The problem with cyclists is they never obey any rules--then they expect cars to do so and drive at 10mph as they ride down the middle of a lane because 'I'm a vehicle!'
I cycle and this is utter bollocks. I obey the rules, as do many of the people I know that cycle. I do not expect cars to drive at 10mph, but I do expect them to not knock me off. It grinds my gears (ahem) when cyclists don't obey the rules. I would certainly never run a red light except once about twenty years ago when I turned left as I realised I had my toe clips on too tight for a different pair of shoes and couldn't get my feet out the clips, when I turned left and onto the pavement. In the end I had to just stop and fall over then finally get my toes out of the clips at the end of my trip. If I know it's going to be a long light and wish to proceed I dismount, move my bicycle over the pavement to the other part of the junction a reasonable distance from the junction, remount, and move off when it is safe to do so.
I've been knocked off my bicycle twice by cars, and both times it was because drivers didn't see me or look for me, when turning left (this is in the UK) when I was continuing straight, and were initially behind me so I had no way to determine their intention to turn left in the space where I was. Luckily in both cases I was able to pick myself up, straighten my handlebars and carry on, as I was still not fully in the junction by the time they turned left and they just side swiped me. In one instance I did their car much more damage with the metal end grips on my handlebars than they did to me. But it has made me more aware of this possibility when driving.
There are some tearaways (not me!) and occasionally people ride two-abreast on narrow roads, but that's not common and by and large they are pretty good. And in these days of LEDs lights are much, er, lighter for a given level of light, which helps too this time of year.
There's also the Zoe from alliance partner Nissan. The Zoe is sold only in Europe, currently 2.5k/month. So it makes a direct comparison difficult. The Tesla is selling pretty well, though, if you look at market sizes. I couldn't see any obvious figures for total European electric car sales, but in France Zoes are about 60% of the sales, so that suggests total electric car sales in France of about 2k per year (not really very many), but scaling up by population that's about 10k in a Western Europan nation, so that suggests that Tesla is selling better than the Leaf and Renault. 50% of Zoe sales are in France.
I did consider a Zoe a while ago, but the previous version's range of 150 miles was a bit low. The current range of 250 miles would be tempting if I was in the market for a new car, and it had a bit more towing capacity. In a few years, if there's something for a reasonable cost with that sort of range and a decent towing capacity, that's probably what I will get.
Renault charges a fee per month (I presume given Nissan and Renault's close ties, Nissan is probably the same), based on distance travelled, then replace your batteries when required. You are effectively paying for the battery replacement over time, but it also means that the resale value of the vehicle remains good, as the new owner can continue that payment. It's somewhat in the interest of the vendor to make new batteries cheaper, as then the replacement cost is reduced, and they get to bank the difference, but it's also useful for them to be able to cost-effectively extend the lifetime of the battery too, as then the cost/time is reduced for the vendor, but also it can reduce the monthly charge, which would stimulate sales. When I considered buying a Renault when the range was around 130 miles (which in the end I felt would be too limiting) it was of the order of half the cost of petrol for a similar size of vehicle, although that ratio depends on the fuel price at any given time.
Yes, VW, Daimler, BMW, GM, Ford, fiat-chrysler, Nissan, etc have spent the last 5-7 years telling us how they would beat Tesla. Now, like cold fusion, it will happen real soon.
Several of those currently sell electric cars (and Renault) in the non-luxury market. They haven't been selling massively well in earlier incarnations, but the Nissan and Renault ones are now pretty reasonable (230 mile range), which is where they really become competitive with ICE cars. That's not a direct competitor to Tesla as it's a different market segment. BMW, Audi. etc., I would also see as a threat to Tesla outside the USA, but in the luxury market, as they are highly experienced manufacturers with a large and effective workforce. Tesla is probably pretty secure within the USA, though.
Add Kia, Toyota into the etc in the list. And I am sure there are others that are likely to be working on things, such as Vauxhall/Peugot/Citroen.
Looking at the data, households on the top 20% have as few as zero and as many as 5 workers, or more. The average of 2.0 hides a lot, even if the modal value is 2. But $150,000 average with five working might not be impressive per person!
It's not important as to whether they are correct, rather than they make predictions which can be tested. The last I read is that recent evidence didn't seem to be supporting String Theory(ies) but rather the Standard Model, which means science is working as intended.
Traditionally, the activation function was f(W) where W is the set of incoming neuron activations over the synaptic connections, such that a=f(W) where a is an activation level, then the neuron fires if a>A, but now it's more common to use spiking neurons, i.e. a time-series element such that, say, a neuron that's had some pattern of events come in, then fires. That's what SpiNNaker uses, see https://en.wikipedia.org/wiki/... is an initial bit of reading, which includes "SpiNNaker (Spiking Neural Network Architecture), designed at the University of Manchester, uses ARM processors as the building blocks of a massively parallel computing platform based on a six-layer thalamocortical model.[5]". You could probably consider it, at any instant, to a'=f(W,a)-d(a) where W is the incoming signal at the current instant, a is the activation, and d is some decay, probably a function of a, with the neuron firing when a > A, although I don't know what mathematical formalism they use, and there are probably many more details. Quite how SpiNNaker implements that in detail (synchronous, asynchronous, etc).
Indeed. I mentioned this to my wife and suggested that a fully physically analagous implementation would need dozens of people to swap around an old-style telephone exchange patch board.
Implementing neurons directly in hardware is problematic, as it means you lose (1) flexibility to change the implementation, and (2) economies of scale from using commodity off-the-shelf components. You can offset that somewhat by using FPGAs, which gives you some economy of scale, along with flexibility. There have certainly been a number of neuron-like hardware systems in the past (e.g. CAM, although that's very simplistic), and other systems which are weight adders. I've worked on such systems. However, you can simulate some of it effectively, in terms of cost:compute on GPGPUs and other systems, even if it's overall more efficient to use weight adders directly. Steve Furber's taken a different route to using GPGUs, possibly because of the interconnect topology issues - i.e. if you have a Minsky machine (named after the founder of modern AI) you have a lot of GPGPU horsepower, but apart from NVlinks your distributed connectivity is going to be big box to big box. If you are building something at the scale of SpiNNaker, I doubt that would work, especially if you wanted something hyper-connected.
It's research funding. It covers the salary of a number of people, plus the costs of equipment (a dozen racks of computer system), office space, pencils, travel, etc. There are 38 staff in the Advanced Processor Technology group (headed by Steve Furber) at Manchester University.
Back in 'the day', by which I mean the days of my parents and grandparents, coffee with water and a little milk was called coffee, coffee with just milk was called 'church coffee'. Now, I suppose, it would be a latte, but without the steam and froth, as that would probably be sinful.
The theories make predictions, which by-and-large can ultimately be tested (at the expense of machines that smash things into other things with higher energies, perhaps). Even now, new data comes along, e.g. from CERN, and demolishes or supports various aspects. Ultimately, pen-and-paper is cheaper than particle accelerators, so theory development can run ahead of the available experimental information.
Are you assuming one neuron per core? You can have a neural network with many, many units running on a single CPU core. In that case the number of units in in the hundreds to thousands, typically, which means that neuron-to-neuron communication is relatively easy to handle as you can simply use shared memory. The trick with SpiNNaker and similar efforts is being able to marshall communication with more diverse connections and communication, and that gets complex when communications are not within the locality (physically speaking) of a neuron's computational location. An analogy (and one that Wolfram would approve of!) is to look at cellular automata (I did a fair bit of work on these in the past) where again it is possible to relatively easily do synchronous updates for CA that are locally connected, but much more complex for those that include a neighbourhood with less local connections for the update rule (cf. activation function). Sometimes using asynchronous updates can be useful, as long as they are reasonably timely, and you accept some jitter, but if the activation function is appropriate, it is possible for it to cope with the noise that asynchronicity injects. In my work that is where neural networks came in, as hand designing update rules becomes impossible, and you end up needing a neural network to learn the update complex function based on desired state transitions. But then you still need really good organisation of the distributed computing element to make if efficient in terms of computation/time and even more skill to create something that is efficient in terms of computation/energy. If you look at Steve Furber's credentials with ARM, you can see why he's a good lead for the latter, and he has assembled an excellent team working on all aspects of the problem.
Slashdot Mirror
Yes, I annoy other motorists by refusing to scare the living daylight out of cyclists when I am driving too.
The problem with cyclists is they never obey any rules--then they expect cars to do so and drive at 10mph as they ride down the middle of a lane because 'I'm a vehicle!'
I cycle and this is utter bollocks. I obey the rules, as do many of the people I know that cycle. I do not expect cars to drive at 10mph, but I do expect them to not knock me off. It grinds my gears (ahem) when cyclists don't obey the rules. I would certainly never run a red light except once about twenty years ago when I turned left as I realised I had my toe clips on too tight for a different pair of shoes and couldn't get my feet out the clips, when I turned left and onto the pavement. In the end I had to just stop and fall over then finally get my toes out of the clips at the end of my trip. If I know it's going to be a long light and wish to proceed I dismount, move my bicycle over the pavement to the other part of the junction a reasonable distance from the junction, remount, and move off when it is safe to do so.
I've been knocked off my bicycle twice by cars, and both times it was because drivers didn't see me or look for me, when turning left (this is in the UK) when I was continuing straight, and were initially behind me so I had no way to determine their intention to turn left in the space where I was. Luckily in both cases I was able to pick myself up, straighten my handlebars and carry on, as I was still not fully in the junction by the time they turned left and they just side swiped me. In one instance I did their car much more damage with the metal end grips on my handlebars than they did to me. But it has made me more aware of this possibility when driving.
There are some tearaways (not me!) and occasionally people ride two-abreast on narrow roads, but that's not common and by and large they are pretty good. And in these days of LEDs lights are much, er, lighter for a given level of light, which helps too this time of year.
There's also the Zoe from alliance partner Nissan. The Zoe is sold only in Europe, currently 2.5k/month. So it makes a direct comparison difficult. The Tesla is selling pretty well, though, if you look at market sizes. I couldn't see any obvious figures for total European electric car sales, but in France Zoes are about 60% of the sales, so that suggests total electric car sales in France of about 2k per year (not really very many), but scaling up by population that's about 10k in a Western Europan nation, so that suggests that Tesla is selling better than the Leaf and Renault. 50% of Zoe sales are in France.
I did consider a Zoe a while ago, but the previous version's range of 150 miles was a bit low. The current range of 250 miles would be tempting if I was in the market for a new car, and it had a bit more towing capacity. In a few years, if there's something for a reasonable cost with that sort of range and a decent towing capacity, that's probably what I will get.
Renault charges a fee per month (I presume given Nissan and Renault's close ties, Nissan is probably the same), based on distance travelled, then replace your batteries when required. You are effectively paying for the battery replacement over time, but it also means that the resale value of the vehicle remains good, as the new owner can continue that payment. It's somewhat in the interest of the vendor to make new batteries cheaper, as then the replacement cost is reduced, and they get to bank the difference, but it's also useful for them to be able to cost-effectively extend the lifetime of the battery too, as then the cost/time is reduced for the vendor, but also it can reduce the monthly charge, which would stimulate sales. When I considered buying a Renault when the range was around 130 miles (which in the end I felt would be too limiting) it was of the order of half the cost of petrol for a similar size of vehicle, although that ratio depends on the fuel price at any given time.
Yes, VW, Daimler, BMW, GM, Ford, fiat-chrysler, Nissan, etc have spent the last 5-7 years telling us how they would beat Tesla. Now, like cold fusion, it will happen real soon.
Several of those currently sell electric cars (and Renault) in the non-luxury market. They haven't been selling massively well in earlier incarnations, but the Nissan and Renault ones are now pretty reasonable (230 mile range), which is where they really become competitive with ICE cars. That's not a direct competitor to Tesla as it's a different market segment. BMW, Audi. etc., I would also see as a threat to Tesla outside the USA, but in the luxury market, as they are highly experienced manufacturers with a large and effective workforce. Tesla is probably pretty secure within the USA, though.
Add Kia, Toyota into the etc in the list. And I am sure there are others that are likely to be working on things, such as Vauxhall/Peugot/Citroen.
Just because you cycle to work it doesn't mean you can't shuttle kids at other times, surely?
3. Houses have become much more expensive
They are also much bigger. New houses today are twice as big as houses built 50 years ago, despite families getting smaller.
Adjusted for inflation, the average cost per square-foot has barely changed.
They are only much bigger in the USA, maybe also Canada. In the UK they are much more expensive, and if anything smaller.
Looking at the data, households on the top 20% have as few as zero and as many as 5 workers, or more. The average of 2.0 hides a lot, even if the modal value is 2. But $150,000 average with five working might not be impressive per person!
There are no bachelors in the top 20%?
Skynet is the name of the UK's military satellite network, and has been since the 1970s.
I see I'm behind the times on the full complexity and need to do more reading.
It's not important as to whether they are correct, rather than they make predictions which can be tested. The last I read is that recent evidence didn't seem to be supporting String Theory(ies) but rather the Standard Model, which means science is working as intended.
In fact, this is one way we use to filter out the fakers.
Opinions differ, obviously, http://thedatascientist.com/ma....
Traditionally, the activation function was f(W) where W is the set of incoming neuron activations over the synaptic connections, such that a=f(W) where a is an activation level, then the neuron fires if a>A, but now it's more common to use spiking neurons, i.e. a time-series element such that, say, a neuron that's had some pattern of events come in, then fires. That's what SpiNNaker uses, see https://en.wikipedia.org/wiki/... is an initial bit of reading, which includes "SpiNNaker (Spiking Neural Network Architecture), designed at the University of Manchester, uses ARM processors as the building blocks of a massively parallel computing platform based on a six-layer thalamocortical model.[5]". You could probably consider it, at any instant, to a'=f(W,a)-d(a) where W is the incoming signal at the current instant, a is the activation, and d is some decay, probably a function of a, with the neuron firing when a > A, although I don't know what mathematical formalism they use, and there are probably many more details. Quite how SpiNNaker implements that in detail (synchronous, asynchronous, etc).
I should go and read the papers.
Indeed. I mentioned this to my wife and suggested that a fully physically analagous implementation would need dozens of people to swap around an old-style telephone exchange patch board.
Thanks for confirming my suspicion. This is just a basic tunnel, nothing special or interesting, doesn't demonstrate anything new or innovative.
It depends on the cost to dig it. It might be innovative it was cheaper than would otherwise be expected, by a significant margin.
First of all, we do not understand how or why consciesness exists.
What makes you think SpiNNaker is necessarily designed to 'discover' consciousness?
Implementing neurons directly in hardware is problematic, as it means you lose (1) flexibility to change the implementation, and (2) economies of scale from using commodity off-the-shelf components. You can offset that somewhat by using FPGAs, which gives you some economy of scale, along with flexibility. There have certainly been a number of neuron-like hardware systems in the past (e.g. CAM, although that's very simplistic), and other systems which are weight adders. I've worked on such systems. However, you can simulate some of it effectively, in terms of cost:compute on GPGPUs and other systems, even if it's overall more efficient to use weight adders directly. Steve Furber's taken a different route to using GPGUs, possibly because of the interconnect topology issues - i.e. if you have a Minsky machine (named after the founder of modern AI) you have a lot of GPGPU horsepower, but apart from NVlinks your distributed connectivity is going to be big box to big box. If you are building something at the scale of SpiNNaker, I doubt that would work, especially if you wanted something hyper-connected.
It's research funding. It covers the salary of a number of people, plus the costs of equipment (a dozen racks of computer system), office space, pencils, travel, etc. There are 38 staff in the Advanced Processor Technology group (headed by Steve Furber) at Manchester University.
Back in 'the day', by which I mean the days of my parents and grandparents, coffee with water and a little milk was called coffee, coffee with just milk was called 'church coffee'. Now, I suppose, it would be a latte, but without the steam and froth, as that would probably be sinful.
To a 95% confidence level, this is true, but there's still a chance that 0% or 100% are made up.
The theories make predictions, which by-and-large can ultimately be tested (at the expense of machines that smash things into other things with higher energies, perhaps). Even now, new data comes along, e.g. from CERN, and demolishes or supports various aspects. Ultimately, pen-and-paper is cheaper than particle accelerators, so theory development can run ahead of the available experimental information.
Are you assuming one neuron per core? You can have a neural network with many, many units running on a single CPU core. In that case the number of units in in the hundreds to thousands, typically, which means that neuron-to-neuron communication is relatively easy to handle as you can simply use shared memory. The trick with SpiNNaker and similar efforts is being able to marshall communication with more diverse connections and communication, and that gets complex when communications are not within the locality (physically speaking) of a neuron's computational location. An analogy (and one that Wolfram would approve of!) is to look at cellular automata (I did a fair bit of work on these in the past) where again it is possible to relatively easily do synchronous updates for CA that are locally connected, but much more complex for those that include a neighbourhood with less local connections for the update rule (cf. activation function). Sometimes using asynchronous updates can be useful, as long as they are reasonably timely, and you accept some jitter, but if the activation function is appropriate, it is possible for it to cope with the noise that asynchronicity injects. In my work that is where neural networks came in, as hand designing update rules becomes impossible, and you end up needing a neural network to learn the update complex function based on desired state transitions. But then you still need really good organisation of the distributed computing element to make if efficient in terms of computation/time and even more skill to create something that is efficient in terms of computation/energy. If you look at Steve Furber's credentials with ARM, you can see why he's a good lead for the latter, and he has assembled an excellent team working on all aspects of the problem.