[FoRK] Making robotics a priority

Stephen Williams sdw at lig.net
Thu Jul 14 11:19:20 PDT 2011

On 7/14/11 10:55 AM, Aaron Burt wrote:
> On Wed, Jul 13, 2011 at 12:57:29PM +0200, Eugen Leitl wrote:
>> On Tue, Jul 12, 2011 at 08:11:08PM -0700, Stephen Williams wrote:
>>> On 7/12/11 7:30 PM, Adam L Beberg wrote:
>>>> And you propose we get to post-scarcity though peak-oil,
>>>> peak-phosphorous, peak-water, peak-fiat, lack of education, 60/40
>>>> gender imbalances, and rapid destruction of the planet... how exactly?
>>> More robotics / AI / automation.  Didn't I make that clear?
>> Had you said nanotechnology and superintelligent AI I'd agree.
> I'm trying not to read that as "resource-intensive pinnacle-technology-
> of-the-moment #1 and #2."

Nanotechnology and AI is resource intensive?

>> However, we're not getting these anytime soon. Because almost nobody
>> is working on it.
> Nanotech is getting a *lot* of work.  But we're still at the materials
> stage, not the moving-parts stage.  Think iron age, not industrial age.

You haven't been paying attention.  While slow going, we've had molecular moving parts for years.  There were several versions 
of a molecular memory that used a rotating ring molecule around a bar-like molecule to denote 1 or 0.

We're now building shapes and materials at the protein level, and often by proteins or protein-like mechanisms.  Sometimes these 
are for biological purposes, sometimes not at all.

After getting extremely good at imaging individual atoms and small and large molecules, we're now very good at pushing and 
placing individual atoms.  A good chunk of the supercomputing capacity right now is working on protein folding to both 
understand and engineer molecular shapes and mechanisms.

> AI is also getting a lot of work.  But it's not sci-fi brain-in-a-chip AI,
> it's Google "teach computers to do what we mean" AI.  What's your take on
> Charlie Stross' http://www.orbitbooks.net/2011/07/08/artificial-stupids/ ?

He's wrong about flying cars, they're shipping next year.  He's also wrong about flying being necessarily hard: there are dozens 
(at least) of UAVs flying that are completely automatic.  Most airliners land themselves now.  Pilots at some airlines are 
forbidden to land manually in bad weather.  How is that somehow impossible for a flying car?  In real production models, we'll 
have multiple redundant mechanisms (parachutes, airbags, maybe even last-resort retro rockets), but it isn't impossible.

The rest about AI seems silly and myopic.
>> Robotics is about the only thing to make scalable organic
>> agriculture happen, which at least postpones the malthusian
>> reckoning.
> Hm.  Spend mineral and energy resources on farming so humans don't have to
> do the work?  Sounds a whole lot like our current fossil-based agriculture.

Take the fossil out.  Spend a bit on equipment that is long-lived and self-repairing perhaps, but probably much less on energy.

> You may have some misconceptions about the efficacy of sustainable and
> organic agriculture.  It ain't automated like industrial grain production,
> but then, very little can be automated to that degree.

It could be automated.  Avoiding all advances in commercial agriculture would be a mistake, but cutting down pesticides is nice.

> OTOH, while we can meet nitrogen needs w/o synthetics, phosphorus and
> micronutrients can be problematic at feed-everyone scales.  Organic farmers
> use semi-renewable resources like guano and various sorts of bone meal, but
> I haven't calculated whether the supply would meet the demand if everyone
> ate a sustainable diet.

If we need resources, we'll find a way to get them.  Sea water, sea floor sludge, rocks, whatever.

>> Key factor is renewable energy and synthons, which
>> is too little and too late.
> If by synthons you mean chemical feedstock, renewable-sourced feedstocks
> are ramping up very nicely, thank you.
>> A softer depopulation option than die-off by degraded ecosystem
>> (assisted by a nuclear exchange) would be engineered plagues.
> Why engineered?  Usually the natural ones arise in stressed, crowded
> populations.  Civil wars are an even softer option.

We just need to stop having too many children.  Attrition happens fast enough for everything but 


More information about the FoRK mailing list