[ExI] Fermi Paradox and Transcension

[Anders Sandberg]

On 15/09/2012 17:42, BillK wrote:
> Now, if he is expecting all this in only ten years..... what will be 
> possible by the time Mbrains come along? They might not even need to 
> be near the sun at all.

Thermodynamics puts a limit to how little energy dissipative 
computations require: you need to pay kTln(2) Joule per erased bit. 
Reversible computations can do things more cheaply, but you typically 
need to slow down - and error correction still requires power.


> If they are orbiting further out, then heating and cooling might be 
> negligible. And smaller nodes will be more powerful. A DVD sized disk 
> might be a whole Mbrain civilisation.

A 3 K disk is just a hundred times more effective than a 300 K disk. 
Maybe it is more effective to use a lot of energy to cool the computers 
rather than do the computation? Hmm, I suspect not, but I am not a 
thermal wizard.


-- 
Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University