Re: Linda Nagata's Attriums

Anders Sandberg (
Tue, 13 Aug 1996 13:35:05 +0200 (MET DST)

On Mon, 12 Aug 1996, Crosby_M wrote:

> If you haven't read Linda Nagata's "The Bohr Maker" (Bantam Spectra, Apr'95,
> ISBN 0-553-56925-2), it's well worth reading even if you normally have
> little time for fiction. Her story (with well-developed characters & drama
> for a first-time novel) involves a near-future solar system based on
> nanotechnology, 'attrium' implants for full immersion into VR and the
> transmission of mind copies between distributed attriums.

I also recommend the book, I have no doubt many extropians will enjoy it.
By the way, it is "atrium", not "attrium" (probably from the open room at
the center of roman villas where visitors were received).

> Plus, there's a
> libertarian struggle against a 'Gaian' World government that exists for the
> primary purpose of avoiding a 'gray-goo booboo' AND proscribing transhuman
> technologies.

It is a bit more subtle than that. I think the Commonwealth represents
just the kind of state (or in this case, international organization) that
extropians fear and want to fight again (guess why the book is so fun
reading then). But it isn't naive - there is a real fear, and a real
danger, so public opinion generally supports it (_Tech heaven_ give some
hints of its beginnings). All transhuman technologies are not proscribed,
just the too drastic (independent AI, uncontrolled assemblers, multiple
physical copies of the same person).

> While it may be feasible that such a device itself could eventually be built
> and implanted, a few practical interim questions are (a) how would an
> 'attrium' interact with the organic human brain and nervous system;

As I see it, Nagata's idea is that the atrium itself is a nanocomputer
network that send out interface threads to neurons in the owner's brain.
For example, it might be located above the cortical surface, sending down
bundles of nanofibres into the sensory and motoric pathways so that it
can override/augment the outside signal (this could be done by inhibiting
the primary sensory imput and sending in new input; a quite fun
neuroelectronic problem). With nanotechnology this is probably quite
simple, the hard part is making sense of the signals (in the I/O parts of
the brain there is probably less variation than the cognitive systems).

Creating ghosts seem to imply that the nanofibers can read *all* of the
brain, and transmit the uploaded pattern. This is much harder and less

> how could the necessary wide-bandwidth and far-ranging communications be
> implemented, especially given the current worries about the effects on
> organic functions of ELF and more high-powered electromagnetic energies?
> (When even cellular phones are suspected of having some impact on brain
> tumors in certain genetically susceptible individuals, it seems unlikely
> that a radio transponder could be implanted in human brains without serious
> biological side effects.)

What about a body area network a la those under development at MIT?
Signals could also be sent to a transponder (or several) outside the head
area, and medical nanites could keep the surrounding tissues healthy.
Although I would probably opt for a infrared system, which interacted
with the surrounding local networks (but in _The Bohr Maker_ the system
seems to be able to function by radio).

> As a communication alternative, there is some speculation in the quantum
> physics community over the potential for non-local communication, i.e., once
> they have interacted, two quantum particles remain immediately correlated
> regardless of their distance from each other. Some progress in quantum
> computing based on these principles has been made.

Quantum entanglement isn't useful as an information carrier (I think this
has been proven quite rigorously), but it is useful in cryptography and
quantum computation. If I have one of a pair of correlated particles and
you the other, any change I do to the particle will destroy the
correlation and does not cause any detectable change to your particle (it
is the initial states that are correlated).

> Perhaps it is for these reasons that Linda Nagata envisioned fully-immersive
> communication between attriums taking place only via the downloading of
> 'ghosts', extracts of a person's sensibilities, memories & mind, able to
> transmit to another's attrium, reside there 'off line' from the originator's
> mind, and able to return the experiences to the originator only by
> disengaging and uploading back to the 'home address'.

This seems to be reasonable over interplanetary distances too (even on
the same planet; the delays become comparable to neural delays on a scale
of just thousands of kilometers).

Interesting idea: "squatter ghosts" and other infomorphs that move
between the atria of people, using the empty buffer space used to
receiving ghosts.

> 07/20/96. ATTRIUM DESIGN LEVELS. While Nagata doesn't go into any of the
> following, I think there would likely be several stages in the development
> of such attriums:
> A level 1 Attrium would basically be a thought & experience recorder: able
> to replay sensations and subvocalized thoughts; which also implies a
> capability to generate something like a 'hypnotic state' for controlled
> overriding of sensations from ones' immediate physical environment during
> playback.

Sounds almost like an enchanced reality version of the current prototype
wearable computers, with an useful user interface.

> Key areas to be explored would include
> agents that could simultaneously monitor 'mood' from within the
> hormone-filled ventricles of the brain

Just a small correction: the hormones in the ventricles are rather bad
predictors of mood (especially since most hormones are released into the
blood) - I suggest that you look at the activity in the limbic system

> A practical goal would be a system for handicapped people
> that provides a simulated keyboard in their minds that could be used to
> capture mentally spelled-out sentences.

There have been some success in interpreting arm movements in monkeys due
to cortical activity patterns (Schwartz et al, see SciAm a few months
back), so this may be feasible.

Anders Sandberg Towards Ascension!
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y