Here is a draft proposal of ten levels of complexity that might be necessary for WBE: Rather than groups certain attributes into a category called 'exotica', (which suggests that the level of complexity necessary for WBE is understood but that some fruitcakes like to include fanciful extra components for their own amusement,) I suggest the levels of complexity be given in a straight-forward way to indicate that the correct level of detail for WBE is truly unknown. If WBEs became possible, then maybe it would be fun to make a virtual environment for them and give them virtual bodies, but is this really a requirement for achieving WBE? Perhaps real world visual and audio input, and real audio output is sufficient for achieving a HAL-like WBE. Why not just connect a CCD camera to the WBE's retina? This solution connects the WBE to the real world and doesn't require the additional computational bandwidth for photorealistic scene simulation.

In the computer requirements part, environment simulation is described for photorealistic visual reproduction. How else would you know whether the WBE had consciousness?

The WBE needs to interact with the real world to demonstrate that it is working. It seems to me that if you built a WBE contained within a simulated environment you run the risk of creating something completely vacuous. I don't understand why an environment simulator is a good idea, in general, or why it should be included in this WBE roadmap.

Nfs most wanted pc portable 1 link