Originally humans were motivated by a fear of pain (starvation, injury, death). These gave us the knowledge of things to avoid as individuals. However this knowledge did not prevent large-scale pain, such as famine, plague, war, etc. We collectively began to invent things to mitigate these problems as well, such as agriculture, stone masonry & construction, weapons, government. The rest is history. We have become so good at survival that there is almost nothing that can hurt us. We grow and multiply freely, our numbers increasing without bound. These days we have so much time on our hands that we have forgotten, or at least have time to ask, what it is that makes us "human".

No matter how fast a computer is, if it cannot sense its surroundings it will never be human. If it does not have a body and able to feel sensation, hear, see, smell, and taste at least in a similar way to humans, it will always be handicapped. It must be able to do this for extended periods, perhaps a decade or two, to accumulate a basic understanding of the world. Simulations may give a computer experience, it could learn from 3D environments, as long as they are constructed of enormously detailed high resolution textures. At some point however, it will have to move around in the real world. It must also be motivated by pain. There must be sufficient resources available for the computer to repair itself, and physical self-repair must be possible. Either that or there must be groups of computers with the ability to interact and support each others total survival needs.

Perhaps robotics engineers should develop artificial skin so robots can "feel" temperature & pressure; stereoscopic eyes for real 3D vision; sniffer tubes; chemical tasters for sugar concentration, acid concentration, base concentration, dryness or wetness. Such components require highly developed technology, and it is hard to see how a computer could repair itself without assistance.

I think sentient computers would be highly interdependent, for redundancy, for survival.