I was thinking along these lines myself the other day. You've got someone trapped in rubble, and you want to send in an animal trained to find them. Now, the person is already stressed out. What are they going to do when a rat, or a mouse, or possibly even a snake comes along.
Besides, you still have the problem of getting a radio signal to penetrate steel, concrete, etc.
They use carrier pigeons for when phones, power and radio are down, for example during hurricans.
I think that they'd face a huge packet loss rate if they are using pigeons in a hurricane.
Actually, avian IP transmission schemes face some of the same limitations as token ring networks. You can only have a limited number of packets "on the fly" at any particular time. Each avian carrier is similar to a token. If it fails to reach its destination, the bandwidth of the entire network drops. Unless there is a pool (or "coop") of backup carriers, you would have to wait a considerable period for a new carrier to be created (or "hatched" as it is sometimes called by those running these types of networks).
Using this type of network in hurricane conditions virtually guarantees that you're going to lose all of your packet carriers. This will bring down your network completely, until the next brooding season.
Put lots of energy in to break everything apart and hope the bits come together in the right way with a means to carry off the excess energy (so the acids stay together)
And if they don't come all the way apart, how do you know?
Every square centimeter of every piece of lab equipment everywhere on the planet is covered in bacteria and virii. Merely killing the little critters is not enough for this type of experiment to be valid. Their bodies must be done away with. All amino acid and amino acid fragments must be removed. Not 99% removed. Not 99.9999% removed. Everything must be gone. Otherwise, all you're showing is that:
Raw material + energy + amino acids -> amino acids
instead of
Raw material + energy -> amino acids
Until all organic compounds are removed from the system (which we can't do), claims of creating spontaneous amino acids are invalid. In fact, the only thing that these experiments demonstrate is how difficult it is to wipe them out.
Good points! I've always had a problem with not having a model of the world. When the lights go out, I can still navigate in my home. I have some sort of model of the world in my mind. Yes, the world may not match the model, but the model does have a lot of predictive power. I will not wind up in the living room when that nighttime craving for one more piece of chocolate cake hits, because I know I need to turn left at the bottom of the stairs, not right.
Byte magazine had an article, many winters ago, about nested feedback loops and brain layers. It seemed promising to me at the time, but out of my league. I was a lowly student who knew nothing about simulating neural networks. I don't think I even had a computer at that point, and besides, I was more intrigued by the possibilities of expert systems... but I digress.
The article talked about using layers of feedback loops modelled on the brain's mechanisms. A single feedback loop can control the the tension of a muscle. A second feedback loop in control of the first one, can control how the muscle moves the joint to the desired point. A loop on top of these two controls how fast the joint flexes. A loop on top of those controls the position of one joint relative to another. Unfortunately, I do not have access to the article any longer. It would be good to read it again.
The world may be its own best model, but that means squat if you want to develop complex bots. For example: Face a wall, turn off the lights, and walk to the wall. Chances are you won't run into the wall. Why? You have an internal model of the world that tells you how far you can walk before you plant your nose into the drywall.
The subsumtion architecture model works fine for reflex level stuff. But in order for it to become useful, it must take a supporting role to some sort of hierarchical planning/goal seeking mechanism.
A washing machine is not a robot. It does not sense its environment and adjust itself to it. It cannot modify its own behavior. It does not make any decisions. A washing machine blindly obeys the timing cycles built into it. Furthermore, those timing cycles are typically hard wired. You want a longer spin cycle, you have to replace a physical part.
A robot must be able to sense and respond to its environment. A robot needs to be able to decide between different possible behaviours, and a robot's behaviour set must not be hard wired.
We have figured out what causes evolution, and we do control it.
"Evolution" is a combination of mutation, and inheritance. Mutation causes a change in the genentic makeup of an organism, and inheritance passes this change to that organism's offspring. If that change does not hinder the survival of the organisms, it is passed on. If it does, then the change does not propogate through the population.
We control evolution in two ways: Selective breeding, and lately, through gene splicing. Mankind has been controlling the evolution of many organisms: Dogs, cats, goldfish, horses, corn, tomatos, wheat, etc. The list is extensive, and dates back as far as we can track our own history.
Recently, we have taken a more direct approach. Instead of controlling the inheritance end, we control the mutation end. We take a DNA sequence we like, graft it into the target, and voila! glow in the dark bananas, or disease resistant potatos.
It is unlikely that they would bring it down. They would either replace the defective part, or slap a strip of duct tape, refuel and then relaunch it in orbit.
$600/month will get you a small, bug infested low end basement apartment. Groceries would cost another $100/month at least. Certain luxury items like heat, water, electricity would run you $25/month. Transit back and forth to work runs you around $20/Week.
I don't know where you live, but you won't get by on $600/month in most major cities.
No, that is the average intensity. Yes, there is a thermal delay. Yes, it takes time for the filament to heat up, and time for it to cool off. But the filament does heat and cool in time with the voltage fluctuations. Granted, it may not cool off enough to stop glowing, but the light given off does vary. A light bulb filament just doesn't have the thermal "mass" needed to moderate/buffer 60Hz AC into a flat intensity.
I'm not sure which side of the issue you're trying to argue, but it sounds like you're saying that incandescents are not modulated by AC.
Let's assume that when the voltage is at it's highest, the bulb is at it's hottest. From that point of the sine wave, to the zero crossing at the 180 degree mark, the filament is cooling off. Now, as you say, it will still be hot when we reach the zero crossing, but the voltage on the wire cannot maintain that temperature. In fact the filament will continue to cool until we reach a voltage that will maintain the filament temperature. From that point on, the filament will heat up until we reach the voltage maximum (minimum) at the 270 degree mark. At that point we would again cool off until we pass the 360 mark.
In other words, there are two heat/cool cycles per power cycle. Connect a solar cell to a speaker, and you will hear a 120Hz hum.
Twice a year is right. Take an ellipse and select a point on the ellipse. Pretend that this is the Earth. Now select another point outside the ellipse (voyager), and draw a line from the voyager point, through the Earth point and continue it. This line is the signal path. You will notice that unless you chose an Earth point on the ellipse that made the signal path line tangential to the ellipse, your signal path line will intersect the ellipse twice.
(Assuming the craft doesn't wobble, that it is in the plane of the earth's orbit, that it's antenna is pointed to somewhere on the earth's orbit, and that the signal path doesn't pass through any intervening stellar phenomenon (sun block:-), then this argument holds)
He is the civilian equivalent to General Protection Fault (and coincidentally, his nephew). Rumour has it that he was in the same class at the academy as Kernel Panic, but for some reason he was discharged. It's all very hush-hush.
Slashdot Mirror
Not to mention buoyancy. Air is a fluid, so all objects are "lighter" by the weight of the air they displace.
Note: This is different than lift, which is the result of pressure differences caused by air moving at different speeds.
I was thinking along these lines myself the other day. You've got someone trapped in rubble, and you want to send in an animal trained to find them. Now, the person is already stressed out. What are they going to do when a rat, or a mouse, or possibly even a snake comes along.
Besides, you still have the problem of getting a radio signal to penetrate steel, concrete, etc.
Linux was also used to render scenes from "Titanic" as I recall.
This is where the name "Justin Case" would come in handy. Maybe "Anita Name"...
Probably one of those spam mail order doctorates.
Has this been fixed yet? IE restoring the coyright notice?
OK, you share your credit card number and expiry date with me, and I'll tell you my shoe size.
Some information is owned.
Read your bible. God has made several personal appearances. In fact, what do you think that whole Jesus incident was about?
Submarine races, eh? Is that what you told your folks?
I think that they'd face a huge packet loss rate if they are using pigeons in a hurricane.
Actually, avian IP transmission schemes face some of the same limitations as token ring networks. You can only have a limited number of packets "on the fly" at any particular time. Each avian carrier is similar to a token. If it fails to reach its destination, the bandwidth of the entire network drops. Unless there is a pool (or "coop") of backup carriers, you would have to wait a considerable period for a new carrier to be created (or "hatched" as it is sometimes called by those running these types of networks).
Using this type of network in hurricane conditions virtually guarantees that you're going to lose all of your packet carriers. This will bring down your network completely, until the next brooding season.
Put lots of energy in to break everything apart and hope the bits come together in the right way with a means to carry off the excess energy (so the acids stay together)
And if they don't come all the way apart, how do you know?
Every square centimeter of every piece of lab equipment everywhere on the planet is covered in bacteria and virii. Merely killing the little critters is not enough for this type of experiment to be valid. Their bodies must be done away with. All amino acid and amino acid fragments must be removed. Not 99% removed. Not 99.9999% removed. Everything must be gone. Otherwise, all you're showing is that:
Raw material + energy + amino acids -> amino acids
instead of
Raw material + energy -> amino acids
Until all organic compounds are removed from the system (which we can't do), claims of creating spontaneous amino acids are invalid. In fact, the only thing that these experiments demonstrate is how difficult it is to wipe them out.
Good points! I've always had a problem with not having a model of the world. When the lights go out, I can still navigate in my home. I have some sort of model of the world in my mind. Yes, the world may not match the model, but the model does have a lot of predictive power. I will not wind up in the living room when that nighttime craving for one more piece of chocolate cake hits, because I know I need to turn left at the bottom of the stairs, not right.
Byte magazine had an article, many winters ago, about nested feedback loops and brain layers. It seemed promising to me at the time, but out of my league. I was a lowly student who knew nothing about simulating neural networks. I don't think I even had a computer at that point, and besides, I was more intrigued by the possibilities of expert systems... but I digress.
The article talked about using layers of feedback loops modelled on the brain's mechanisms. A single feedback loop can control the the tension of a muscle. A second feedback loop in control of the first one, can control how the muscle moves the joint to the desired point. A loop on top of these two controls how fast the joint flexes. A loop on top of those controls the position of one joint relative to another. Unfortunately, I do not have access to the article any longer. It would be good to read it again.
The world may be its own best model, but that means squat if you want to develop complex bots. For example: Face a wall, turn off the lights, and walk to the wall. Chances are you won't run into the wall. Why? You have an internal model of the world that tells you how far you can walk before you plant your nose into the drywall.
The subsumtion architecture model works fine for reflex level stuff. But in order for it to become useful, it must take a supporting role to some sort of hierarchical planning/goal seeking mechanism.
A washing machine is not a robot. It does not sense its environment and adjust itself to it. It cannot modify its own behavior. It does not make any decisions. A washing machine blindly obeys the timing cycles built into it. Furthermore, those timing cycles are typically hard wired. You want a longer spin cycle, you have to replace a physical part.
A robot must be able to sense and respond to its environment. A robot needs to be able to decide between different possible behaviours, and a robot's behaviour set must not be hard wired.
A washing machine fails on all three counts.
We have figured out what causes evolution, and we do control it.
"Evolution" is a combination of mutation, and inheritance. Mutation causes a change in the genentic makeup of an organism, and inheritance passes this change to that organism's offspring. If that change does not hinder the survival of the organisms, it is passed on. If it does, then the change does not propogate through the population.
We control evolution in two ways: Selective breeding, and lately, through gene splicing.
Mankind has been controlling the evolution of many organisms: Dogs, cats, goldfish, horses, corn, tomatos, wheat, etc. The list is extensive, and dates back as far as we can track our own history.
Recently, we have taken a more direct approach. Instead of controlling the inheritance end, we control the mutation end. We take a DNA sequence we like, graft it into the target, and voila! glow in the dark bananas, or disease resistant potatos.
It is unlikely that they would bring it down. They would either replace the defective part, or slap a strip of duct tape, refuel and then relaunch it in orbit.
$600/month will get you a small, bug infested low end basement apartment. Groceries would cost another $100/month at least. Certain luxury items like heat, water, electricity would run you $25/month. Transit back and forth to work runs you around $20/Week.
I don't know where you live, but you won't get by on $600/month in most major cities.
PEBCAK
Problem Exists Between Chair And Keyboard
No, that is the average intensity. Yes, there is a thermal delay. Yes, it takes time for the filament to heat up, and time for it to cool off. But the filament does heat and cool in time with the voltage fluctuations. Granted, it may not cool off enough to stop glowing, but the light given off does vary. A light bulb filament just doesn't have the thermal "mass" needed to moderate/buffer 60Hz AC into a flat intensity.
Let's assume that when the voltage is at it's highest, the bulb is at it's hottest. From that point of the sine wave, to the zero crossing at the 180 degree mark, the filament is cooling off. Now, as you say, it will still be hot when we reach the zero crossing, but the voltage on the wire cannot maintain that temperature. In fact the filament will continue to cool until we reach a voltage that will maintain the filament temperature. From that point on, the filament will heat up until we reach the voltage maximum (minimum) at the 270 degree mark. At that point we would again cool off until we pass the 360 mark.
In other words, there are two heat/cool cycles per power cycle. Connect a solar cell to a speaker, and you will hear a 120Hz hum.
What do you think is pushing all that data down the fibreoptic cable, a candle and Cowboy Neil?
(Assuming the craft doesn't wobble, that it is in the plane of the earth's orbit, that it's antenna is pointed to somewhere on the earth's orbit, and that the signal path doesn't pass through any intervening stellar phenomenon (sun block :-), then this argument holds)
He is the civilian equivalent to General Protection Fault (and coincidentally, his nephew). Rumour has it that he was in the same class at the academy as Kernel Panic, but for some reason he was discharged. It's all very hush-hush.
For a robot to build a robot with a lack of first law constraints would be a violation of the first law.
Dangit! Where did I leave that deed to that dilithium mine...