[ExI] Power satellites

Keith Henson hkeithhenson at gmail.com
Sat Apr 25 23:07:07 UTC 2009

On Sat, Apr 25, 2009 at 9:29 AM, BillK <pharos at gmail.com> wrote:


> But the main thing I don't like is the timescale of 15 to 50 years.
> How can you guarantee funding and political will for a project like that?

You can't.  Which is why the crash program to reach at least half a
GW/day construction rate by 2015.  Impossible you say?  I agree it
can't be done in the US, we are no longer the country that did the
Manhattan project in a few years.  China/Japan may be a different

> PV cells are improving month by month, becoming cheaper and more efficient.
> By 15 years time, I would expect every building to be off the grid,
> using spray on PV cells.

Short of full on nanotechnology, I don't think spray on PV cells are
possible.  Assuming they were, how are you going to make connections
to them?  What voltage/current are you talking about?  PV cells make
around half a volt/cell.  How do you suggest wiring them in
series/parallel to get a reasonable voltage?  Storage?

>And who knows what other energy improvements
> will come along in that time period.

The current energy per person in the US is ~10kW.  It is obviously
possible to run a human scale intelligence on 20 W.  There is a 500 to
one improvement.

> By the time it gets half built, people will be wondering whether they
> should be spending all this money on a project that is looking as
> though it might not be needed after all.

Possible, likely even.

>From the "Standard Gauge" story

The grid was lightly loaded though not as much as the 98.8 % drop in
active population would indicate; supporting close to half a billion
minds and keeping the associated bodies in cold storage and their
memories updated took a modest amount of power.

On the way down to the train, the power sats looked like a chain of
beads across the southern sky.  Amanda turned to Jenny:

"Jenny, if you stay up a while, you can see power sats blink out as
they enter the earth's shadow."

"Remember what I said about humans not being very good at anticipating
the future?


"Well, there you can see it."  The chain of power sats had gaps in them.

"By the time most of them were built, the population and the demand
for power was falling fast."

"There are a few up there that were never finished."

"You won't see it this time of the year, but in the spring, the power
beams are used to warm up fields and control weeds."

"How does that work?"  Jenny asked as they approached the station with
its Russian looking cupola on the top of the clock tower looming in
the near dark.

"The farmers put up pilot beams in the middle of a big circle of
farmland and several power sats focus their microwaves on the pilot
beams.  It's just like warming up food in a microwave oven."

Jenny had figured out microwave ovens before she was 3 so the concept
was easy to grasp.

"And the heat kills the weeds?"  Jenny said.

"Weeds, insects, nematodes, they mildly cook the top few inches of soil."


Thirty years before the AIs who were tasked with remembering and
making presentations to CMU visitors would run up a palace of utility
fog on the mall and present a 3D docudrama on the historical events
around the emergence of AIs at CMU.

Now, in deference to the attempt to raise children in a retro
environment with features of the 1950s, the adults were directed by
messages to their neural interfaces to the McConomy Auditorium, a 110
year-old theater in Carnegie Mellon’s central buildings.  The
20-minute presentation to the adults and older kids (the younger ones
could watch or play on the lawn) was in black and white newsreels
format, much of it converted from video of press conferences.

“Even with a nearly complete historical record from those times, it’s
hard to pin down when the first AIs became full personalities.”   The
narrator spoke in a voice over showing primitive robots and computers.

“The problem isn’t unique to AI history, there is a similar problem
about the first railroad."  (Montage of drawings and photographs of
early trains.)  "About the best we can say is that what we now think
of as AIs didn’t exist before 2032 and definitely did by 2036.  In
that year there were more than a hundred scientific papers co-authored
by AIs.   Carnegie Mellon was in the forefront of this effort."
(Shots of University labs and bits of recorded slow interactions with
early AIs.)  "The key insight was to equip AIs with carefully selected
human motivations."

“The two biggest problems of the early 21 century were energy and
medical treatment.  CMU researchers contributed to both.

“Solar power from orbit solved, in fact, over solved, the first by
2035.”  (Shots of space elevators and power satellites in orbit,
photos of rectenna farms.)

“Integrating AIs into nanomedicine clinics solved the medical
treatment problem.  It took only a few years.  After that AIs and
clinics could be “grown” at low cost and they did their own upgrades,
a lot of it in the field in Africa.  They were too late for the
smallpox epidemics that swept out of the Mid East.

“A side effect of the clinics and widespread use of virtual reality
caused a physical world population crash in the mid 2050s and the
mothballing of the cities.”  (Simulated video of dense freeway traffic
dwindling to an occasional car and then none.)

The rest of the presentation was subtle propaganda mainly to the
children and directed to the goal of them doing their part in
enlarging the population.


My work on power sats built using space elevators wasn't serious.  Hu
Davis asked me to run the energy payback analysis for rockets.  The
payback times were really good, under 100 days even for rockets, but
the economics sucked, $8-10 a gallon gasoline, at least ten cent a kWh
power.  Not to mention the boggle factor, launching a vehicle twice
the size of a Saturn V every hour.

In thrashing around I looked at nuclear tugs to get stuff from LEO up
to GEO.  Better, but not good enough.  Considered lasers to accomplish
the same thing and with Jordin Kare's suggestions got the traffic
model down to four 300 ton sub orbital launches with the second stage
being pushed by a 4 GW ablation laser.  The laser has to be bounced
from GEO to get the very long path length required for large payloads
and a relatively small laser.  Mass ratios are 3 for the first stage
and as little as two for the second.

More information about the extropy-chat mailing list