<html><body><div style="color:#000; background-color:#fff; font-family:times new roman, new york, times, serif;font-size:14pt"><div id="yiv510374316"><div><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);"><div id="yiv510374316"><div id="yiv510374316yui_3_2_0_19_131935505043256"><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);" id="yiv510374316yui_3_2_0_19_131935505043257" class="yiv510374316ms__id23203"><div id="yiv510374316"><div id="yiv510374316yui_3_2_0_17_1319355050432168"><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);" id="yiv510374316yui_3_2_0_17_1319355050432169" class="yiv510374316ms__id15766 yiv510374316ms__id23204"><div id="yiv510374316"><div
id="yiv510374316yui_3_2_0_16_1319355050432133"><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);" id="yiv510374316yui_3_2_0_16_1319355050432134" class="yiv510374316ms__id7808 yiv510374316ms__id15767 yiv510374316ms__id23205"><div><span id="yiv510374316yui_3_2_0_17_131935363901756">I wrote:</span></div><div id="yiv510374316yui_3_2_0_17_1319355050432208"><span id="yiv510374316yui_3_2_0_17_1319355050432206"><br>> One of the largest challenges in the development<br>> of AI is financial. Some ten years ago I
estimated<br>> my favored approach would take some 2 billion<br>> dollars in lab equipment and infrastructure plus<br>> the budget to keep a minimum of 50 people<br>> involved - call it 100 million per year for 10-15<br>> years. Adjust for current pricing.</span></div><div id="yiv510374316yui_3_2_0_19_131935505043273"><span id="yiv510374316yui_3_2_0_16_1319355050432202"></span> </div><div id="yiv510374316yui_3_2_0_16_1319355050432250"><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span></div><span id="yiv510374316yui_3_2_0_16_1319355050432193"></span></div><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);" class="yiv510374316ms__id7808 yiv510374316ms__id15768 yiv510374316ms__id23206"><span><div><font id="yiv510374316yui_3_2_0_19_131935505043276" size="2" face="Arial"> </font><br
id="yiv510374316yui_3_2_0_16_1319355050432176">> That estimate
seems rather pessimistic to me. What </div><div id="yiv510374316yui_3_2_0_16_1319355050432221">> is your reasoning?</div></span><div id="yiv510374316yui_3_2_0_16_1319355050432221"></div><div id="yiv510374316yui_3_2_0_19_1319355050432145"><span id="yiv510374316yui_3_2_0_16_1319355050432173">http://en.wikipedia.org/wiki/Monolithic_microwave_integrated_circuit</span></div><div> </div>
<div id="yiv510374316yui_3_2_0_16_131935505043293"><span id="yiv510374316yui_3_2_0_17_131935363901765">My experience in the costing is primarily related to having</span></div>
<div id="yiv510374316yui_3_2_0_17_1319353639017313"><span id="yiv510374316yui_3_2_0_17_1319353639017221">worked on the <span id="misspell-0" class="mark">MMIC</span> development program as well as</span></div><div><span id="yiv510374316yui_3_2_0_16_131935505043281">graduate classes in sensors, physics based information theory,</span></div><div id="yiv510374316yui_3_2_0_16_131935505043294"><span id="yiv510374316yui_3_2_0_16_131935505043286">and assorted optics. The laboratory where I worked could</span></div><div><span id="yiv510374316yui_3_2_0_16_1319355050432104">start from <span id="misspell-1" class="mark">GaAs</span> raw stock and process through to finished</span></div><div id="yiv510374316yui_3_2_0_19_131935505043264"><span id="yiv510374316yui_3_2_0_16_1319355050432111"><span id="misspell-2" class="mark">MMIC</span> devices - as well as design devices and circuits.</span></div><div><span
id="yiv510374316yui_3_2_0_19_1319355050432116"></span> </div><div id="yiv510374316yui_3_2_0_19_1319355050432119"><span id="yiv510374316yui_3_2_0_19_1319355050432121">http://en.wikipedia.org/wiki/Active_Electronically_Scanned_Array</span></div><div><span id="yiv510374316yui_3_2_0_19_1319355050432123"></span> </div><div id="yiv510374316yui_3_2_0_19_1319355050432126"><span id="yiv510374316yui_3_2_0_19_1319355050432129"><span id="misspell-3" class="mark">MMIC</span> devices are used to form phased array radars also developed</span></div><div id="yiv510374316yui_3_2_0_19_1319355050432142"><span id="yiv510374316yui_3_2_0_19_1319355050432133">at the lab where I worked. The analogy being <span id="misspell-4" class="mark">MMIC's</span> act as the</span></div><div id="yiv510374316yui_3_2_0_19_1319355050432144"><span id="yiv510374316yui_3_2_0_19_1319355050432138">neurons of the phased array radar.</span></div><div><span
id="yiv510374316yui_3_2_0_19_1319355050432162"></span> </div><div><span id="yiv510374316yui_3_2_0_19_1319355050432165">Other analogies are phased array solid state lasers and massively </span></div><div id="yiv510374316yui_3_2_0_19_1319355050432181"><span id="yiv510374316yui_3_2_0_19_1319355050432179">parallel </span><span id="yiv510374316yui_3_2_0_19_1319355050432172">computers composed of many individual serial processors</span></div><div id="yiv510374316yui_3_2_0_19_1319355050432188"><span id="yiv510374316yui_3_2_0_19_1319355050432183">- <var id="yiv510374316yui-ie-cursor"></var>also worked on where I went to school.</span></div><div id="yiv510374316yui_3_2_0_19_131935505043294"><span id="yiv510374316yui_3_2_0_16_1319355050432116"></span> </div><div id="yiv510374316yui_3_2_0_16_1319355050432228"><span id="yiv510374316yui_3_2_0_16_1319355050432120">The AI design I have in mind would require a lab with
similar</span></div><div id="yiv510374316yui_3_2_0_16_1319355050432167"><span id="yiv510374316yui_3_2_0_16_1319355050432128">capabilities plus the ability to fabricate optical devices.</span></div><div id="yiv510374316yui_3_2_0_19_1319355050432106"><span id="yiv510374316yui_3_2_0_16_1319355050432230"></span> </div><div><span id="yiv510374316yui_3_2_0_16_1319355050432233">I would consider the pricing I did to be a bargain. </span></div><span id="yiv510374316yui_3_2_0_16_1319355050432245"></span></div><div style="color: rgb(0, 0, 0); font-family: times new roman, new york, times, serif; font-size: 14pt; background-color: rgb(255, 255, 255);" class="yiv510374316ms__id7808 yiv510374316ms__id15770 yiv510374316ms__id23208"><span></span><div id="yiv510374316yui_3_2_0_16_1319355050432254"><br><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span></div><div><br>> What do you think your odds of success are?</div><div
id="yiv510374316yui_3_2_0_16_1319355050432262"> </div><div id="yiv510374316yui_3_2_0_17_1319355050432225">It is my opinion that given the funding and good</div><div>management a near certainty it could be done in the</div><div id="yiv510374316yui_3_2_0_16_1319355050432287">time frame outlined as long as government
doesn't</div><div id="yiv510374316yui_3_2_0_16_1319355050432289">interfere. It could even be accomplished within a</div><div id="yiv510374316yui_3_2_0_16_1319355050432293">government program if they returned to a management</div><div id="yiv510374316yui_3_2_0_16_1319355050432297">R&D structure more like the 1950s to early 1960's </div><div id="yiv510374316yui_3_2_0_16_1319355050432316">when choking to death on paperwork [excessive </div><div>overhead, regulation, and external and internal </div><div>micromanagement] was not as much an issue.</div><div id="yiv510374316yui_3_2_0_16_1319355050432283"> </div><div id="yiv510374316yui_3_2_0_17_1319355050432235"><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span><br><br>> What is your basic approach?</div><div> </div><div>The key is the neuron equivalent. In my approach</div><div>the processor communicates optically with
other</div><div>processors. The processors are in an optically</div><div>friendly 3-D support structure where they can</div><div>communicate directly with thousands of other</div><div>processors.</div><div id="yiv510374316yui_3_2_0_16_1319355050432330"> </div><div id="yiv510374316yui_3_2_0_16_1319355050432332">I wrote:</div><div id="yiv510374316yui_3_2_0_17_1319355050432239"><br>> This concept
would be a continuation
of the idea<br>> of a series of minor human upgrades - possibly<br>> involving a kind of smart phone technology and<br>> minor implants and added sensor
capability.</div><div id="yiv510374316yui_3_2_0_16_1319355050432343"> </div><div id="yiv510374316yui_3_2_0_17_1319355050432249"><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span><br><br>> Why do you think now is the right time to attempt that?</div><div id="yiv510374316yui_3_2_0_16_1319355050432350"> </div><div>We have a whole generation of people hungry for knowledge</div><div>and prolific with smart phones. We also have a generation</div><div>of soldiers who have just been introduced to mobile sensor</div><div>displays, networking in the field, and computer on person</div><div>performing tasks for them as they work. Similar work done</div><div>in cockpit diplays has now moved to mobile individual users.</div><div> </div><div><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span><br><br>> I would very much like to hear a more detailed
proposal.</div><div> </div><div>I can think of many applications. Picking the most marketable</div><div>to detail is the trick. <br></div><div>I wrote:<br> <br>> The growth of the system<br>> can help pay for the AI work. The AI is<br>> integrated into the network as portions come on<br>> line. It would be expected that over the time<br>> frame of development there would be<br>> improvements in product delivery, improved<br>> sensors, improved implants, and improvements<br>> in the AI structure and ways to integrate it into<br>> the database.</div><div> </div><div><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span><br><br>> Do you think there
is a market? How much
work </div><div>> do you think you will need to service that market?</div><div> </div><div>You need to create a product in order to generate a</div><div>market. The market for automobiles was always there</div><div>- but the equipment and infrastructure were not for </div><div>millenia. The devices and increasing infrastructure</div><div>will have to grow together. You cannot plan up front</div><div>what all will be needed in great detail as it will evolve</div><div>depending on the mechanisms produced and how</div><div>the market of customers react to it.<br><var id="yiv510374316yui-ie-cursor"></var></div><div><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span><br><br>> My concern about the type of bionic technology </div><div>> available today is far too primitive and that if you </div><div>> use it, you'd be boxing yourself into a dead-end </div><div>>
upgrade and would have great difficulty transitioning </div><div>> to a more powerful technology. The body is resilient </div><div>> but it is easily injured. A poorly designed upgrade </div><div>> could do serious damage to both the patient and to </div><div>> the reputation of transhumanism as a whole.</div><div> </div><div>I agree which is why I support the first generation</div><div>devices using existing sensory input with subtle</div><div>human interaction - like a smart heads up display</div><div>with sound systems. The military has already</div><div>been working on eye movement controlled heads</div><div>up displays and silent digital communication.</div><div>Integrate into that sophisticated external sensors</div><div>and wireless capabilities and you have something</div><div>well beyond smart phones but still no direct</div><div>medical issues. The next obvious step from</div><div>there is first
generation commerical implants</div><div>taking the most successful processes a step</div><div>further.<br><br>I wrote: <br><br>> Like anything it would necessarily start small<br>> with a tiny number of applications, sensors,<br>> and implant options.<br><br><span id="yiv510374316yui_3_2_0_16_1319355050432205">Alan Grimes wrote:</span></div><div> </div><div>> Again, I see AI as a necessary pre-requisite </div><div>> to a highly capable neural interface. Existing </div><div>> neural interfaces are problematic for a number </div><div>> of reasons. Also, they don't really interface with
</div><div>> the brain at an interesting level yet.</div><div> </div><div>We can't start from the final answer. It would be</div><div>like planning for the interstate highway system</div><div>based on experience with wagon technology prior </div><div>to the railroads.</div><div> </div><div>My view:</div><div> </div><div>Step one - non-medical upgrades with external</div><div>senses and processing - more like personal</div><div>computing assistants and heads up displays.</div><div> </div><div>Step two - money from step one funds implant</div><div>and AI research. Implant trials before attempting</div><div>any marketing of product. AI rearch continues.</div><div> </div><div>Step three - 1st generation implant products come</div><div>on line. What happens after step three and AI</div><div>being developed is too far out to project. At that</div><div>point everything will snowball very
quickly.</div><div> </div><div>Dennis May<var id="yui-ie-cursor"></var></div></div></div></div></div></div></div></div></div></div></div></div></div></div></body></html>