<HTML><BODY style="word-wrap: break-word; -khtml-nbsp-mode: space; -khtml-line-break: after-white-space; "><BR><DIV><DIV>On May 18, 2006, at 10:39 PM, jeffrey davis wrote:</DIV><BR class="Apple-interchange-newline"><BLOCKQUOTE type="cite"><BR><BR> <DIV><SPAN class="gmail_quote">On 5/18/06, <B class="gmail_sendername">Samantha Atkins</B> <<A onclick="return top.js.OpenExtLink(window,event,this)" href="mailto:sjatkins@mac.com" target="_blank">sjatkins@mac.com</A> > wrote:</SPAN> <BLOCKQUOTE class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid"><BR><BR>It is not irrelevant as the type of technology available determines the costs of such a project and its feasibility. Sufficiently advanced automation to accomplish this task as well as sufficient resources and sustaining technology may require nanotechnology and AI. I think that it will. If you think otherwise then please make your case. <BR><BR>- samantha</BLOCKQUOTE> <DIV> </DIV> <DIV>You're perhaps familiar with "Advanced automation for space missions"(AASM), a seminal work on self-replication by Freitas et al. sometimes referred to as the 1980 NASA summer study. A quarter century ago Freitas declared self-replication doable, and on the moon no less, with the attendant severe restrictions on human on-site assistance. So it's not really my case but Freitas's. <BR></DIV></DIV></BLOCKQUOTE><DIV><BR class="khtml-block-placeholder"></DIV><DIV><BR class="khtml-block-placeholder"></DIV>Assuming sufficient resources, energy, control and logics that can't be locally replicated without something like MNT, yes? </DIV><DIV><BR><BLOCKQUOTE type="cite"><DIV><DIV> </DIV> <DIV> </DIV> <DIV>Engineering-wise, it's about control systems. Our current industrial system with humans in the loop has 100 percent closure. Replacing the humans requires control systems.</DIV></DIV></BLOCKQUOTE><DIV><BR class="khtml-block-placeholder"></DIV>How sophisticated are these likely to need to be to build infrastructure capable of supporting large numbers of humans in a hostile environment. Where is the case that we have that sophistication remotely in hand or will have with less than major AI advancements?</DIV><DIV><BR><BLOCKQUOTE type="cite"><DIV> <DIV> </DIV> <DIV>Since 1980 we've seen how many doublings of computational capacity, which translates into vastly cheaper (and/or vastly more capable) control systems components? Using Moore's law as a rough guide, in the twenty-five years since AASM, control element costs have fallen, or capability risen, by a factor of 10e6. </DIV> <DIV><BR></DIV></DIV></BLOCKQUOTE><DIV><BR class="khtml-block-placeholder"></DIV><DIV>Total hand waving. Chip density and raw speed to not remotely directly translate to increased autonomous control capabilities.</DIV><BR><BLOCKQUOTE type="cite"><DIV><DIV> </DIV> <DIV>Beyond that, the obstacles to implementation remain vision, creativity, the size of the project(very big), and perhaps political will. Personally, I prefer to dispense with political will and go with vision and creativity. </DIV> <DIV> </DIV> <DIV>That said, many smart folks still contend that the problem is "too hard". Add the daunting size of the undertaking and it becomes a non-trivial matter to mobilize enough folks to "Just give it a try and well see if it can be done." That's where the creativity comes in. <BR></DIV></DIV></BLOCKQUOTE><DIV><BR class="khtml-block-placeholder"></DIV><DIV><BR class="khtml-block-placeholder"></DIV>Thanks for for the rah-rah non-answer.</DIV><BR class="khtml-block-placeholder"><DIV>- samantha</DIV><DIV><BR class="khtml-block-placeholder"></DIV></BODY></HTML>