<html><body><div style="color:#000; background-color:#fff; font-family:times new roman, new york, times, serif;font-size:14pt"><div id="yiv1893516410"><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="yiv1893516410"><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="yiv1893516410yui_3_2_0_19_1317734478527261" class="yiv1893516410ms__id48426"><div id="yiv1893516410yui_3_2_0_18_131773447852743"><span id="yiv1893516410yui_3_2_0_18_131773447852757">Anders <span id="misspell-0" class="mark">Sandberg</span> wrote:</span></div><div><span id="yiv1893516410yui_3_2_0_18_131773447852763"></span> </div><div id="yiv1893516410yui_3_2_0_18_131773447852767"><span id="yiv1893516410yui_3_2_0_18_131773447852769">"...keynote talk was by Hubert Dreyfus and largely consisted
</span></div><div><span id="yiv1893516410yui_3_2_0_18_131773447852773">of him gloating about how he had been right about the </span></div><div id="yiv1893516410yui_3_2_0_18_131773447852778"><span id="yiv1893516410yui_3_2_0_18_131773447852782">failures of all the AI pioneers"</span></div><div><span id="yiv1893516410yui_3_2_0_18_131773447852787"></span> </div><div id="yiv1893516410yui_3_2_0_18_131773447852796"><span id="yiv1893516410yui_3_2_0_18_131773447852790">Everyone took short-cut routes instead of developing and</span></div><div id="yiv1893516410yui_3_2_0_18_1317734478527113"><span id="yiv1893516410yui_3_2_0_18_1317734478527109">implementing large scale </span><span id="yiv1893516410yui_3_2_0_18_131773447852799">neural networks - no mystery</span></div><div id="yiv1893516410yui_3_2_0_18_1317734478527121"><span id="yiv1893516410yui_3_2_0_18_1317734478527116">there. Unfortunately I still don't see any serious long
term</span></div><div><span id="yiv1893516410yui_3_2_0_18_1317734478527124">effort in that direction - just larger and larger short-cuts</span></div><div><span id="yiv1893516410yui_3_2_0_18_1317734478527129">with no plans
for follow-on except more and more short</span></div><div><span id="yiv1893516410yui_3_2_0_18_1317734478527137">term projects.</span></div><div><span id="yiv1893516410yui_3_2_0_18_131773447852784"></span> </div><div id="yiv1893516410yui_3_2_0_18_1317734478527143"><span id="yiv1893516410yui_3_2_0_18_1317734478527145"><span id="yiv1893516410yui_3_2_0_18_131773447852757">Anders <span id="misspell-1" class="mark">Sandberg</span> wrote:</span></span></div><div><span><span id="yiv1893516410yui_3_2_0_18_1317734478527150"></span></span> </div><div id="yiv1893516410yui_3_2_0_18_1317734478527156"><span><span id="yiv1893516410yui_3_2_0_18_1317734478527160">"I think this shows a general malaise of many fields: are </span></span></div><div id="yiv1893516410yui_3_2_0_18_1317734478527165"><span><span id="yiv1893516410yui_3_2_0_18_1317734478527169">there incentives for progress or incentives for churn?"</span></span></div><div
id="yiv1893516410yui_3_2_0_18_1317734478527187"> </div><div>Anders lists a number of observations of symptoms of</div><div id="yiv1893516410yui_3_2_0_18_1317734478527202">a problem. The root of the problem is not hard to </div><div id="yiv1893516410yui_3_2_0_18_1317734478527233">understand - it is central planning versus free markets.</div><div>Central planners represent a serial computer with a number</div><div>of bottlenecks in speed and a limited number of inputs</div><div id="yiv1893516410yui_3_2_0_18_1317734478527235">[limited knowledge]. Free markets are a massive parallel</div><div>computational project with nearly unlimited inputs and</div><div id="yiv1893516410yui_3_2_0_18_1317734478527238">outputs. You can build remarkable serial and/or partially </div><div>parallel computers [Watson] and conclude their successes </div><div>imply a path forward [seen versus unseen]. Or you can</div><div>learn from history
and what works in nature and unleash</div><div>the much greater <span id="yiv1893516410misspell-2"><span>potential</span></span> of what amounts to a massive</div><div>parallel operation in essentially neural network form </div><div>[much greater knowledge] - free markets.</div><div> </div><div>I would think that people in a forum like this would</div><div>understand information bottlenecks and serial versus</div><div>parallel processing and be able to apply it in economic</div><div id="yiv1893516410yui_3_2_0_19_1317734478527265">and political situations.</div><div id="yiv1893516410yui_3_2_0_19_1317734478527270"> </div><div id="yiv1893516410yui_3_2_0_19_1317734478527272">Big government and central planners may be able to</div><div id="yiv1893516410yui_3_2_0_19_1317734478527278">pull off a moon shot and build atomic bombs but the</div><div id="yiv1893516410yui_3_2_0_19_1317734478527282">science [Goddard] and [<span id="misspell-2"
class="mark">Szilárd</span>] was the work of a</div><div>tiny number of individuals. In the days of 6%-8%</div><div>real growth unleashed by free markets there would</div><div>have been real money available in free markets to</div><div>invest in new technologies versus government </div><div>siphoning it all off to buy votes [seen versus unseen].</div><div>Money in private hands can and did build huge</div><div>projects in the past - something central planners </div><div>would have us believe cannot be done today.</div><div> </div><div id="yiv1893516410yui_3_2_0_19_1317734478527260">Dennis May</div></div></div></div></div></div></body></html>