Searching \ for '[OT:] Fatal software disaster - lessons to learn' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=fatal+software+disaster
Search entire site for: 'Fatal software disaster - lessons to learn'.

Exact match. Not showing close matches.
PICList Thread
'[OT:] Fatal software disaster - lessons to learn'
2004\08\08@094727 by Jake Anderson

flavicon
face
tag fixed

{Quote hidden}

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-requestKILLspamspammitvma.mit.edu

2004\08\08@170625 by Jinx

face picon face
I had this wee debate with someone on another group at the
weekend about engineering failures, in particular power plants
(the thread was originally about the electricity grid in NZ and
nuclear waste)

He wrote :

> Yes, human error can be reduced, but it can't be eliminated
> altogether. The machines have to be run by humans

My argument was that if the spirit is willing, any engineering
undertaking can be made reliable and safe. Maybe that's just
me looking through my idealists' rose-tinted glasses.

Redundancy, backup, failsafes, use of appropriate technology,
reducing human involvement and so on. My claim was that any
failure could be traced back to corruption, cost-cutting or care-
lessness/incompetence, all of which are fixable

For example, nuclear waste is not dangerous if managed
properly, which it can be. And a mains socket isn't dangerous
if you don't stick your finger in it

--
http://www.piclist.com hint: To leave the PICList
KILLspampiclist-unsubscribe-requestKILLspamspammitvma.mit.edu

2004\08\08@194901 by Russell McMahon

face
flavicon
face
> For example, nuclear waste is not dangerous if managed
> properly,

By definition.

> which it can be.

Now that's less certain :-(.
Why should high radiation level nuclear waste be "safe" for millenia, given
our successes in lesser disciplines.

The lifetime of the worst classes of nuclear waste so far exceed man's
proven ability to create viable engineering solutions that we have no way of
knowing if mooted waste management systems will work. But mankinds ability
and indeed eager willingness to mess things up, get it wrong, cut corners,
cheat where cheating is insane or fatal etc is so well demonstrated that
there is every reason to think that nuclear waste storage is beyond us, at
least currently. "Safely stored" wastes in solidified material were found to
be migrating material surfacewards when living organisms living inside the
solidified waste were found using the material as a food source. Just
because THEY can live in such unbelievable conditions, it doesn't mean
anything else can. Leakproof containers ultimately aren't. Impermeable rock
caverns may well be for years decades or centuries - possibly even millenia.
But SOME won't prove to be next week/month/year/decade - let alone century.
Which type have you chosen to live near?

Any active human engineeering solution that lasts more than a century is a
miracle of design, excecution and luck. Passive engineering designs can last
somewhat longer. (Examples of each may include an 1904 automobile or an
Eiffel tower). But in each case maintenance, repair, refurbishment and
monitoring is essential. Left to themselves all 1904 autos and eiffel towers
would be long gone. You can find pieces of roman seweres and walls and
buildings, but functionally they are long gone.

I think we have a way to go yet before we can claim to be able to safely
manage nuclear waste. Arguably the safest solution is to concentrate it very
highly and then dump it in deep space. Would be rather expensive - but once
it achieves escape velocity its pretty safe. The odd one that 'doesn't
quite' excape is then just part of the cost of doing busienss :-(


       RM

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu

2004\08\08@201840 by Jinx

face picon face
> I think we have a way to go yet before we can claim to be able to
> safely manage nuclear waste. Arguably the safest solution is to
> concentrate it very highly and then dump it in deep space

The argument against transporting nuclear waste by any means
is "what if there's an accident or terrorism?". Back to humans again.
Pragmatists and realists could argue this forever

I'd have thought a very deep shaft would suffice. Way down in
bedrock, and by the time the waste re-appeared by tectonic
subduction, its half-life would have been far exceeded, resulting
in a little volcanic puff of lead

> Would be rather expensive - but once it achieves escape velocity
> its pretty safe. The odd one that 'doesn't quite' excape is then just
> part of the cost of doing business :-(

Well, you make your societal choices as to what standard of living
you want (or indeed have thrust upon you). Woud we be safer as
greenies living like Amish ? Probably, but safety comes a distant
second behind curiosity and the need to explore and invent

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu

2004\08\08@211528 by Howard Winter

face
flavicon
picon face
Russell,

On Mon, 9 Aug 2004 11:38:46 +1200, Russell McMahon
wrote:

>...<
> I think we have a way to go yet before we can claim to
be able to safely manage nuclear waste.  Arguably the
safest solution is to concentrate it very highly and
then dump it in deep space.

But you're potentially giving someone else the problem -
I think going the other way would be safer - into the
Sun.  The little blip of radioactivity we're talking
about wouldn't even be noticed against the vast energy
there.

> Would be rather expensive - but once it achieves
escape velocity its pretty safe. The odd one that
'doesn't quite' excape is then just part of the cost of
doing busienss :-(

That is the Big Problem, of course - fine when it works,
potential disaster when it doesn't!  And failed
rocket-launches aren't terribly rare, unfortunately.

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-requestEraseMEspamspam_OUTmitvma.mit.edu

2004\08\08@222323 by Russell McMahon

face
flavicon
face
> The argument against transporting nuclear waste by any means
> is "what if there's an accident or terrorism?". Back to humans again.
> Pragmatists and realists could argue this forever

It doesn't really matter WHY it happens. Universal experience proves that
things go wrong often enough for one to disbelive people who claim otherwise
for systems which really matter - especially so for those which have to last
essentially for ever.

> I'd have thought a very deep shaft would suffice. Way down in
> bedrock, and by the time the waste re-appeared by tectonic
> subduction, its half-life would have been far exceeded, resulting
> in a little volcanic puff of lead

Indeed - many people have thought just that and there are various 'safe
'storage' schemes which rely on deep burial. What could go wrong? Who knows?
Can something go wrong? You bet! How do we know? Universal statistical
experience of human endavour.

> Well, you make your societal choices as to what standard of living
> you want (or indeed have thrust upon you). Woud we be safer as
> greenies living like Amish ? Probably,

I don't think so personally. I would be very surprised if the Amish
lifestyle mortality / danger etc rate is the lowest that can be achieved.
I'd expect that something with a more more technological bent would be
safer.

> but safety comes a distant
> second behind curiosity and the need to explore and invent

I assume that in that sentence the word "need" can be translated "desire"
;-)


       RM

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspamTakeThisOuTmitvma.mit.edu

2004\08\08@222324 by Russell McMahon

face
flavicon
face
part 1 2227 bytes content-type:text/plain; (decoded 7bit)

>> Arguably the
>> safest solution is to concentrate it very highly and
>> then dump it in deep space.

> But you're potentially giving someone else the problem -

Not really. Inner solar system space is a far from benign, high radiation
environment. The odd 'flask" of high concentration nuclear waste would make
an unnoticeable difference. And if the material is 'finely dovoded" and you
rupture the storage container once in space you get it spread infinitely
thinly.

> I think going the other way would be safer - into the
> Sun.

I just suggested "beyond escape velocity". No direction implied. But into
the sun is very much more expensive. Scrubbing off the energy required to
get to the sun is about as energy intensive as attempting to leave the solar
system.

>The little blip of radioactivity we're talking
> about wouldn't even be noticed against the vast energy
> there.

Yes. Same applies to anywhere in inner system space.

> > Would be rather expensive - but once it achieves
>> escape velocity its pretty safe. The odd one that
>> 'doesn't quite' excape is then just part of the cost of
>> doing busienss :-(

> That is the Big Problem, of course - fine when it works,
> potential disaster when it doesn't!  And failed
> rocket-launches aren't terribly rare, unfortunately.

Rocket launches can be made extremely reliable. How reliable depends on what
you wish to spend. The whole manned space program has had a single
(acknowledged) fatal launch event. (recovery has had several). The Atlas II,
III & V programs achieved 100% successs over 68 consecutive launches as at
the end of 2003. Don't know the stats since then but I don't think they've
had a loss since then (yet) - and that's an unmanned program. And it you
REALLY must you can make the payload able to survive re-entry. Then you just
have to find it :-).

Googles  ... Atlas .... 72 consecutive launches as at May 19 2004
       www.spacedaily.com/news/launchers-04l.html
There it goes .....


       RM

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestEraseMEspam.....mitvma.mit.edu




part 2 1556 bytes content-type:image/jpeg; (decode)

2004\08\08@223359 by Russell McMahon

face
flavicon
face
This came offlist - but had points worth addressing on list - no personal
stuff therein but won't say who sent it lest they don't want their views
identified.

> >Passive engineering designs can last
> > somewhat longer.

> The pyramids in Egypt come to mind.  And perhaps a sign on the outside
> saying that anone who entered would suffer a curse might be telling the
> truth.

Good example. Good for 2000 years ++ without a decent earthquake. Wonder how
they'd work for the longer lasting wastes.

> > Arguably the safest solution is to concentrate it very
> > highly and then dump it in deep space.

> I believe I recall reading an engineering analysis of this proposal years
> ago and they determined that this had substantially higher risk than any
> of the alternatives proposed.

Maybe - but it's the risk to US and NOW. Not our descendants sometime later.
And you could make the payload re-entry capable.
Success probability can be made better than 99%. That's pathetically low
compared to what other systems CLAIM for their safety levels, and amazingly
high compared to what is liable to actually be achieved by other means.

> Someone suggested dumping the containers in one of the deep ocean trenches
> and then provoking a landslide to bury the containers.

Experience to date suggests that nature can easily outwit such schemes.

> I think I remember someone suggesting using particle beams to tear apart
> the long lived material, turning it into much shorter lived material and
> making use of the resulting released energy.  It seems like they had a
> recycling process in mind where they had some percentage success with each
> pass through the system including time delays and filtering out the safe
> material each time.

That sounds promising if it can be made to work. Its owning the problem
rather than assing it on to our (putative) descendants.

I'm not anti-nuclear per se - just don't believe that anyone has solved the
problems that must be solved to make it closed-system safe.
(Interim solutions only need to exceed a human lifespan to be attractive to
many).




       RM

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspammitvma.mit.edu

2004\08\08@235708 by Jinx

face picon face
> Woud we be safer as greenies living like Amish ? Probably
>
> I don't think so personally. I would be very surprised if the Amish
> lifestyle mortality / danger etc rate is the lowest that can be achieved.
> I'd expect that something with a more more technological bent would
> be safer

If Amish or the like took full advantage of medical care, the best it's
ever been, and continued to refrain from those bugbears of modern
life - stress, overweight, inactivity, personal polution (smoking, over-
indulgence in a bevvy or two, cosmetics, electronic), I'm sure they'd
be as healthy as one could wish to be. Obviously you can't rule out
natural causes but you can minimise the external ones. Such a life-
style is definitely not fulfilling for most people, even with the upside
that the only serious injury risk on a daily basis might be being on
the receiving end of a kick in the head from a grumpy carthorse

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestEraseMEspamEraseMEmitvma.mit.edu

2004\08\09@000815 by Bob Axtell

face picon face
I have enjoyed this thread quite a bit. Love your comments Russell:
shows a LOT of clear thought.

At the tender age of 19, as an EE student, I suddenly realized what
mankind was doing to itself
just to generate electricity. I was appalled at how truly stupid nuclear
power was. I became suddenly
suspicious of governments, and have been increasingly suspicious the
older I became.

The problems of nuclear energy are so many and so aggregious that this
forum could never come close.

But the most serious problem is nuclear waste. To be rendered safe,
these highly corrosive, incredibly
toxic compounds must be physically contained for 50,000 years or more.
People just don't understand
how insurmountable that problem is. I'm a great lover of engineering
tasks, but _I_ have ZERO confidence
that this can be done.

Even sadder than this, is the fact that more energy falls on the PARKING
LOT of the Hanford nuclear facility
every day (as sunlight) than Hanford generates. Where I live (southern
AZ) sunlight reaches over 200w per
square meter for 8+hrs/day for 320+ days a year. Free electricity. Clean
electricity. No nuclear waste.
a 10-mile square of  useless arid desert could generate the entire
electricity needs of the US and Canada
combined with enough left over to fill the tanks of everybodyy's
electric car. But is anybody interested?

Just think. We elected politicians that CAUSED this kinda stuff to
happen. Scary, really scary, isn't it?

--Bob


Russell McMahon wrote:

{Quote hidden}

--

Note: Attachments must be sent to
RemoveMEattachspam_OUTspamKILLspamengineer.cotse.net, and
MAY delay replies to this message

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@001842 by Jake Anderson

flavicon
face
last I calculated to power the world we needed
PV cells roughly half the size of australia?.

people say solar is so nice but they seem to
forget what goes into making solar cells is pretty
damn nasty stuff. Aresnic and a whole bunch of
other poisinous substances.

If the loop can be closed in nuclear power
(IE such that the end result is power and
very low level waste) then it seems that
it would be the better solution.

Of course the hydrogen fusion reactor is the
(second, rember ZPE) best answer to this.
as always you build it i'll sell it.
any PIClisters have their own pet fusion
reactor ideas? i'd be glad to hear them.



> {Original Message removed}

2004\08\09@004409 by Engineering Info

picon face
I don't have any "pet fusion reactor ideas" but I do have two Fuel Cells
instead which is close enough for me. Cost about $9,000 each.

BTW Russell, the newest version are now up to 45 watts @ 12 volts
continous power instead of the 30 watt versions I currently have.  Put
two of those together and you'll be close to that 100 watts you wanted.
Aproximately 17" tall by 3" diameter bottles for about 1000WHrs of power.

Jake Anderson wrote:

{Quote hidden}

>>{Original Message removed}

2004\08\09@013459 by Russell McMahon

face
flavicon
face
> last I calculated to power the world we needed
> PV cells roughly half the size of australia?.

US energy consumption 1998 was about 100 billion kWh (E11 kWh)

A square meter at my lattitude gets usable thermal energy of from about 3
kWh/day winter to about 8 kWh/day summer.
If we ever went solar we would expect efficiencies of at least 30% so lets
average the above and divide by 3 to get about 1.5 kWh/m^2/day or about
550 - say 500 kWh/m^2 per annum.
So to get E11 kWh/year needs about 2 x E8 m^2 or about 200 square kilometres
of solar cells given the above assumptions . E&OE as usual.
That doesn't sound bad - especially if you can bias generation locations
towards the south. I suspect you can find a 1000 or so square km in NM/
Arizona/ Texas for such purposes.

At present cost is prohibitive. Payback periods are inferior to those for
more usual technology. But if research keeps plugging away as it has been
for years now we'll hopefully arriove at the level in Kim Stanley Robinson's
utterly superb "Red Mars" where you arrive at a site, roll out as much solar
cloth as you need and start cooking.

> Of course the hydrogen fusion reactor is the
> (second, rember ZPE) best answer to this.
> as always you build it i'll sell it.
> any PIClisters have their own pet fusion
> reactor ideas? i'd be glad to hear them.

I'm looking forward to "them" importing solar Helium III mined from the
Moon's surface. At the rate things are going it may be the Chinese who ar
doing it :-)


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@080512 by Dave Tweed

face
flavicon
face
Russell McMahon <RemoveMEapptechTakeThisOuTspamspamPARADISE.NET.NZ> wrote:
> I just suggested "beyond escape velocity". No direction implied. But into
> the sun is very much more expensive. Scrubbing off the energy required to
> get to the sun is about as energy intensive as attempting to leave the
> solar system.

Not at all. You just need an orbit that intersects the sun, regardless of
the energy level. A slingshot past Venus, for example, could do it very
easily, but there is of course the risk that you crash your load into Venus
itself instead of sending it past.

-- Dave Tweed

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@082622 by Howard Winter

face
flavicon
picon face
Russell,

On Mon, 9 Aug 2004 14:28:19 +1200, Russell McMahon wrote:

> I'm not anti-nuclear per se - just don't believe that anyone has solved the problems that must be solved to
make it closed-system safe.

I agree, if the closed system you mean is Earth! :-)

> (Interim solutions only need to exceed a human lifespan to be attractive to many).

I think you're being too kind - to some people they only have to exceed *their* life, and to others just their
term of office...

Reading the Three Mile Island report that someone cited here the other day, I am amazed that apparently the
most effort by far seemed to be expended in arse-covering and PR, trying to convince people that there was no
need to panic.  As if the worst thing that could happen was panic!  It looked like they'd much rather people
were calmly contaminated than escaped that by evacuating without being told to do so.   Hmmmm....

I think if I lived close to a nuclear plant (not that I would!) at the first hint of an incident I would
calmly go somewhere else for a few days until things were sorted out.  I'd also have my own geiger-counter so
I could check for myself.  Oh - I do have one!  (Remembering that the first the West knew of Chernobyl, was an
individual in Sweden who noticed a sharp rise in "background" radiation - currently reading 0.03mr/hr, so
things seem to be OK at the moment :-)

Am I right in thinking that NZ has no nuclear power plants?

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@103323 by Howard Winter

face
flavicon
picon face
Bob,

On Sun, 8 Aug 2004 21:03:25 -0700, Bob Axtell wrote:

> To be rendered safe, these highly corrosive, incredibly toxic compounds must be physically contained for
50,000 years or more.

And just to put that in perspective, 50,000 years ago was Paeleolithic times - early stone age.  Ten times as
long ago as the building of Stonehenge!  When we look at how things have changed in the last 100 years it's
just staggering to think how different the World will be in 500, let alone 50,000 years's time.  The only
thing we can be sure of is that if humans are still alive, there will still be politicians...  I wish I hadn't
thought of that, now I'm feeling all depressed!

> Even sadder than this, is the fact that more energy falls on the PARKING LOT of the Hanford nuclear facility
every day (as sunlight) than Hanford generates. Where I live (southern AZ) sunlight reaches over 200w per
square meter for 8+hrs/day for 320+ days a year. Free electricity. Clean electricity. No nuclear waste. a
10-mile square of  useless arid desert could generate the entire electricity needs of the US and Canada
combined with enough left over to fill the tanks of everybodyy's electric car. But is anybody interested?

I am, but there's the Rest of the World too, and here in SE England the energy landing is a bit more modest.
I've put a couple of solar panels on the roof of my shed as an experiment, and over the past few weeks when on
a sunny day (about half of them) they collect about 5 AmpHours at 24V, and they're probably 1/3m^2 so I'm
getting a maximum of 36Wh/m^2 on a good day.  I'd love to be use free green energy, but my garden isn't big
enough!  Not to mention the cost - those panels cost about £150 (say US$270) so they are never going to be
cheaper than buying from the grid, unless there's a major improvement (at least an order of mangitude) in
price/performance.  While green is good (IMHO) I'm not going to pay 10x as much for it!

> Just think. We elected politicians that CAUSED this kinda stuff to happen. Scary, really scary, isn't it?

Indeed, but at least over this side of the pond there is a viable "Green Party" to vote for.  They don't get
that many votes, but they do manage a presence in the various parliaments.  As a matter of interest, are there
any politicians in any of the various US legislatures whose main platform/policy is "green"?  The indications
we have from the news here are that there aren't, but that may because it's just not interesting enough to
feature.

Cheers,



Howard Winter
St.Albans, England

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@104916 by Howard Winter

face
flavicon
picon face
Jake,

On Mon, 9 Aug 2004 14:18:25 +1000, Jake Anderson wrote:

> last I calculated to power the world we needed
> PV cells roughly half the size of australia?.

And since most of Australia is uninhabited, that sounds like the solution!  We just need really long power
cables...  :-)

> people say solar is so nice but they seem to
> forget what goes into making solar cells is pretty
> damn nasty stuff. Aresnic and a whole bunch of
> other poisinous substances.

Yes, but if they are incorporated into the products, and not discharged as waste, I don't see this as a
problem.  Arsenic is a naturally-occurring element, so I don't think having it somewhere on the planet is
necessarily a problem.  NIMBY, of course!  :-)))

The great thing about Solar / Wind / Wave energy is that the energy is arriving now, so we aren't using
anything up.  All the other forms are using energy that came from the Sun some time ago (or from even earlier
stellar activity) so are depleting whatever it is, and will eventually run out.  Obviously we won't run out of
Hydrogen as long as we have the energy to take it from water (running out of water is unthinkable - if we did
the energy problem would be lowered in our priorities!) but I'm not sure where *that* energy comes from in a
fusion system.

> If the loop can be closed in nuclear power (IE such that the end result is power and very low level waste)
then it seems that it would be the better solution.

There's still the accident potential, though.  I don't know what the waste from a graphite-core
pressurised-water reactor is like, but the former inhabitants of Pripyat would have something to say about the
results of an accident!

> Of course the hydrogen fusion reactor is the (second, rember ZPE) best answer to this.  as always you build
it i'll sell it.  any PIClisters have their own pet fusion reactor ideas? i'd be glad to hear them.

Yes, but I'm keeping them to myself until I've perfected the process... probably about 2080!  ;-)

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@105956 by David VanHorn

flavicon
face
At 09:32 AM 8/9/2004, Howard Winter wrote:

>Bob,
>
>On Sun, 8 Aug 2004 21:03:25 -0700, Bob Axtell wrote:
>
>> To be rendered safe, these highly corrosive, incredibly toxic compounds must be physically contained for 50,000 years or more.
>
>And just to put that in perspective, 50,000 years ago was Paeleolithic times - early stone age.  Ten times as long ago as the building of Stonehenge!


There's a book on this, called "Deep Time"  Highly recommended.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@182137 by Russell McMahon

face
flavicon
face
> Am I right in thinking that NZ has no nuclear power plants?

Whole country is officially "nuclear free" including no nuclear ship visits.
This is the subject of much ongoing discussions, especially near elections.

We have medical and industrial radiation sources but no reactors whatsoever.
Nearest is "Lucas Heights" in Sydney about 1500 miles away across the ocean.

A key point is that the shape of our country and our dependence on primary
exports and "clean green" image mean that ANY nuclear accident here would be
a horrendous disaster financially. The true risk-benefit of nuclear is much
worse for us than for many countries. eg imagine that Chernobyl had occurred
in Luxembourg. How would it have affected them there compared to where it
actually happened. In NZ a 60 mile radius circle ALWAYS touches the sea on
both sides.

We have large amounts of hydro generation, natural gas, coal, some oil, some
wind power. Never "enough" of course.


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@184710 by Jinx

face picon face
> > Am I right in thinking that NZ has no nuclear power plants?
>
> Whole country is officially "nuclear free" including no nuclear ship
> visits. This is the subject of much ongoing discussions, especially
> near elections

I wonder if nuclear power would ever be a goer in NZ anyway. A
country like the US or France with a large population and high
electricity demand can afford to have excess nuclear capacity. If
one plant goes off-line, there's another available to fill the gap. In
a country like NZ with a fraction of the population and an even
smaller fraction of heavy industry, the energy needs are far more
modest and you have to consider what could be available as
backup. From memory the last time there was an "energy crisis"
it took quite some time to get mothballed fossil-fuel plants up and
running. As late as last month there were still no firm plans to get
Marsden B running - it's 26 years old and never been used !

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\09@203750 by Russell McMahon

face
flavicon
face
Jinx, on NZ power:

> From memory the last time there was an "energy crisis"
> it took quite some time to get mothballed fossil-fuel plants up and
> running. As late as last month there were still no firm plans to get
> Marsden B running - it's 26 years old and never been used !

As of recently, the government (ie taxpayers) are now paying for surplus
generation capacity to be held in reserve for emergency use. This was always
done as part of the order of things before privatisation. Nowadays it has to
be done explicitly. It just took them a few decades and a couple of
emergencies to realise this. Actually, it's really to *acknowledge it
publicly* - the people who knew have always known that where we were going
with privatisation and market forces was untenable in the long term. Market
forces are fine things until your hot water isn't on a winter morning and
your lights and heaters won't on a winter night, and your (manu)factories
can't during random periods of peak demand  - as all happened. No amount of
market forces (ie realistically conceivable  magnitude ones) are liable to
pull you out of that in an acceptably short time frame - the price/demand
cycle is far too slow, but then political forces step in - especially with
our finely balanced political position. So the government pays for the
reserve power that we always used to have and we all pay for it in taxes
like we always used to do.



       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

2004\08\10@065942 by Howard Winter

face
flavicon
picon face
Jinx,

On Mon, 9 Aug 2004 09:07:27 +1200, Jinx wrote:

> I had this wee debate with someone on another group at the
> weekend about engineering failures, in particular power plants
> (the thread was originally about the electricity grid in NZ and
> nuclear waste)
>
> He wrote :
>
> > Yes, human error can be reduced, but it can't be eliminated
> > altogether. The machines have to be run by humans

Well actually they don't, but they have to be designed by humans and so far not a single person has
demonstrated 20/20 foresight...  the problem of allowing for *every* possibility without human intervention is
much more difficult to solve than leaving humans in the loop, even if just in an intervention role.

> My argument was that if the spirit is willing, any engineering
> undertaking can be made reliable and safe. Maybe that's just
> me looking through my idealists' rose-tinted glasses.

'Fraid so!  If it was purely an engingeering problem then it's possible, but nothing is ever free from
economics, politics, ego, and nay-saying busybodies.  And time-limits, of course!

> Redundancy, backup, failsafes, use of appropriate technology,
> reducing human involvement and so on. My claim was that any
> failure could be traced back to corruption, cost-cutting or care-
> lessness/incompetence, all of which are fixable

Well first of all "reducing human involvement" is a very dangerous aim, especially if we're talking about
safety.  If I'm in an aircraft with a problem, I'm much rather the person with the responsibility of sorting
it out was in a position to be first on the scene of the accident, as now, rather than at a cocktail party
having just been presented with an "automation achievement" award!  :-)

> For example, nuclear waste is not dangerous if managed properly, which it can be. And a mains socket isn't
dangerous if you don't stick your finger in it

Yes, but something that is "not dangerous as long as everything is done properly" *is* dangerous!
This is completely ignoring the human factor - and removing humans isn't the answer to that.  :-)  Electrical
sockets in the UK have shields over the two "live" entrances, so only inserting a plug will retract them.
Most adults don't stick screwdrivers, knives, forks, whatever, into sockets, but children do, and with the
earlier socket designs without shields, a number of them were killed by it.  You might say "Darwin applies",
but even I am not *that* heartless!

The safe way to design things, including processes like "safe handling" of things like nuclear waste, is to
assume that the person involved will not necessarily act correctly at all times, and especially times when
things are going wrong.  To do otherwise is setting up for disaster.

For a programmer of weapons-guidance systems, "fail safe" is obviously the watchword.  Bt what is safe in this
case?  In the Falklands war there was an incoming "hostile" which was identified by a ship's radar operators -
they designated it as a target, and shortly afterwards its radar image crossed over another (also hostile)
aircraft.  When they seperated, the operator gave the command to fire, but the system did nothing.  Because
the radar image had merged and seperated, the system wasn't sure it still had the designated target, so it did
the "safe" thing and didn't fire.  I believe there was no way for the software to say to the operator: "I'm
not sure - confirm if this is the one you mean" - it just refused to fire, and by the time the humans had had
time to realise there was a problem, there wasn't time to re-designate the target and fire.  The result was
that the hostile got through and sank a ship.

All good disasters are a culmination of a number of factors, absence of any one of which would have prevented
them.  That's because we tend to design for individual failures or problems, rather than a cascade of them.
For this reason I believe it is very hard to ever say that something is safe, meaning that it will never fail
catastrophically.  Having a human on hand to second-guess the one who designed the system is usually a Good
Thing, because we are fairly good at re-assessing an unexpected situation and coming up with a novel solution.
But, as in the Theriac-25 saga, the human has to know that they are expected to do this, and "trusting the
machine" leading to complacency and abdication of responsibility must be avoided.  A machine that "cries wolf"
is a really Bad Thing in this case...

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@075249 by Jinx

face picon face
Howard, you're right with all your points of course. Humans should
be involved in the process - but hopefully with the Peter Principle not
in effect

What's your view re a self-aware machine that could make human-
like decisions ? Who knows when in the future this could come to
pass. I'm guessing some time away, but an unforeseen breakthrough
could happen any time to speed things up

For example, some discussion points

http://ebtx.com/nman/nman14e.htm

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@082025 by Russell McMahon

face
flavicon
face
> What's your view re a self-aware machine that could make human-
> like decisions ? Who knows when in the future this could come to
> pass. I'm guessing some time away, but an unforeseen breakthrough
> could happen any time to speed things up

What would it take for you (whoever you are) to accept that a machine was
"truly" self aware?
Or intelligent? Or thinking? Or alive?
Would its passing the Turing test be enough to convince you of any of the
above?
Even if it could pass the TT in all disciplines, would that suffice?


       RM

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@083100 by Jake Anderson

flavicon
face
my question is if you built such a machine
could you in good conscience turn it off?

{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@083515 by D. Jay Newman

flavicon
face
> > What would it take for you (whoever you are) to accept that a machine was
> > "truly" self aware?
> > Or intelligent? Or thinking? Or alive?
> > Would its passing the Turing test be enough to convince you of any of the
> > above?
> > Even if it could pass the TT in all disciplines, would that suffice?

> my question is if you built such a machine
> could you in good conscience turn it off?

(Corrected for top-quoting)

This is an interesting question. A coworker and I were discussing this
yesterday.

If I believed that I might have created something sentient (even as aware
as an animal I would consider for a pet) I would feel that I have no right
to destroy it.

In other words, if I am convininced that it even *might* be self-aware
I would feel obligated to treat it as if it provisionally were self-aware.

Not that I think I'll have to worry about this for several years.  :)
--
D. Jay Newman           ! DCX - it takes off and lands base down,
jaySTOPspamspamspam_OUTsprucegrove.com     !       as God and Robert Heinlein intended.
http://enerd.ws/robots/ !

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@084137 by Jinx

face picon face
> What would it take for you (whoever you are) to accept that a machine
> was "truly" self aware?

Able to safeguard its own existence must be at/near the top of a list

For example, how does a young animal, especially one with no parental
guidance (eg most insects or reptiles) know to avoid prey ? Would it be
cheating to imbue a machine with "instinct" ?

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@084346 by Jake Anderson

flavicon
face
{Quote hidden}

what if you programmed it such that it didnt mind being switched off?
the "pig" in hitch hikers guide to the galaxy's resteraunt at the
end of the universe. (a sentient animal genetically engenered to
want to be eaten and which rather vocally advocates that fact)

is it ethical to create a life form without a "will to live"
or self preservation. seems to me it would be the ultimate slave then
If you do create a sentience with the will to survive will it
"kill you and eat your babies" in a darwinian move if given half
a chance. Or might it just out compete us, and use us in the
same way we use chimps now? (for psyc type stuff)

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@084930 by Jinx

face picon face
> my question is if you built such a machine
> could you in good conscience turn it off?

Remember the Tamagotchi fad and the basket-cases (mostly
temporary I'm sure) it created in its wake ? Most parents would
have happily stamped on them with no qualms at all

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@085139 by Howard Winter

face
flavicon
picon face
Jinx,

On Tue, 10 Aug 2004 23:54:05 +1200, Jinx wrote:

> Howard, you're right with all your points of course. Humans should
> be involved in the process - but hopefully with the Peter Principle not
> in effect
>
> What's your view re a self-aware machine that could make human-like decisions ?

In a nutshell:  I hope it doesn't happen, especially in my lifetime!

The ramifications (if it's even possible) are huge, assuming it is given the reponsibility to make decisions
that matter, rather than just an experiment.  How do you make sure it makes the "right" decision?  You'd have
to teach it somehow what is "right" and "wrong" and make sure it follows the right path, while it can see that
the rest of the World does wrong things.  Would you invent some sort of religion to keep it in line?

Don't forget that people "go wrong" due to problems in upbringing, mental diosorders, and other causes.  Are
we prepared to allow a machine to "go wrong"?  If it's self-aware, would it be immoral to turn it off ("kill"
it)?

If it gets creative or adventurous, how do you stop it doing something awful?  Just look at the way kids (and
some adults!) act when they think they can get away with it.  I forget which aircraft it was (may have been
the 747) where during a test flight the test-pilot thought "I think I can barrel-roll this thing" and did so.
Afterwards someone senior in the company took him aside and said:  "You know it can do it safely, I know it
can do it safely - Don't do it again!"  Imagine what a machine might do if it didn't have a conscience?  I
cannot even begin to work out how to teach conscience!  :-)

I can see a whole new profession evolving: Machine Psychology (Heaven help us! :-)

> Who knows when in the future this could come to pass. I'm guessing some time away, but an unforeseen
breakthrough could happen any time to speed things up

Personally I think it's unlikely.  We don't even know what makes awareness - a complex set of connected nerves
somehow doesn't seem to be sufficient to my mind (!).  The Internet may be approaching the complexity of a
brain (not ours, perhaps something a lot further down the food chain) but is it showing even a glimmer of
awareness?  I think not, and I don't think it could.  Our brains continue to act while we are asleep, and they
are no less complex at that time, yet we are unconscious.  So the question of what makes us aware perhaps
could look at how we turn awareness off for a third of our lives, and how we turn it on again when needed,
without apparently losing anything in between.

> For example, some discussion points
>
> http://ebtx.com/nman/nman14e.htm

Very interesting, and it certainly points out some of the problems I can see.  In the film "Saturn 3" a robot
with an artificial brain (although apparently a biological one) was taught by being connected to a person who
had a terminal installed in his head for the purpose.  It also showed the problems that followed from the
teacher being unsuitable for the task, when it started attacking the people...  I rest my case!  :-)

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@085347 by D. Jay Newman

flavicon
face
> what if you programmed it such that it didnt mind being switched off?
> the "pig" in hitch hikers guide to the galaxy's resteraunt at the
> end of the universe. (a sentient animal genetically engenered to
> want to be eaten and which rather vocally advocates that fact)

I think that it was a cow.

However that may be, I am not talking about deleberate creation, but
rather creating sentience accidentally. Who knows, it may happen...

> is it ethical to create a life form without a "will to live"
> or self preservation. seems to me it would be the ultimate slave then

I wouldn't want to create something without a "will to live". While
I build machines to help me and I have no real problem with machine
"slavery", I would prefer that it tries to preserve itself.

As an example, I have a dog. I expect Bug (the dog) to obey me. When
push comes to shove, I expect my will will prevail.

However, I do try to do the best I can to provide her with a good life.

> If you do create a sentience with the will to survive will it
> "kill you and eat your babies" in a darwinian move if given half
> a chance. Or might it just out compete us, and use us in the
> same way we use chimps now? (for psyc type stuff)

When we can create something that intelligent, I hope that we keep it
carefully guarded.
--
D. Jay Newman           ! DCX - it takes off and lands base down,
KILLspamjayspamBeGonespamsprucegrove.com     !       as God and Robert Heinlein intended.
http://enerd.ws/robots/ !

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@093716 by Howard Winter

face
flavicon
picon face
Jinx,

On Wed, 11 Aug 2004 00:51:38 +1200, Jinx wrote:

> > my question is if you built such a machine
> > could you in good conscience turn it off?
>
> Remember the Tamagotchi fad and the basket-cases (mostly
> temporary I'm sure) it created in its wake ? Most parents would
> have happily stamped on them with no qualms at all

Not to mention the woman who was driving when the Tamagotchi on her keyring started bleating to be fed - while
her attention was distracted in trying to "save its life", the car went off the road and she lost her own.


Howard Winter
St.Albans, England

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@095542 by Russell McMahon

face
flavicon
face
> > What would it take for you (whoever you are) to accept that a machine
> > was "truly" self aware?
>
> Able to safeguard its own existence must be at/near the top of a list
>
> For example, how does a young animal, especially one with no parental
> guidance (eg most insects or reptiles) know to avoid prey ? Would it be
> cheating to imbue a machine with "instinct" ?

Ability  to protect ones-self appears to be *inversely* correlated with
intelligence and generalisation. "Higher" animals (eg mammals) are the most
helpless at birth. Humans are essentially unable to fend for themselves for
in excess of 5% of their lifespans. (Arguable at what age a human could
reasonably survive if treated as an "adult". Most 5 year olds wouldn't. Many
10 year olds would)

And, do remember :-)

"The Skynet Funding Bill is passed. The system goes on-line August 4th,
1997. Human decisions are removed from strategic defense. Skynet begins to
learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time,
August 29th. In a panic, they try to pull the plug."
'And Skynet fights back.'

25 days to become self aware seems fairly slow :-)

I don't believe we have more than the vaguest feel for any of this.
God knows where we will end up when we start to play with this area in
earnest*.
Experience in every conceivable field of human endeavour in zillions of
cases show that some people will push the limits of ability with absolutely
no knowledge of or concern for consequences. It's happening now in the
Genetic Engineering field and whether or not we will reap the whirlwind
there as we seek the holy grail is entirely outside our control.

In the field of machine intelligence we have no real idea of what "self
aware" really entails or even means.
We have no idea of what it would take to be sure a machine creation
"genuinely" had such an ability.
No idea what machine intelligence entails, or the consequences.
Don't even know how WE work. Lots of ideas, but we can't make something
which we can convince ourselves "thinks" as genuinely as the simplest
biological entity which we agree does think. At higher levels we can make
something which emulates eg a cat, but we have no idea whether our models
are useful or meaningful cat models.

Interesting times ahead are a certainty.



       RM



* Just as well ;-)

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@132339 by Wouter van Ooijen

face picon face
> What's your view re a self-aware machine that could
> make human-like decisions ?

Wouldn't that just replace the bad decisions made by humans with equally
bad descisions made by machines?

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@162149 by M. Adam Davis

flavicon
face
Equally bad, but /faster/.  A bad decision made quickly can be as good
as a good decision made too late, right?  If you can make 1000's of
decisions a second, you always have time to do it over...  ;-)

-Adam

Wouter van Ooijen wrote:

{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@184942 by John Tserkezis

flavicon
face
M. Adam Davis wrote:

> Equally bad, but /faster/.  A bad decision made quickly can be as good
> as a good decision made too late, right?  If you can make 1000's of
> decisions a second, you always have time to do it over...  ;-)

 More importantly, who would get the blame, the robot or designer?

--
Linux Registered User # 302622                         <http://counter.li.org>

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\10@193341 by M. Adam Davis

flavicon
face
John Tserkezis wrote:

> M. Adam Davis wrote:
>
>> Equally bad, but /faster/.  A bad decision made quickly can be as good
>> as a good decision made too late, right?  If you can make 1000's of
>> decisions a second, you always have time to do it over...  ;-)
>
>
>  More importantly, who would get the blame, the robot or designer?
>
This may depend on how the first real EULA judgement turns out.  A lot
of people speculate that an EULA doesn't really protect the maker of the
software/hardware, but AFAICT a case has never gone to court.  I suspect
that manufacturers may not want it tested, and would rather settle than
have to change the EULA or accept some liability for their software.

Considering that the software industry is still really in its infancy, I
expect things will be completely different by the time we get to robots
that could accept blame and punishment.  Of course, if a robot can
accept blame and punishment, then one could argue that they have
feelings and are 'alive', and as such should be given rights (ie, no
punishment without rights, taxation without representation, etc).  Once
this occurs then you must keep them in good working order unless you
want to be punished as their owner.  Robots can live forever, so the
problem then becomes worse than simply who gets the blame - it becomes
about robot overpopulation, robot rights, etc.  Then humans do end up
serving robots, even if they are not forced to do so by the robots, but
by their own laws.

I doubt we'll ever get to that point (thinking, feeling robots),
though.  Interesting thought experiments...

-Adam

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics

2004\08\11@001609 by William Chops Westfield

face picon face
On Aug 10, 2004, at 5:30 AM, D. Jay Newman wrote:
>
>> my question is if you built such a machine
>> could you in good conscience turn it off?
>>
It doesn't matter so much, since you can back it up to offline storage
and turn it back on again.  Or duplicated it as much as you want.
That's a big FEATURE of digital; lossless copy/backup.  (used rather
effectively in Hogan's "The Two Faces of Tomorrow")

BillW

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@074527 by D. Jay Newman

flavicon
face
> On Aug 10, 2004, at 5:30 AM, D. Jay Newman wrote:
> >
> >> my question is if you built such a machine
> >> could you in good conscience turn it off?
> >>
> It doesn't matter so much, since you can back it up to offline storage
> and turn it back on again.  Or duplicated it as much as you want.
> That's a big FEATURE of digital; lossless copy/backup.  (used rather
> effectively in Hogan's "The Two Faces of Tomorrow")

You're making the assumption that the AI is built using solely digital
technology.

Also I think the question is phrased wrong. I would have asked:
 Could you in good concience destroy the machine or turn it off
 perminantly?
--
D. Jay Newman           ! DCX - it takes off and lands base down,
EraseMEjayspamEraseMEsprucegrove.com     !       as God and Robert Heinlein intended.
http://enerd.ws/robots/ !

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@081020 by Russell McMahon

face
flavicon
face
> > It doesn't matter so much, since you can back it up to offline storage
> > and turn it back on again.  Or duplicated it as much as you want.

> You're making the assumption that the AI is built using solely digital
> technology.

It's extremely likely that such a creation would be at least in part (ie
fatally) backuppable and turn offable. Also highly likely that digital
techniques will be a significant or total part of it. Of course, this
likelihood could change drastically with new discoveries. We may find how to
store information reliably and compactly and with low energy in an organic
matrix using low level analog signals - perhaps with some sort of level
triggering and stimulus / response characteristic. I think that this could
be made quite compact if you could work out how to build it. In such a case
backup may be difficult.

> Also I think the question is phrased wrong. I would have asked:
>   Could you in good concience destroy the machine or turn it off
>   perminantly?

In the case of an "AI" - definitely if necessary*. If turnoff became
desirable for some compelling reason then temporary shutdown would be by far
the preferred route, and failure to not build such a disabling capability
into a device potentially more "intelligent" than a human would be a
ludicrous oversight. As would allowing a system to progress to such a stage
that it produced such a device that did not have this capability. Something
akin to Asimov's laws would be essential. However, forcing ALL players to
implement such precautions would be impossible and the probability that
someone would circumvent such provisions is 1 (based on a very very very
large case history of human behaviour).(Current GE developments provide a
prime example of how willing we are to risk destroying the whole human race
in the pursuit of knowledge/financial reward/power/Nobel prizes and
more.)(Believe me not?: check this space (if still viable) in 10 years
time).




       Russell McMahon

* I'd 'probably' be willing under very very very qualified conditions (none
of which I'd like to even think about) to kill another human being. Not
something I hope will ever happen. Given that, a man-made creation would be
a far easier choice. If anyone can't see the difference then odds are that I
can't explain it to them.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@194115 by Robert B.

flavicon
face
----- Original Message -----
From: "Russell McMahon" <@spam@apptech@spam@spamspam_OUTPARADISE.NET.NZ>
<snip>
>
> "The Skynet Funding Bill is passed. The system goes on-line August 4th,
> 1997. Human decisions are removed from strategic defense. Skynet begins to
> learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern
time,
> August 29th. In a panic, they try to pull the plug."
> 'And Skynet fights back.'
>
> 25 days to become self aware seems fairly slow :-)
>

<smip>

It was, after all, only learning at a geometric rate.  I've wondered ever
since I first saw the movie why a machine so well connected learned so
relatively SLOW.  Geometric rate means, of course, that it was learning
linearly.  It seems to me that the more it knew the more quickly it could
make sense of new data, and thereby learn at some exponential rate.

FEX: if you know nothing at all about PICS (on topic? ;-) it would take a
large amount of basic digital/analog electronics before you could make any
sense of the datasheets.  But once you had the basic knowledge you could
easily digest many many datasheets and learn at a faster rate.

But then again, if it can only pipe so much data through per unit time, I
guess it could never learn any faster than linearly.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@203549 by Jinx

face picon face
> It was, after all, only learning at a geometric rate.  I've wondered
> ever since I first saw the movie why a machine so well connected
> learned so relatively SLOW

Number 5 in Short Circuit ?

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@203550 by Russell McMahon

face
flavicon
face
> It was, after all, only learning at a geometric rate.  I've wondered ever
> since I first saw the movie why a machine so well connected learned so
> relatively SLOW.  Geometric rate means, of course, that it was learning
> linearly.  It seems to me that the more it knew the more quickly it could
> make sense of new data, and thereby learn at some exponential rate.

Geometric is in fact exponential.

An arithmetic progression is

   Tn+1 = Tn + k

Geometric is

   Tn+1 = Tn x k

So

       Tn+x = Tn x K^x

The bigger it gets the faster it grows.
(The bigger it gets the bigger it gets).

25 days to learn enough to decide that you are sure you are able to take
over the world is actually quite fast by our standards. Even Hitler took a
decade or more. It had to ensure sensor adequacy, resources, security and
more.

In the event, it got it wrong. We think.
(So did Hitler - but we largely have the execrable Joseph Stalin (and his
100 million or so largely commendable friends) to thank for that)
It probably should have waited 26 days.


       RM

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@204213 by Russell McMahon

face
flavicon
face
> > It was, after all, only learning at a geometric rate.  I've wondered
> > ever since I first saw the movie why a machine so well connected
> > learned so relatively SLOW
>
> Number 5 in Short Circuit ?

Terminator.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@204422 by Robert B.

flavicon
face
I'm missing something here...   what is #5 in short circuit?


----- Original Message -----
From: "Jinx" <spamBeGonejoecolquittspamKILLspamCLEAR.NET.NZ>
To: <.....PICLISTspam_OUTspamMITVMA.MIT.EDU>
Sent: Wednesday, August 11, 2004 8:36 PM
Subject: Re: [OT:] Fatal software disaster - lessons to learn


> > It was, after all, only learning at a geometric rate.  I've wondered
> > ever since I first saw the movie why a machine so well connected
> > learned so relatively SLOW
>
> Number 5 in Short Circuit ?
>
> --
> http://www.piclist.com hint: The PICList is archived three different
> ways.  See http://www.piclist.com/#archives for details.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\11@204835 by Randy Glenn

picon face
#5 was the designation of a defense robot in the movie Short Circuit
which became intelligent and self-aware due to a lightning strike.
Perhaps better known by the name "Number Johnny Five"

On Wed, 11 Aug 2004 20:44:08 -0400, Robert B. <TakeThisOuTpiclist.....spamTakeThisOuTnerdulator.net> wrote:
> I'm missing something here...   what is #5 in short circuit?
>
>
>
>
> {Original Message removed}

2004\08\11@205043 by M. Adam Davis

flavicon
face
Number 5 is a character in the movie Short Circuit:
http://www.imdb.com/title/tt0091949/

One of several military destined robots, he gets zapped by lightening
and becomes self aware.  Hilarity ensues.

-Adam

Robert B. wrote:

>I'm missing something here...   what is #5 in short circuit?
>
>
>{Original Message removed}

2004\08\11@210914 by Jinx

face picon face
> > Number 5 in Short Circuit ?
>
> Terminator

I know - I was comparing "actors" ; one made of metal, one
made of wood (work it out)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

2004\08\12@054516 by Peter L. Peres

picon face
>I don't have any "pet fusion reactor ideas" but I do have two Fuel Cells
>instead which is close enough for me. Cost about $9,000 each.
>
>BTW Russell, the newest version are now up to 45 watts @ 12 volts
>continous power instead of the 30 watt versions I currently have.  Put
>two of those together and you'll be close to that 100 watts you wanted.
>Aproximately 17" tall by 3" diameter bottles for about 1000WHrs of power

Hmm .. add two more and you got a minimal car (tricycle or bicycle) drive.
200W is considered to be the power needed to run a bicycle in flat terrain
at a reasonable speed. Then you'd have a slow bicycle that cost as much as
5 compact cars. How long until you break even given the cars could run on
LPG conversion (done for $1000-2000 per car) ?

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email TakeThisOuTlistservKILLspamspamspammitvma.mit.edu with SET PICList DIGEST in the body

2004\08\12@060927 by Russell McMahon

face
flavicon
face
> >I don't have any "pet fusion reactor ideas" but I do have two Fuel Cells
> >instead which is close enough for me. Cost about $9,000 each.

> >BTW Russell, the newest version are now up to 45 watts @ 12 volts
> >continous power instead of the 30 watt versions I currently have.  Put
> >two of those together and you'll be close to that 100 watts you wanted.
> >Aproximately 17" tall by 3" diameter bottles for about 1000WHrs of power
>
> Hmm .. add two more and you got a minimal car (tricycle or bicycle) drive.
> 200W is considered to be the power needed to run a bicycle in flat terrain
> at a reasonable speed. Then you'd have a slow bicycle that cost as much as
> 5 compact cars. How long until you break even given the cars could run on
> LPG conversion (done for $1000-2000 per car) ?

Once they double the watts per power for the same money (any time now) the
break even, for my blue-sky never happen project, would be a year's use.
It's hard for a person to carry a compact car with LPG conversion (let alone
5 :-) ).

Such fuel cells are going to be really niche for a long time yet - but the
niches are getting bigger by the month.


       RM

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservspamRemoveMEmitvma.mit.edu with SET PICList DIGEST in the body

2004\08\12@083307 by Bob Ammerman

picon face
Probably the start of the genre:

Colossus: The Forbin Project

Anybody else remember?

Bob Ammerman
RAm Systems

{Original Message removed}

2004\08\12@083722 by D. Jay Newman
flavicon
face
> Colossus: The Forbin Project

Yes. I also read the books. The movie was "interesting" by today's
standards of computers.

But then, movies rarely portray computers accurately.
--
D. Jay Newman           ! DCX - it takes off and lands base down,
RemoveMEjayspamspamBeGonesprucegrove.com     !       as God and Robert Heinlein intended.
http://enerd.ws/robots/ !

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spamBeGonelistserv@spam@spamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body

2004\08\12@131023 by Howard Winter

face
flavicon
picon face
On Thu, 12 Aug 2004 08:30:09 -0400, D. Jay Newman wrote:

> movies rarely portray computers accurately.

I think  you're being kind - I can't remember *ever*
seeing computers portrayed accurately on TV or in films.

Except "Deep Thought", in HHGTTG, where it's awkward,
pedantic and annoying...

Cheers,


Howard Winter
St.Albans, England

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email TakeThisOuTlistservspamspammitvma.mit.edu with SET PICList DIGEST in the body

2004\08\12@163601 by Peter L. Peres

picon face
>> What's your view re a self-aware machine that could make human-like
>> decisions ?
>
> Wouldn't that just replace the bad decisions made by humans with equally
> bad descisions made by machines?

I think it would replace decisions which humans delusionally believe to be
infallible and well motivated with decisions which even humans can see to
be less infallible and less well motivated. This on the grounds that both
the humans and the machine will be imperfect by design and if the machine
carries over the human's ideas and intentions then it will be less good
than its creator (perfection degree < 1 for both, perfection degree of
machine is then x * y, which is less than both x and y if x,y < 1. Worse,
if a group of N individuals is coopted to perfect the machine, the joint
contribution will cause an even more imperfect result ...). The non-$DEITY
involving proof of imperfection exists in many forms, one of them being
the one involving the flat (2D) people who can't tell the shape of an egg
that is intersected by their flat world.

So then one could invest a machine with initiative and other human-like
impulses, like the impulse to kick an empty can just for the hell of it.
Then you would have a 200lb robot kicking someone nearby just for the hell
of it pretty soon.

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservEraseMEspammitvma.mit.edu with SET PICList DIGEST in the body

2004\08\12@164846 by Dave VanHorn

flavicon
face
>
>So then one could invest a machine with initiative and other human-like
>impulses, like the impulse to kick an empty can just for the hell of it.
>Then you would have a 200lb robot kicking someone nearby just for the hell
>of it pretty soon.

"No, it's because I like the squishy sound they make!"

:)

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservEraseMEspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body

2004\08\16@082704 by Alan B. Pearce

face picon face
>Googles  ... Atlas .... 72 consecutive launches as at May 19 2004

The Delta rocket is also pretty reliable these days - having just had my
part of a project launched on one a month ago.

Mind you it did make everyone laugh on the previous day when they had to
abort the launch sequence with about 3 minutes to go, and the launch
controller intoned "now everyone please refer to page x in appendix a ..."
sounding more like a priest at a high church service than anything else.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

More... (looser matching)
- Last day of these posts
- In 2004 , 2005 only
- Today
- New search...