[quote=dnooman
Who wants to see/needs to see/ or wants to display the back of their monitor? Why would you want a display that had a constantly changing background?
[/quote]
Because you want to project an image in front of the what’s behind it. For example, the Heads Up Display of an aircraft.
I thought you were going to say the whole “pre crime” tech.
LOL..
Off topic, but last year when my office got robbed, I had the opportunity to sit with the cops (and half my team) going through the security camera footage on my laptop. Of course, with every key stroke I’m like: typetypetypetypetype
“Enhance…” typetypetypetypetype
“Enhance…” typetypetypetypetype
“Enhance…”
A true story about ultralights and the issue of “pilot error”.
A guy I used to work with bought a powered paraglider. It’s essentially a motorized parachute. As far as I can tell, they’ll sell them to anyone and anyone can fly one if they feel up to it.
This guy hops in and starts flying. Within a week, he’s “crashed” it. (It wasn’t a real crash because like I said, it’s essentially a parachute. So he basically had an unexpected landing and was uninjured.)
And guess what the problem was? He had run out of gas while he was flying. Because he forgot to refill the tank.
Now if you want to ride around on your motorcycle or your ATV or your speedboat, it’s up to you if you check the gas tank. But if you’re going to go flying, you need to check the gas tank everytime before you take off. Two hundred feet up in the air is not a good place to run out of gas.
Like I said, this guy survived because of the unusually forgiving nature of the vehicle he was flying. But it illustrates the point that flying is a much more serious than driving and a cavalier attitude towards safety that’s no big deal in a driver is a fatal problem in a pilot.
They’ll be used very much like the cars of the time. You don’t think cars will still be piloted by humans in the future, do you? A “car” will be a personal ground vehicle that takes you where you tell it to. A “flying car” will be the same thing, that travels through the air. I’m sure we won’t call them “flying cars” unless there’s a strong retro-future fad involved. They’ll be called… who knows. No reasonable way to predict that. Calling such a thing a “flying car” is like calling a railroad an “iron horse”. It’s a transitional name for the concept, not a prediction of what they’ll actually be. But let’s not quibble about semantics. There will be personal air vehicles used casually for transportation, much the way that we casually use cars now.
Human pilots aren’t safe. No argument here. Hell, human drivers aren’t safe, either. That’s why future vehicles (of all kinds) will not be piloted by humans.
I think there’s at least two generations of technology we need, before we can make a CGI Bogart. First, we need to be able to make convincing humans, and we’re getting pretty close on that. Next, though, we’d have to be able to make convincing specific humans. For comparison, it’s easy to make me look like a real human (I manage to pull it off every day), but it’s very hard to make me look like Bogie.
I think the failure to get a decent young Jeff Bridges in Tron: Legacy, but only missing it by >< that much, is a salient sign. Where ILM failed was in the eyes, and translating the performance, but everything else was really good. Whereas where WETA succeeds is in the eyes and performance capture, but falls down on skin and getting the lighting right (it tends to somehow look staged, for want of a better word).
Your friend needed more training. Powered paragliders land the same whether the engine is running or not. The rule of ppg is never to fly above a place where you can’t safely land. I had the engine quit on me a bunch of times when I used to fly one- you just turn it into the wind and flare at the proper time. Truly they are a joy to fly. Mine would pack up and fit in the back of my Civic.
I like your line of reasoning, but I still think the problem will have to be solved in a way that we can’t comprehend yet.
Auto pilot features have manual overrides because there may be mechanical failures, these are intended for professional pilots though. Something similar would need to be built in to a “sky transport” but would have no skilled person to operate it necessarily.
Currently, ATC handles many, many planes coming and going out of airports every hour. They carry hundreds of people on a set schedule. Imagine every one of those people had their own transport, and could arrive whenever they felt like it?
Besides, isn’t modern auto tech much more able to deal with inertia issues as opposed to gravity issues? Maybe auoto-pop parachutes like in small planes would need to be mandatory.
For it to work they’d need to be able to create realistic humans who are seen on screen with real humans and the audience couldn’t tell which was which. Gollum and giant apes are one thing, Bogey is another.
Plus, real human actors use their faces and bodies to create very subtle feelings, so a convincing CGI actor would not only have to look realistic, it’d have to ‘learn’ to act. A convincing Adam Sandler CGI actor, that’d be a cinch. Maybe that’s where they should start and work up to Brando.
Hey, watching not-Arnold square off against Christian Bale was the high point of the latest TERMINATOR movie (for me, and, near as I could tell, for the rest of the audience).
Yes, I understood that. I meant “I don’t know enough about the economics of airships to make a knowledgeable argument, in contrast to other modes of travel, where I do know enough about the economics”.
True. And I know that we may have serious problems with our helium supply in the future. But right now, does having to use He over H a major economic impact on the viability of airships? What fraction of the cost of operating an airship is the cost of the gas to fill it? The relative costs of the gases are important, but not the only consideration.
I’m fully aware that hysterical regulations can have bad economic outcomes. I’m just not convinced that hysteria over the Hindenburg is really what’s keeping blimps from major success.
Yeah, well, plus it was a joke, taking that one sentence out of context. I mean, if we need to make a convincing human, we have the means, though the rendering process takes nine months.
There have been several proposed technologies for “faster than a ship, cheaper than a plane” cargo service. Airships are one, some others are hydrofoils, air cushion vehicles, “ram wing” ground-effect craft, maybe others I’m not remembering. This stuff has been on the pages of Popular Mechanics since the 1960s. But either they aren’t cheap/practical enough or the market simply isn’t there.
Not really. They have manual overrides because there’s no demand for a self-flying aircraft. Creating a self-flying aircraft is a straight-forward engineering challenge and well within the bounds of today’s computing power and sensor technology.
It’s essentially already been done: the recent Mars landers used fairly tricky maneuvers to safely land relatively delicate hardware, and they had to do this autonomously due to the communications delay with Earth. Doing the usual flight, takeoff and landing of an aircraft is trivial by comparison. The F/A-18 fighter is entirely capable of landing itself on a carrier (a pitching, rolling, moving target) in any weather; the system goes unused except in emergencies mainly because to the egos of the people inside them
A modern, fully designed system would not only be capable of handling anything short of a catastrophic mechanical failure, it would do so better than any human pilot. Considering that most aircraft losses are due to erroneous pilot response to a malfunction rather than the malfunction itself.
The problem comes in convincing people of this. “Let our AI fly you!” is not exactly a winning ad campaign for the forseeable future.
If the aircraft are automated, then the ATC is too, and can suddenly be done efficiently at any scale. From a computer science perspective, this is actually a much more difficult problem than that of flying the aircraft.
Funny true story: back in the early 1960s the Air Force was proposing developing the X-20/DynaSoar spaceplane as an operational manned suborbital bomber. But the Air Force insisted that the pilots be able to actually “fly” it to it’s target, which was a meaningless concept because the boost stage and the reentry glide trajectory would have to be 100% controlled by computer and inertial navigation. So the engineers came up with a system whereby the joystick could be manually set to different positions which would correspond to simply telling the flight computer what path to take.
I don’t think that’s quite right. Autopilots have manual overrides because autopilots are still worse than human pilots in extreme cases that the autopilot wasn’t designed to handle. Or, at least, because people think that they’re worse than human pilots in those situations.
But that’s certainly not a fundamental property of autopilots. It really won’t be very long before autopilots are just strictly better than human pilots in all situations. Get enough data and simulate enough strange conditions and failures, and there’s no longer a need for a manual override; it simply won’t help to put a human in the loop.
That’s not to say that an autopiloted vehicle will never crash. There are still mechanical failures or bizarre weather events that simply render the craft unflyable. But it’s not like a human pilot would fare any better.
