Right, there's a significant difference between water vapour and water aerosols, which is what this experiment is about. The second thing is that they are mainly interested in testing whether they can pump the stuff that high, instead of the climatological effects of that. (Which realistically ought to be negligible for a test this small.)
Well, they taught me that internet armchair generals recognise no such thing as innocence in their willingness to defend any crime committed by their chosen side.
Data storage should be in databases. Data manipulation should be done in software. Attempting analyses using giant spreadsheets of interrelating cells that cannot be easily debugged or even understood, using opaque 'functions' that frequently do things wrong, with massive file duplication over and over that prevent searching... That's an idiotic way to do things.
You might ask yourself, who are Slashdot's customers? What's its product? Well, surely its customers are those advertisers you see at the top of the page, and the product they are selling is the loving attention of Slashdot readers via comments like this one.
Well, the wikipedia for the Shinkai 6500 (whose record these guys are specifically trying to beat) explains: "The only manned expedition to have gone deeper was the dive of the Trieste bathyscaphe in 1960. However, the vessel could not navigate along the bottom of the sea bed." So that's the difference.
As per your link, the people responsible for that melamine thing were executed. The Chinese government tends to take these things very seriously, at least some of the time. Your example sucks.
It's not the left. It's just the axis of dumbasses. There's plenty of leftwingers (myself included) that opposed Iraq, want a withdrawal from Afghanistan, and still support this intervention.
Is that really that hard? Out in the country, it should be pretty easy to code a computer to spot signs and comprehend what it means. I doubt whatever they did would work in an urban environment though,
Rush hour traffic is actually pretty easy to program a robot car to drive in. The speeds are low, and because of the high density of the cars, you aren't going to change lanes or do anything complicated anyway. You could pretty much design a car on the principle of 'move somewhat slowly on a preset route, if anything gets in your way, stop' and it'll work moderately well. Or so I've been told by the local robotics research group.
Daylight illumination is fairly uniform throughout. Streetlights are very spotty, with the same thing looking very different depending on whether it's directly under a streetlight or distant away from it. Hence figuring out things at night with just a camera is a comparatively more difficult computational problem.
Why can't they use the model in the airline industry, where autopilots are commonplace? Require that there is always a driver to supervise the AI, and so whenever the car crashes, it would be the responsibility of the supervising driver.
Well, if you are actually interested in the science, the research this car is based on can be easily found, I think, using google. Here, read this article (exhibited at the 2008 IEEE computational intelligence conference hosted in Hong Kong), and if you comprehend it, you can implement their procedure yourself: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4634099&tag=1
Slashdot Mirror
Helllooooo strawman.
Right, there's a significant difference between water vapour and water aerosols, which is what this experiment is about. The second thing is that they are mainly interested in testing whether they can pump the stuff that high, instead of the climatological effects of that. (Which realistically ought to be negligible for a test this small.)
It's likely a massive underestimate of the civilian toll.
Well, they taught me that internet armchair generals recognise no such thing as innocence in their willingness to defend any crime committed by their chosen side.
The submission is quite humorous as a joke about the need for type safety though, though not really encouraging of trust for the devs.
Data storage should be in databases. Data manipulation should be done in software. Attempting analyses using giant spreadsheets of interrelating cells that cannot be easily debugged or even understood, using opaque 'functions' that frequently do things wrong, with massive file duplication over and over that prevent searching... That's an idiotic way to do things.
Super complex spreadsheets... really just shouldn't exist.
You might ask yourself, who are Slashdot's customers? What's its product? Well, surely its customers are those advertisers you see at the top of the page, and the product they are selling is the loving attention of Slashdot readers via comments like this one.
Well, the wikipedia for the Shinkai 6500 (whose record these guys are specifically trying to beat) explains: "The only manned expedition to have gone deeper was the dive of the Trieste bathyscaphe in 1960. However, the vessel could not navigate along the bottom of the sea bed." So that's the difference.
As per your link, the people responsible for that melamine thing were executed. The Chinese government tends to take these things very seriously, at least some of the time. Your example sucks.
It's not the left. It's just the axis of dumbasses. There's plenty of leftwingers (myself included) that opposed Iraq, want a withdrawal from Afghanistan, and still support this intervention.
I'm concerned it'll wipe my credit card.
But what are the military applications? Because let's be honest, if DARPA is doing it, then isn't that what this is really about?
Right, exactly. The study was doing by an internet marketting firm, so you have to realise what that means.
Is that really that hard? Out in the country, it should be pretty easy to code a computer to spot signs and comprehend what it means. I doubt whatever they did would work in an urban environment though,
Rush hour traffic is actually pretty easy to program a robot car to drive in. The speeds are low, and because of the high density of the cars, you aren't going to change lanes or do anything complicated anyway. You could pretty much design a car on the principle of 'move somewhat slowly on a preset route, if anything gets in your way, stop' and it'll work moderately well. Or so I've been told by the local robotics research group.
I don't know about your driving school, but I think you need a bit more judgement than that to overtake safely!
Daylight illumination is fairly uniform throughout. Streetlights are very spotty, with the same thing looking very different depending on whether it's directly under a streetlight or distant away from it. Hence figuring out things at night with just a camera is a comparatively more difficult computational problem.
In fact TFA had this information as well.
"The test also showed the car could cope with potential dangers from other vehicles such as abrupt lane changes.
"The driverless car is much safer because it reacts more quickly than humans. It can respond in 40 milliseconds while human needs at least 500 ms."
During its trip, the driverless car overtook other cars 67 times and had an average speed of 87 kilometers an hour, according to the research team. "
So, no, the road wasn't empty.
Why can't they use the model in the airline industry, where autopilots are commonplace? Require that there is always a driver to supervise the AI, and so whenever the car crashes, it would be the responsibility of the supervising driver.
Did you RTFA? Because that isn't true.
I don't think any of these do overtaking automatically though.
This feels kinda heavily anecdotal. It's not like things like that don't happen in other countries as well. http://www.cyclechat.net/topic/86407-us-driver-deliberately-hits-then-runs-over-cyclist-faces-no-charges/
Well, if you are actually interested in the science, the research this car is based on can be easily found, I think, using google. Here, read this article (exhibited at the 2008 IEEE computational intelligence conference hosted in Hong Kong), and if you comprehend it, you can implement their procedure yourself: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4634099&tag=1