h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Hmm.

3. The most important military fact of the 21st century was demonstrated very early to be that there's no way to repel an attack by a tiny force of sufficiently motivated Muslims.

But to put it in context, there have not been casualties on the scale of good old fashioned war. We are rightly appalled by acts like 7/7, but to our parent's and grandparent's that scale of civilian casualties in their city would count as a quiet night. Think of how terrifying their lives must have been. Not even the monstrosity of 9/11 comes close to a single night's toll of serious warfare.



4. Yeah - Americans aren't on fish and yeast - although many are subsisting on Poverty Food such as burgers and ramen.

Elsewhere...well...global standards of nutrition aren't brilliant. Conceivably *averaged out* it amounts to fish and yeast...which means that next time you tuck in to [name tasty dinner of your choice] you are doing extremely well compared to someone who isn't even at fish'n'yeast level. (I'd argue that you might be dining well at their expense - but that's another issue). So Heinlein simply hadn't factored in increasing global inequalities.

Incidentally...there's not plenty more fish in the sea anymore. Even krill are being over-farmed. However, domestic fish farming may be a solution. It is said that there is an active competition between two of the leading nations in GM technology to produce the first agriculturally useful GM animal. If the US gets there first it will be a disease resistant salmon. If Cuba does it, GM tilapia - one of the world's most commonly eaten fish. Chuck some in an irrigation dish and throw in the odd bit of garbage and you've got food.



15. There's an interesting debate over whether China is Communist - and I've had the conversation lately with various people from PRC. They say not 'Everybody's just out for money these days'. But from what I can see there are still distinct elements of 'Socialism With Chinese Characteristics' in the words that come out from Beijing. It will be interesting to see what nudges on the tiller Li Keqiang makes - he is thought to retain some M-L leanings.

A little subtlety - what we're actually seeing is managed capitalism. Marxism is an analysis of capitalism which tells us how it works. (And Maoism taught the Chinese how to overcome their old power structures). I don't think even the Chinese imagine they've attained Communism. What they're after is the necessary Capitalist conditions that precede Communism. Whether the long term end justifies the way they treat their people in the short term is a separate issue.



17. Why assume that aircraft would be piloted by laypersons? We're not quite there yet with autonomously operating aircraft - but it's not an impossibility. We're pretty close with cars.

Whether it's a good idea...? Think laterally. Better public transport. Bikes.

"Why assume that aircraft would be piloted by laypersons? "

You misunderstand.

I'm saying that a layman, in 1949, might well think "oh, planes will be controlled automatically". This is because, as a layman, he massively underestimates the complexity of the task of piloting, and thinks it susceptible to automation.

"We're not quite there yet with autonomously operating aircraft "

Again with the layperson's massive underestimate of complexity. We're not "not quite there". We're a long, LONG way away.

"We're pretty close with cars."

And that's the giveaway, right there. We're not, in fact, pretty close with cars. Are you suggesting that an autonomous road vehicle is "pretty close"?

I can't buy one. But that's not "pretty close", that's "here". So, here's what I'd say counts as "pretty close": an experimental vehicle is demonstrated into which I can get, specify my destination verbally or one a keyboard or map or similar as with a satnav, and then I can sit and read a book while the car drives itself there. It will need to be capable of getting off my drive, down the road, negotiating junctions, roundabouts, A roads, motorways, level crossings, pedestrian crossings, traffic lights, and the barrier at the entrance to my workplace. It will need to be able to do all these things without my intervention. When a single vehicle has been built which can achieve this, then and only then will I give you "pretty close" for cars. Right now, we're nowhere near being able to do that, much less have such a system certified as safe for sale to anyone.

And that's cars - vehicles travelling at 70mph or less () on a two-dimensional, stationary, visible medium.

Most aircraft would be stalling if they tried to fly as slow as 70mph, and their medium is not stationary, not visible, and three dimensional. Flying is a great deal more complex and difficult than driving. One hint should be that the average car driver has spent ten or twenty hours learning, none of it solo, before passing the simple test required, whereas the average private pilot has spent five or six times longer and passed numerous written examinations... and that's to be let loose with the very lightest and simplest powered aircraft.

I shall be surprised and impressed if there's a commercially available automatic car by the end of the century. I predict with some certainty that there won't be a similarly capable aircraft in that timescale.

What's distracting to this discussion is the existence of semi-autonomous military drone aircraft. Much is made of their ability to take off, navigate to a target, deliver ordnance and return to base and land with minimal operator intervention. However, there are significant operational limits on drones, and furthermore they don't generally operate in the sort of airlaw environment that airliners have to. There's a reason that the British RAF drone wing is based at Creech in Nevada - they can't legally fly in the UK. Also, there's the issue of consequences - even if an autonomous drone crashes, there's relatively little risk it'll hit anything important, and of course there's nobody on board - and that's most of the point. Its control systems would need to be many orders of magnitude more reliable before you could ask an actual human to sit in there and let the thing fly them about.

Flying is *HARD*. It's much, much harder and more dangerous than driving, and we've not automated driving yet. The problem is, laypersons hear of things like "autopilot" and "automatic landing systems" and autonomous drones, and think the job of the modern pilot is to be some kind of caretaker to the computer that controls the plane.

>>Again with the layperson's massive underestimate of complexity. We're not "not quite there". We're a long, LONG way away.

>>"We're pretty close with cars."

>>And that's the giveaway, right there. We're not, in fact, pretty close with cars. Are you suggesting that an autonomous road vehicle is "pretty close"?

Perhaps we can quibble over what is meant by 'pretty close' - but as it happens I'm not entirely a layperson in this field.

Well that's good. Perhaps, then, you'd care to speculate on how long it will be, in decades, before:

(a) an experimental vehicle is demonstrated that could take me, on public roads, from my home to my place of work 20 miles away via two motorways, without my having to do anything more than specify the destination, and

(b) before such technology becomes available in a car I can buy.

As a complete layman in this field, my guess would be it'll be no less than a decade before we get (a), and no fewer than three before we get (b). Make me happy - tell me I'm being pessimistic. Tell me there's a realistic possibility of my being able to buy a self-driving car before I get my pension...



All of which is interesting but not relevant - we were talking about aircraft, the control of which are at least an order of magnitude more complex than cars, and the safety standards for which (at least in civil aviation) are significantly higher still.

(Note: I've seen videos of demos of self-driving vehicles on public roads. I've been seeing them for years, though. There's always some caveat, not least there's invariably a driver on board to take over if/when the computer falls over. Which defeats the point entirely. I want to *sleep* while my car drives me to work - or watch a movie, or read a book, or have sex. If I have to be watching to make sure the computer is driving acceptably... *I* am driving. Hands off, but I'm ultimately in charge. Which is worse than useless.)

It won't happen for various reasons - principally because it isn't needed.

*Technologically* it could happen soonish - but with not quite the same technology. Think laterally. It wouldn't be about cars or aircraft sensing other aircraft/cars and making autonomous decisions to avoid bashing into them. It would involve ad-hoc networking - all the vehicles cooperating in control arrangement. Currently there's little need for the necessary infrastructure (which would have to be ubiquitous) since what we've got already works. Ish. Although some have been considering it. I could name names.

We'll more likely see something of the sort in shipping as more and more 'short shipping' takes place in congested, busy sea lanes to meet the geographical trend of industry and people being located near coasts Not quite such a fast moving environment, sure - but with its own challenges.

So if you've finished the You Know Nothing slapdown...

Jaysus. Did you really think I was talking about military drones? Or that I wouldn't be aware of how military drones work? Now go and make some Teflon or chocolate or whatever. Or talk about proton torpedoes.

>>Note: I've seen videos of demos of self-driving vehicles on public roads. I've been seeing them for years, though. There's always some caveat, not least there's invariably a driver on board to take over if/when the computer falls over.

And early cars had a wee man out the front with a flag.

Jaysus. You've seen the videos, eh? I've met the guys who make 'em.

Now kiss.

So let me see if I've got this right...

You've met the guys who make them - whoopie do for you. We *could* do them "soonish". Except... we won't because? It's "not needed"???

Angry Birds is "not needed". I would suggest that a reliable technology for getting fallible humans out of the loop when it comes to controlling 1000kg lumps of metal moving about in close proximity to pedestrians and bicycles is needed a hell of a lot more urgently than we need 3D TV or a foldable mobile phone display.

Vehicle swarming is interesting (that's what you're on about with the "ad hoc networking", right?), except you need a road that's guaranteed to have no pedestrians or motorcyclists or cyclists or mobility scooters etc. etc. etc. And you need, overnight, for every single vehicle on the road to have it. Which, as I believe I said, is worse than useless. I don't just want to sleep on the motorways.

Your answer to my two questions - and please correct me if I'm wrong, is "Never", on both counts. And that's for cars, not aircraft. Right?

"what we've got already works"

Tell that to the bereaved relatives of the ten people who will be killed today because a driver somewhere wasn't paying attention. And tomorrow, and the next day.

Your intention from the outset has been to demonstrate to everyone that I am stupid and know nothing.

Why?

oh...and do cars need to dodge pedestrians and motability scooters on motorways?

No. They're allowed to run them over on motorways

Actually I did once have to phone the polis because I passed a motability scooter heading down the slip road onto the M77.

When my wife used hers on the roads near Crieff, I'm convinced drivers did think they were allowed to run her over She hasn't used it on the road since.

@31 - you misunderstand (again).

My intention from the outset was to demonstrate that Heinlein's prediction of automatic aircraft was in error, and likely based on his (for a layman, typical) lack of knowledge of the complexity of the task of automating it.

Your level of stupidity or knowledge is entirely irrelevant to this point.

From the outset, you've tried to make this about you, rather than Heinlein. Why?

Because that's what h2g2 is about. You and Ed, rolling around the proverbial floor, sweatily wrestling over ideas that the rest of us are too afraid to grasp.

Well I wouldn't have minded a discussion. But this one degenerated early with 'Again with the layperson's massive underestimate of complexity'.

I came across the following and thought it could be posted here: http://www.youtube.com/watch?v=A1vyB-O5i6E

There's no youtube video of "the" levitating mouse so this will have to do: http://www.dailymail.co.uk/sciencetech/article-1212553/NASA-levitates-mice-using-magnets-simulate-space-travel.html

Google wants driverless cars-

http://www.dailymail.co.uk/sciencetech/article-1386378/Google-pushes-changes-law-make-driver-cars-legal.html

'Self-driving cars could soon be a reality on the roads of Nevada if Google gets its way.'