[FoRK] Air cars? Please?

[silky]

On Thu, Mar 27, 2008 at 6:29 PM, Stephen Williams <sdw at lig.net> wrote:
> silky wrote:
>  > On Thu, Mar 27, 2008 at 1:05 PM, Stephen D. Williams <sdw at lig.net> wrote:
>  >
>  >> We're actually ahead of a lot of his notions, especially on the
>  >>  computing / cell phone / software side.
>  >>
>  >>  A lot of the rest is doable except for either monetization issues or
>  >>  more commonly the likely disruption of current monetization. Those local
>  >>  minima are tough to bust out of. House construction, for instance, could
>  >>  be simplified and greatly reduced in price except that high end builders
>  >>  already have high end margins and would risk cannibalizing their market
>  >>  before they'd get ahead.
>  >>
>  >>  Domes aren't doable yet, but automated high speed traffic is totally
>  >>  within reach, except for those pesky manual drivers everywhere and poor
>  >>  (i.e. often difficult) roads taking up all of the public space. My car,
>  >>  which can be had starting at $25K, can already comfortably do 150mph,
>  >>  it's just not legal to go even half that anywhere in the US outside a
>  >>  race track. This is true even though many Interstates have zones that
>  >>  could handle at least 120mph without significant increase in accidents.
>  >>
>  >
>  > you've got to be kidding about the computer-driven traffic control
>  > system? no way would i drive in a world where a internet-connected
>  > computer system was controlling the scheduling.
>  >
>  It depends on how such a system would work.  Clearly, the Internet (or
>  internet) connected part would only provide high level information, just
>  slightly more detailed than we have now with Google maps in the Bay
>  Area, as an example.  This data would be suitable for people as well as
>  computers.
>
>  All actual driving would have to be locally controlled in a sealed
>  system, just like a person or, on the machine side, like an avionics
>  system.  If you are flown by large commercial airlines, you are already
>  under computer control for pretty much the whole time, with humans on
>  standby unless they want to get a little real stick time here and there
>  to stay sharp.
>
>  Barring something like turning all traffic lights green or changing
>  lights / routing to send opposing traffic at each other or disabling
>  railroad safety signs, things already potentially exposed, the only
>  Internet-level threat would be traffic jams, not actual crashes.
>
>  Most of the tech is a solved problem, at least for good conditions and
>  most roads.  If there were a big economic reason to do it, we could
>  probably have some form implemented in 5 years.  It's probably going to
>  take 20 until we get around to it.

it's not a technical probably at all, it's a problem of security and
responsibility and system correct-ness.

if the system can't provably show that it will never enduce a crash,
never allow any unauthorised modification - yet STILL allow for
emergency adjustments, like a rock falling on the road, or a rogue car
attacking other cars, etc, etc, then i wouldn't want to use it. and
even if it could show that, how can we be sure? is our understanding
of technology and processes for writing it so well thought-out and
tested and implemented by that time that we do have languages and
operating systems that come with proof's of function?

i doubt it.

anyway, i'd argue that if computer-controlled traffic is what you
want, using that on cars is stupid/ineffective/wasteful. better to use
a series of tubes.


>  sdw

-- 
http://lets.coozi.com.au/

There's not a problem I can't fix, because I can do it in the mix.