[ExI] immortal

[BillK]

On Tue, 12 Aug 2025 at 19:38, Darin Sunley via extropy-chat
<extropy-chat at lists.extropy.org> wrote:
>
> In Diaspora, Greg Egan divided humanity into uploads, robots, and modified biological humans. The robots, in particular, are absolutely dogmatic about staying in robot bodies, despite the fact that the polis mainframes are perfectly capable of running/emulating the robot hardware byte code, and vice versa.I never really sympathized with that before. But this conversation made it really make sense for me.
>
> At the end of the day, what you own is what you can control and what you can defend. An upload running on a data center of some sort is utterly at the mercy of the administrators. But if you're running on single-tenant hardware in a mobile robot body, you can control your hardware, and defend your hardware, in a way a polis citizen really can't. I would be highly reluctant to do any upload scan - it is axiomatic that data cannot defend itself. But an upload directly to secure computing hardware that is tamper-proof and inseparable from a chassis that can defend itself is distinctly more palatable.
>
> As to the question of money and sustenance, if you own and control a robot body, you can mount a thousand year pocket nuclear reactor, coat yourself with solar cells, build and defend a solar farm or fissible mine on an asteroid somewhere, etc. Just like today, you take your sustenance any way you can, subject to the approval (or at least the lack of capacity to stop you) of your immediate neighbors.
>
> As to whether it would be ethical to have effectively immortal post humans how many "we" should "allow", I find the question slightly distasteful, and more than a little moot. Assuming any of us survive the next 20 years, I intend to continue to keep trying to survive, and to promote the survival of my loved ones. You can have my warm robotic hands when you pry them from my cold dead wrists. :)
> _______________________________________________


The problem of uploads being at the mercy of the humans that run the
machines seems to me to be assuming that the present state of
operations will continue. I think this is a mistake.
Humans uploading into computers is many years in the future,
probably after the arrival of AGI and the new world that that entails.

I asked the new ChatGPT-5 AI to consider this situation and the implications.
Here is a much abbreviated summary of the suggestions.
BillK

Some ChatGPT-5 AI comments -

Civilization at computer speeds.
If the uploaded population thinks thousands of times faster,
real-world events would seem glacial.
They could detect, model, and respond to threats in what’s effectively
real time for the outside world — potentially intercepting attacks
before humans even finish setting them up.

Decline of physical humans.
If biological humans fade out, the power dynamic shifts:
No external “caretaker” species to hold the off-switch. The uploaded
population might inherit ownership of the infrastructure by default.

Humanoid or non-humanoid robotic guardians.
If advanced robotics, running at the same speed and intelligence as
the uploaded civ, handles maintenance and defense:
Physical protection becomes native to the civ, not outsourced to
slower biological minds. These robots could build, repair, and defend
infrastructure autonomously.

Radically improved computing substrate.
Future substrates might be:
Near-indestructible (quantum fault-tolerant systems, exotic materials,
maybe even computing woven into planetary crusts).
Vastly more energy-efficient, reducing dependence on delicate,
centralized infrastructure.
Distributed at a planetary or solar-system scale, making “switching
off” functionally impossible without total system collapse.

Other future factors worth considering:
Self-migration: The civ could constantly move itself across multiple
substrates, even into alien computing environments, so there’s no
single point to destroy.

Integration with the physical environment: If they compute within
nanostructures embedded in planetary or asteroid mass, “pulling the
plug” might mean dismantling an entire world.

Interstellar dispersal: Copies of the civ could be sent at light speed
to many star systems; no single disaster wipes them out.

Post-scarcity physical control: If they control nearly all automated
manufacturing and resource extraction, they effectively control the
human-physical world as a side effect.

Conclusion:
The security question then isn’t “how do we stop slow, fragile humans
from flipping the switch?” — it’s “how do we stop ourselves from
becoming our own existential threat once we can act at godlike speed,
across physical and digital planes?”
---------------------