[ExI] Meat v. Machine (was simulation)

Eugen Leitl eugen at leitl.org
Wed Dec 29 09:26:43 UTC 2010

On Tue, Dec 28, 2010 at 08:50:35PM -0800, The Avantguardian wrote:

> > And if you don't like it, you don't have to do it.
> Imagine you are on a desert island with Dr. Jeckyll and dozens of innocent 
> bystanders. Dr. Jeckyll offers to share an elixer with you that he strongly 
> believes will transform anyone who quaffs it into a hirsute, immensely strong, 
> and violently homicidal brute. He tells you that he will almost certainly drink 
> it but, "if you don't like it, you don't have to do it." What would you do?

Well, if you really believe that, you should alert the authorities,
or walk the same path Unabomber did.
It's pretty obvious that the military would be extremely interested
in using artificial and/or biologically inspired control on the battlefield,
maintain leadership (hence arms race) in such capabilities, and there's
plenty of lunatic-fringe research about soldier augmentation to realize
how the bacon fries.
> > The information pattern between your ears is also pretty volatile.
> But the information between my ears is not directly coupled to the information 
> in my physical makeup. I can't think myself shorter, taller, or into a 
> worm. Uploads would enjoy no such disconnect between implicit and explicit 
> information. Some seem to think it may be an advantage, but I think it 
> could pose serious risks for the uploaded individual. What happens if an upload 
> thinks himself into a philosophical zombie, computer virus, or other entity not 
> able to think? Or worse poses a danger to the rest of the world? The danger of a 
> stray thought being capable of ending ones existence might give pause to 
> some. And yes you could point out that suicide is possible in meat body too, 
> however, more than the mere passing thought of suicide is required to kill 
> oneself in real time. 

Strawman. You're describing a particular pathology which could be
implemented, in theory, and declare it's the norm. No, it would be a 
nontrivial problem to build a particular pathology like that, and 
of course the answer to that is -- if you don't like it, don't do it.
> > > > Only, postbiology is a lot fitter than you.
> > > 
> > > What is your evidence for this assumption? Or is it an article of faith? Do 
> >you 
> >
> > 
> > Functionality concentration per amount of volume.
> Sigh. Uploaders seem to constantly underestimate how incredible meat is from a 

If it helps, I'm not entirely ignorant of biochemistry. 

> design standpoint. Some of the simplest meat out there, E. coli bacteria, puts 
> engineered technology to shame. The following is something akin to the 
> engineering specs for E. coli:
> http://redpoll.pharmacy.ualberta.ca/CCDB/cgi-bin/STAT_NEW.cgi
> Functionality per unit volume is difficult to measure because functionality is 
> not well defined, but here is a back of the envelope calculation based on the 

We can look at information processing per unit volume/unit energy, information 
storage density, power density, the concentration of particular widgets/volume, 
and so on. 

See e.g. http://nanoengineer-1.com/content/index.php?option=com_content&task=view&id=39&Itemid=49

Now think how you would build that enzymatically,
in a wet context.

> assumption that distinct genes represent distinct functions: 4377 genes 
> per bacterium divided 
> by a cell volume of 10^-15 liters, gives functionality per unit volume of a 
> bacterium as 4.377 *trillion* functions per cubic millimeter. What technology 

Machine-phase nanotechnology.

> does that? Computers may push numbers around with similar functional density, 

3d integrated nanoelectronics are not merely similar. 

> but a bacterium is pushing atoms around, cutting them apart and welding them 

Of course pushing the atoms around is the whole point of machine-phase
nanotechnology. If it can't self-reproduce and build other structures,
it will never amount to much.

> together. Sure people can imagine technology achieving this kind of functional 
> density someday, but I certainly wouldn't take it for granted.   

Then the problem disappears into thin air, and you can sleep safely.
> >  Repertoire of accessible
> > structures.
> Frankly, I don't know how to assess this other than to say that the diversity of 
> biological structures is vast.

If you representatively sample the space of all possible chemical
structures (it's a computationally tractable problem) 
then these points and regions occupied by biology are negligible.
> > Operation temperature range.
> Admittedly I'd score this one for the machines.

A range of about 800 K, or thereabouts.
> > Facultative volatile use.
> How about Alcanivorax borkumensis? As its name suggests, it eats alkanes like 
> oil and gasoline. 

Plunk it down on Phobos, and see how well it fares.

> > Fully
> > static design.
> Coral reefs? I am not sure why static design would be an advantage. "You're a 
> steam shovel and that's all you'll ever amount too!"

Static in the energy sense. Imagine shutting down your CNS completely,
not even consuming nW, and resume it a decade down the road. A lot
of baseline metabolism is just for homeostasis. A fully static
spintronic design doesn't need energy to hold state. When sufficiently
degraded, the structure can be rebuild by an external agency, retracted
completely so only computationally relevant elements remain.

If you don't like that, you can of course implement the current paradigm
where each structure contains according functionality to maintain it 

You're not forced to pick a particular path. You can even reconfigure
at runtime, if you're kinky that way.
> > Energy efficiency.
> Meat has an energy efficiency of 68 percent as long as it has oxygen to breath. 

Information theoretic computation limits.

> That's really close to the *theoretical* efficiency of the ideal Carnot engine 

Electrochemical energy sources are not Carnot-limited.

> and as efficient as a really good rocket engine in practice. 
> http://www.ece.umd.edu/~nsw/ench482/energy.htm
> > Separation of fabrication and operation.
> Again, I am not sure why this is an advantage? Would maintenance and replacement 
> of worn parts constitute fabrication or operation? How about sex?

In separate units. Facultatively, mark.
> > Ability to metabolize the entire PSE.
> Solving your all problems by eating them? Ok so grey goo has this advantage over 
> meat although I hesitate to call it an advantage. But who wants to be grey 
> goo?   

Not grey go, ability to build structures not just limited to CHNOPS, and
a few trace elements.
> > I could go on for pages, but you're probably seeing where I'm getting.
> > Yes, they're way meaner and leaner than us.
> Actually I hoping for more substantive evidence and less speculative hand waving 

Go here http://www.rfreitas.com/
> for your assertion that postbiology is superior to biology. You ever seen what 
> happens to an automobile after a few years in a rainforest? It gets reclaimed by 

And your point is...?
> nature. I saw one a few years ago. I wish I had taken pictures of it. 
> > > mean more fit than me in the vacuum of space or in the jungle? In a virtual 
> > 
> > Everywhere. And jungle is crunchy, and good with ketchup.
> Not if the fire ants have any say in the matter and they very well could.
> http://www.youtube.com/watch?v=3vVUh-194vU

What do ants do when it starts raining liquid air?
> > The nice thing is that you can halt state, copy over, and resume.
> > No such options for us.
> halt state := freeze, 

Please freeze your head within this very ps, without dropping 
a single spike.

> copy over := clone, 

How do you clone all the structures within of your frozen head down to the nm?

> resume := thaw

Please unfreeze your mythical nm-scale copy within this very ps,
without dropping bits that you don't even notice what happened.
> Albeit it only works for cells and tissues right now, give it time.

Well, yeah, I've worked at 21cm and CCR. 
> > I use the term exactly because machine-phase is like biology, only more
> > so, and is successor to current biology (both can't co-exist, due to
> > fitness delta and incompatbility (they can eat you, you can't eat them)).
> There are bacteria that eat iron so why would you assume they couldn't compete 

Your logic is interesting.

> with machine-phase biology? You think the 
> conclusion is foregone, I think it would be quite the contest perhaps the 
> main event of 
> a geologic epoch-- liquid-crystal life versus solid-state life. If machine-phase 

I don't know how long prebiotic Earth and first autocatalytic sets coexisted,
but it must have been very quick. Where are they now? The atoms are still (mostly)
there, but the reearangement is completely different.

> life is so inevitable and so 
> superior, where are the Von Neumann probes?

They're not "von Neumann machines". They're people.

The answer to Fermi's paradox is that we're not in 
anybody's smart light cone. Probably. There are
alternative explanations, but most of them a lot
more complicated. 
> > But they were not aware of what they're doing. We are, yet we're still doing 
> >it.
> The universe is notoriously indifferent to ones mental state. It only 
> evaluates actions and not the rationalizations behind them. To the gazelle being 
> stalked by the lion, the consequences of indifference and unawareness are the 
> same.  

I wasn't referring to the universe's view (it has none).
> > It's less about perks like living forever, it's more about transcending the
> > limitations of being a bipedal primate. It tends to cramp your style a bit.
> Solid-state organisms would have limitations too so you would simply be 
> trading one set of limitations for another.

Of course they have limitations, you're subject to the laws of this
spacetime. But their limitations are a lot less than us, and that's
the whole point of venturing beyond biology.
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE

More information about the extropy-chat mailing list