certain class of automation problem where getting to a 90% solution is easy, getting to 95% takes a lot of work, and getting to 97% is extremely hard. That is, 90% of the parts coming down the assembly line are easy to categorize correctly, the next 5% you can do with a lot of effort, and so on. Unfortunately that last 2 or 3% are damn near impossible due to problems with how good our sensors are, or how good our algorithms are, or how good our mechanical sorting solutions are.
These problems are notorious for causing run-on projects that slurp up money but never end. That's because your initial effort appears to produce amazing results - 90% with almost no effort. How hard can the remaining 10% be? My first encounter with one of these problems was a barcode-reading system at an industrial facility reading barcoded tags with a camera instead of a barcode reader. The problem was that the barcodes were becoming more worn and faded over time, and management believed that if we used a camera instead of a barcode reader we'd be able to enhance the image, etc., and get a good read because clearly a human looking at the picture can clearly see the bars and the human-readable text below it. This project went on for months, and then years, always creeping closer to 100%, but never making that leap to 100%, having thrown several different engineers at the problem and bringing in outside machine vision specialists.
In most cases these problems come from over-estimating the capability of your sensors. A sensor with a little dirt on it suddenly gives the wrong result, or temperature fluctuations mess up the calibration, or the dreaded, "sensor seems to be giving valid values, but they're just wrong for no reason." Even if your sensor values are reliable, in many cases you'll end up with a measurement that doesn't fall clearly into the known-A or known-B range.
That's where "AI" is supposed to save us, but my limited experience with AI shows it falls into the same class of engineering problem: you can quickly build an AI that correctly categorizes 90% of your input correctly, and then with effort you can improve it and improve it some more, but you'll never reach that always-correct answer.
This is where engineering projects fail, because you can always find a manager or an optimistic engineer who can hand-wave away the ambiguity and say, "humans aren't perfect either" and "we can just keep making the AI better and better." That's convenient when you don't put a physical number on it. How good can you make the AI with the available sensors? We know the sensors are in some ways better than human perception, but in other ways they're worse. In what quantitative ways are they worse, and how are you compensating for that?
If I were going to tackle some problem like this, I'd start with a standardized sensor suite and data format. You can't have everyone developing AI based on proprietary sensor data because it's too opaque. You also need to standardize the system output format (accelerator percent, braking percent, steering value, etc.) Plus you need to standardize the parameters of the vehicle. Once you've got that you need to start collecting and publishing this data in this standard format - hundreds of thousands or millions of test case scenarios available for every researcher to use, and in each case you need to have an expert specify what the correct set of outputs should be (or correct range at least) for each scenario. Then you can develop your AI or algorithms and you can then run these through a test suite so your AI has to pass all of these scenarios before it can be certified. As we have crashes then we add to the list of scenarios, and if you make changes to the AI, it has to pass that new scenario and still pass all the old ones.
I get the sense this is what the companies doing research are trying to do, but how do we validate their product? If their databases are proprietary, and their sensor format and data isn't in a standard format, and we can't run the tests ourselves, then how can we trust their systems? Of course we can't.
The comment form probably looked like this:
Comment: _______________________________________
Please enter your name: _____
Please enter your address: _____
[ ] Check this box to certify this is really you.
Somehow I don't think this commercial airliner will have quite the thrust-to-weight ratio to pull off the same feat. :)
1984 was about absolute and total control through fear, whilst Brave New World was all about social engineering. In 1984 there's also some controlling of what people thought too, but Brave New World is much closer.
Campaign finance isn't about the first amendment/free speech. It's about money. In a democracy everyone's supposed to have an equal vote and an equal voice. Money changes that because people with money can buy more influence and drown out their opposition which gives a minority (people with money) more than their equal voice. The first amendment is about making sure the government can't silence you. Campaign finance law is to regulate the power money has to subvert democracy. That's exactly what this story is about.
But a neural net is ultimately just a (complicated) nonlinear function that produces a deterministic output depending on it's input. It's completely algorithmic. The "learning" is tweaking of the function to get results closer to a desired output. I'm not saying AI is impossible, just that we're not "on the cusp."
"we are on the cusp of introducing a new life form; a self-aware AI." - citation needed!
Just because the media and a bunch of silicon valley types are throwing around the acronym AI suddenly doesn't mean we're close to solving any of the fundamental problems of AI research that we've been grappling with over the last half century or more. Artificial neural nets are just algorithmic ways to generate a nonlinear function for classifying things. We've had artificial neural nets for many years, and yes, now we have more computing power than ever, and neural nets do benefit from the increasing scale of parallel computing. We're not going to get to self-awareness anytime soon, unless you use an almost trivial definition of self-awareness in which case computers have already been self-aware for a very long time. Maybe when you say self-awareness you mean consciousness. Nobody in AI research is suggesting artificial neural networks are going to achieve consciousness.
Everyone here seems to think this is no big deal, and that's a bit odd. The main reason it's a big deal is because it skirts campaign finance laws, which encourages American politicians to court relationships with foreign states who want them elected. If you have a foreign state financing your campaign, that's something the American people really should know. More information and more transparency is better.
The next reason why it's a big deal is that Russia's reason for doing this wasn't about getting Trump elected (they reportedly thought it was impossible) -- it was to weaken the US by stoking more hatred amongst Americans. They're encouraging the far left & right and suppressing the middle group. They've actually succeeded in making you all believe that your "side" winning is more important than working together. Divided you fall.
Technically, if I were the programmer, I'd much rather write the software for controlling a flying car than one that drives on roads. Drone software's pretty much a solved problem since it's up, over, down and you have far fewer things that you need to actually detect. Not that I'd want the liability either way.
Yeah, here's an example question from today:
how to create app desktop Management queues in c#?
good people I want to create a desktop application, management queues in C #.
What should I do, I have no idea, I would like help
It's run by Joel Spolsky, and I've met him and as entrepreneurs go he isn't much of an asshole. Also, if it's closed as a Dupe then it'll point you to the original canonical question. The rest of your post is pointless and appears to be incorrect as well. Sorry if that makes you feel unwelcome here.
I don't think you remember what it was like before StackOverflow. Remember ExpertS-exChange.com? Or all the forum threads that end with the person who asked the question saying, "Thanks, I figured it out!"