[ExI] Silence in the sky-but why?

[Tomasz Rola]

On Sun, 22 Sep 2013, Eugen Leitl wrote:

> > without the ability to develop or even maintain current technology.  
> > Then we
> 
> We already been unable to maintain our current technology for several
> decades. The regression is not uniform, and sufficiently slow so that
> it can be ignored by many.

I'm not sure what exactly you mean by this (example[s] would help) but I 
am willing to believe it, to some extent at least.

> > would be unable to advance beyond what I have envisioned as an AD1000-ish
> > level of technology for lack of easily-available energy.  Even at current
> 
> We have no idea what technology level is sustainable if we have allowed
> us to run out of energy.

Most probably, some mix of 17 and 18 century (in high tech level places), 
with occasional inclusion of hand-make electronics. For the record, 
internet is not supportable at all. Mostly, because it may be 
hard/impossible to recreate even 60-ties level of computing. A lot of 
current production processes is automation - the hardware may last, say, 
20-30 years if properly maintained, but some failures and parts 
cannibalising are inevitable. Once education fails, maintainance fails 
too. Actually, I'd speculate this is already happening, with manuals from 
A to Z followed by expendable staff, which otherwise does not have an idea 
what goes on under the hood and have hard time thinking outside the book. 
Since we are talking a lot about energy shortages and such, it is perhaps 
worth noting that to educate an engineer from the grounds up it requires 
about 15+ years in relatively stable environment. Okay, maybe it's 
possible to do this in 10 years if children are educated with program 
designed for this one purpose. No engineers, no industrial production. Not 
enough food -> brain development issues -> retarded engineers.

I'm afraid any kind of survival plan that does not include 
sustaining/recreation of engineering is in fact a plan to slowly ride into 
dark ages and then even darker. There was a time in Europe, from what I 
have heard, when people knew Greek and had access to ancient math 
treatises but simply could not make anything out of it. Without certain 
type of education, math above certain level starts to sound like pure 
gobbledygooke.

In light of this, surviving whatever catastrophe may await us seems 
relatively easy.

Exercises for the impatient:

1. Create your own USB host from handful of parts: usb host connector (I 
am always a bit lost, whats its name, female A?), few wires, programmable 
microcontroller like 8051 or msp430 (I assume you have versions which can 
be interfaced with the said wires). This includes writing your own USB 
stack in assembler. I tried to locate such project but google failed on 
me. Thus, it's possible to have tons of USB based gizmos and nothing to 
connect them to.

2. Create CPU that does not suck (i.e. can support at least LISP 1.5 on 
it) from whatever parts that are not a cpu themselves. FPGA and similar 
stuff is not allowed. 36-bit words are very much welcome, but enough 8-bit 
words should do, too. Just not less than ten kilo of them. Because LISP 
has some requirements and so do nontrivial programs, even in Forth. This 
is doable but my impression is, only handful of people can and maybe few 
handfuls could, given enough time and books. The more time and no missing 
links in books, the better the results of course - in ten years frame, 
maybe even five, who knows, a moron like me could do, too.

3. Try to program Arduino with punched cards. This means, create a reader 
and maybe a puncher, too. Use as many Arduinos as needed. Of course you 
need to devise some card production process, too. Done? Now build a 
sorter. Congrats, you are now somewhere in 1850-1940 period, computing 
wise. It can be sustained provided enough supply of Arduinos and whatever 
other parts are needed.

4. Make a kilobit of your own core memory. Plug it to Arduino. Or whatever 
you please. Repeat 36 times to have one kiloword of core. Repeat last 
sentence 1024 times if you'd like to sport one megaword in your shiny new
supercomputer. A chassis will have to be made of reinforced concrete and 
it'll look very much like a building, but a megaword is worth it.

5. Build your own alfanumeric printer, connect it to CPU and some core 
mem (a 36-bit kiloword may do), add card reader and run a program printing 
logarithmic tables. You have made it to the 1950-ties. Have a big drink. 
Or a bottle. When you wake up and headache is gone, try to build a hard 
drive.

However:

6. Can you make an electric battery delivering certain voltage without 
looking up in google? Can you do so without multimeter? BTW, can you build 
analog multimeter? Can you build a transistor? Integrated circuit? 
Speaking of logarithms - can you build a log scale or Napier bones, or 
hand compute logarithms on paper with Taylor series? Out of your head?

I assume all of the above can be done (even though I cannot any of the 
above), but it is not going to be easy without google and with eventual 
removal of all tech made after certain date. Doing whatever sophisticated 
production without a way of building computing hardware from a scratch - 
no, I don't think so. Some of it can be done with analog regulators, which 
have to be built, too. If "PID controller" rings any bells in a head, then 
maybe. There will be some junk but its quality will degrade quickly, I'm 
afraid. Besides, there will be plenty of mp3 players, digital cameras and 
cellphones, but not so much ARM capable compilers. PC requires huge 
voltages and amperages. Laptops - yeah, maybe, with car adapter. 
Unfortunately, military tech can no longer be trusted for rock solidness, 
after they introduced Windows to their systems (long ago) and started 
replacing Ada with C++ (not so long ago, AFAIK). Mainframes could be best 
bet but they may require unusual power options (like, three phase - I 
really don't know, do they?).

Then again, how many laptops have compiler installed? And the number of 
people who can enter hex in command prompt from their memory is limited 
and it's not sure any of them will make it to times when their services 
are needed. It may be easier to find just a still working Commodore-64 
with its builtin BASIC.

So this junk is going to be very much useless outside the context of 
current civilisation. Most probably it'll be used for jewelry making, to 
create ritual bonding with better, golden times.

This can be called disasterbation. Sure, why not. However, opposite 
thinking can be as well called triumpherbation, unless it can be logically 
proven that the global fail is avoidable [1] and massive life loss is not 
going to happen. Just in case, I'm not lamenting anything. I'm only trying 
to assess certain paths.

[1] If using word "thorium", explain how to build enough plants [2] every 
year - those cannot be 3d printed. Even more important, I'd like to see 
one thorium plant in operation, delivering to the grid, this year. Last 
time I checked, there was none but maybe I missed something. Experiments 
being ran 30 years from now and shut down 20 years ago do not count. If 
using "solar space" please explain effects of solar flares and magnetic 
storms on such orbital station, including rare but possible 
Carrington-like events. When diminishing threats, include Fukushima 
keyword. I assume wind and ground solar are not available in enough 
quantity and/or can't be made fast enough to change anything [3]. But, 
again, I would like to be wrong, so surprise me, anybody. But I'd like to 
read about existing, available right now. Not some lab experimental 
startupmental. Hydrogen based tech? Nobody here mentions it, is it really 
so bonked? I heard this could be used for almost drop in replacing of oil 
fuels (I understand, difficulty is on par with converting normal car to 
gas). Of course it requires plenty of H.

[2] BTW, even existing plants may be in trouble, even if otherwise in fine 
condition, once there is no more folks capable of, say, PDP-11 
programming. Because there is some number of such old computers still used 
in nuke plants or in NASA and, well, despite all the progress and what 
not, spares have to be hunted on ebay and people have to be unmothballed 
from retirement. AFAIK.

[3] Ground solar and wind are great option for individuals and not too big 
groups, IMHO. Maybe for some small countries, too - in a good place. As of 
countrywide solar+grid - I would be delighted to see this, really. I am 
still educating myself about this so, nothing interesting to say.

Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature.      **
** As the answer, master did "rm -rif" on the programmer's home    **
** directory. And then the C programmer became enlightened...      **
**                                                                 **
** Tomasz Rola          mailto:tomasz_rola at bigfoot.com             **