<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.2.2" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: Speaking truth to parallelism</title>
	<link>http://scottaaronson.com/blog/?p=201</link>
	<description>The Blog of Scott Aaronson</description>
	<pubDate>Fri, 10 Sep 2010 04:22:09 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.2.2</generator>

	<item>
		<title>By: The Quantum Pontiff &#187; Welcome to the Feast, Tums Provided</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9920</link>
		<author>The Quantum Pontiff &#187; Welcome to the Feast, Tums Provided</author>
		<pubDate>Fri, 23 Feb 2007 16:51:16 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9920</guid>
		<description>[...] Over at Shtetl-Optimized, Scott goes a little ballistic on criticism he&#8217;s received over the wording of his and Umesh Vazirani&#8217;s letters to the Economist. (As a side note, it is kind of sad to see such a poorly written article in the Economist: I know for a fact that an early Economist article on Shor&#8217;s algorithm drew a lot of great people into the field of quantum computing.) Part of Scott&#8217;s issue is with &#8220;physicists&#8221; insisting on rigor when they themselves are &#8220;the headmasters of handwaving.&#8221; So when Scott says   Today it is accepted that quantum computers could not solve NP-complete problems in a reasonable amount of time. [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] Over at Shtetl-Optimized, Scott goes a little ballistic on criticism he&#8217;s received over the wording of his and Umesh Vazirani&#8217;s letters to the Economist. (As a side note, it is kind of sad to see such a poorly written article in the Economist: I know for a fact that an early Economist article on Shor&#8217;s algorithm drew a lot of great people into the field of quantum computing.) Part of Scott&#8217;s issue is with &#8220;physicists&#8221; insisting on rigor when they themselves are &#8220;the headmasters of handwaving.&#8221; So when Scott says   Today it is accepted that quantum computers could not solve NP-complete problems in a reasonable amount of time. [&#8230;]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Greg Kuperberg</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9779</link>
		<author>Greg Kuperberg</author>
		<pubDate>Sun, 18 Feb 2007 17:44:03 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9779</guid>
		<description>&lt;i&gt;A few words in my defense: I said “qubit MAY not have a precise meaning” as a reference to previous posts, which advance the claim that it does not (see the first post of Robin Blume-Kohout in this discussion).&lt;/i&gt;

Touche, I have met my match at hairsplitting.  :-)</description>
		<content:encoded><![CDATA[<p><i>A few words in my defense: I said “qubit MAY not have a precise meaning” as a reference to previous posts, which advance the claim that it does not (see the first post of Robin Blume-Kohout in this discussion).</i></p>
<p>Touche, I have met my match at hairsplitting.  <img src='http://scottaaronson.com/blog/wp-includes/images/smilies/icon_smile.gif' alt=':-)' class='wp-smiley' /></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Scott</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9775</link>
		<author>Scott</author>
		<pubDate>Sun, 18 Feb 2007 16:14:38 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9775</guid>
		<description>Yeah, I figured that out (see above).</description>
		<content:encoded><![CDATA[<p>Yeah, I figured that out (see above).</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Richard</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9774</link>
		<author>Richard</author>
		<pubDate>Sun, 18 Feb 2007 15:43:28 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9774</guid>
		<description>&lt;i&gt;The article was written by one Stephen Jeffrey, who I assume is male.&lt;/i&gt;

No, that's just the credit for the &lt;i&gt;illustration&lt;/i&gt; in the article. The author's name is not given.</description>
		<content:encoded><![CDATA[<p><i>The article was written by one Stephen Jeffrey, who I assume is male.</i></p>
<p>No, that&#8217;s just the credit for the <i>illustration</i> in the article. The author&#8217;s name is not given.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Steve Demuth</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9771</link>
		<author>Steve Demuth</author>
		<pubDate>Sun, 18 Feb 2007 13:49:37 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9771</guid>
		<description>A few words in my defense: I said "qubit MAY not have a precise meaning" as a reference to previous posts, which advance the claim that it does not (see the first post of Robin Blume-Kohout in this discussion).  Now, by my dictionary, the meaning of the word "may" in this context is clearly to indicate that there is some room for disagreement on the matter.  Q.E.D.

As for my "contentious" definition of "commercially viable:"  I work every day with the product marketing folks of both my own employer, and our actual and potential technology partners, which range from some of the largest and best established tech companies in the world, to some of the smallest and most precariously situated.  In every instance I can document, those folks were working extremely hard to make sure that their products were commercially viable in the sense I gave.  I can't say that there are no scoundrels in the world who misuse the notion (The D-Wave announcement argues strongly otherwise), but I don't think there is anything ambiguous about the phrase "commercially viable" in the technology or marketing community at large.</description>
		<content:encoded><![CDATA[<p>A few words in my defense: I said &#8220;qubit MAY not have a precise meaning&#8221; as a reference to previous posts, which advance the claim that it does not (see the first post of Robin Blume-Kohout in this discussion).  Now, by my dictionary, the meaning of the word &#8220;may&#8221; in this context is clearly to indicate that there is some room for disagreement on the matter.  Q.E.D.</p>
<p>As for my &#8220;contentious&#8221; definition of &#8220;commercially viable:&#8221;  I work every day with the product marketing folks of both my own employer, and our actual and potential technology partners, which range from some of the largest and best established tech companies in the world, to some of the smallest and most precariously situated.  In every instance I can document, those folks were working extremely hard to make sure that their products were commercially viable in the sense I gave.  I can&#8217;t say that there are no scoundrels in the world who misuse the notion (The D-Wave announcement argues strongly otherwise), but I don&#8217;t think there is anything ambiguous about the phrase &#8220;commercially viable&#8221; in the technology or marketing community at large.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: TRRothwell</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9756</link>
		<author>TRRothwell</author>
		<pubDate>Sun, 18 Feb 2007 02:59:45 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9756</guid>
		<description>Interesting fact.  3 of 4 of D-waves theoretical physicists left in summer 2006 and go by the names P.Love A.Maassen van der Brink and A. Smirnov, all rather noted for some nice contributions in JJ and computational physics.  

D-wave quickly took there science team link off in may of thier site and never replaced it (ive been following it rather  rabidly hoping for updates).  The only reason such an exodus would occur is: 

A)  There was no interesting physics left and the hybrid superconducting Q.C was theoretically complete, implying only technical barriers remain.  (hopefully true.  it would be so cool if this is true)
B)  The machine was doomed to failure and the people who actually understand the physics bailed before it tanked. (hopefully not true)
C)  The company is just a VC strategic umbrella to hedge bet large sums of money against the possibility of success versus the all too common practice of the principle stakeholders taking their low strike price options and excercising them with the inside knowledge that B) is true.  (very likely the truth)

Anyways, we all hope that dwave actually makes history and validates their system by submitting to peer review and kicking all of our skeptical butts.  However, the excuse of 'keeping our findings to ourselves' because they are a for profit company doesn't really seem rational to me.  Do they not protect everything they build with iron clad patents and a team of high priced, visciously protective lawyers?  If so, without any real competitors on the horizon, wouldn't they want to submit to us their results and show us they have made history?  As S.Lloyd said, the proof is in the eating of the pudding.  The CTO of d-wave is behaving irrationally by posting here and defending his machine.  Why not just validate it instead, by making demonstrations of comparitive, observable speed-up (the only point unless you just love science) or even that coherent quantum effects are being harnessed? (which is enough for me!)

apologies if I have made ignorant assumptions; I am not a true lifelong expert in the field.  However if i was an investor, I would steer very clear of investment or even association with d-wave until the machine is scaled properly, and produces the aforementioned speed up with coherent quantum effects cleary proven at a feasibly useful scale for solving some currently intractable problems.</description>
		<content:encoded><![CDATA[<p>Interesting fact.  3 of 4 of D-waves theoretical physicists left in summer 2006 and go by the names P.Love A.Maassen van der Brink and A. Smirnov, all rather noted for some nice contributions in JJ and computational physics.  </p>
<p>D-wave quickly took there science team link off in may of thier site and never replaced it (ive been following it rather  rabidly hoping for updates).  The only reason such an exodus would occur is: </p>
<p>A)  There was no interesting physics left and the hybrid superconducting Q.C was theoretically complete, implying only technical barriers remain.  (hopefully true.  it would be so cool if this is true)<br />
B)  The machine was doomed to failure and the people who actually understand the physics bailed before it tanked. (hopefully not true)<br />
C)  The company is just a VC strategic umbrella to hedge bet large sums of money against the possibility of success versus the all too common practice of the principle stakeholders taking their low strike price options and excercising them with the inside knowledge that B) is true.  (very likely the truth)</p>
<p>Anyways, we all hope that dwave actually makes history and validates their system by submitting to peer review and kicking all of our skeptical butts.  However, the excuse of &#8216;keeping our findings to ourselves&#8217; because they are a for profit company doesn&#8217;t really seem rational to me.  Do they not protect everything they build with iron clad patents and a team of high priced, visciously protective lawyers?  If so, without any real competitors on the horizon, wouldn&#8217;t they want to submit to us their results and show us they have made history?  As S.Lloyd said, the proof is in the eating of the pudding.  The CTO of d-wave is behaving irrationally by posting here and defending his machine.  Why not just validate it instead, by making demonstrations of comparitive, observable speed-up (the only point unless you just love science) or even that coherent quantum effects are being harnessed? (which is enough for me!)</p>
<p>apologies if I have made ignorant assumptions; I am not a true lifelong expert in the field.  However if i was an investor, I would steer very clear of investment or even association with d-wave until the machine is scaled properly, and produces the aforementioned speed up with coherent quantum effects cleary proven at a feasibly useful scale for solving some currently intractable problems.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: GMHurley</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9750</link>
		<author>GMHurley</author>
		<pubDate>Sat, 17 Feb 2007 23:31:01 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9750</guid>
		<description>Steve Demuth advances a contentious interpretation of the phrase "commercially viable."  To a corporate financier, it means more or less what he says it does.  To a member of the marketing department, it means something like "we have a fighting chance of selling at least one of these things, and getting cash for it, with no claw-back when it doesn't do what we said it would do."  The threshold for "a fighting chance" is particularly low if you're trying to sell it to some department of government.</description>
		<content:encoded><![CDATA[<p>Steve Demuth advances a contentious interpretation of the phrase &#8220;commercially viable.&#8221;  To a corporate financier, it means more or less what he says it does.  To a member of the marketing department, it means something like &#8220;we have a fighting chance of selling at least one of these things, and getting cash for it, with no claw-back when it doesn&#8217;t do what we said it would do.&#8221;  The threshold for &#8220;a fighting chance&#8221; is particularly low if you&#8217;re trying to sell it to some department of government.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Greg Kuperberg</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9749</link>
		<author>Greg Kuperberg</author>
		<pubDate>Sat, 17 Feb 2007 21:40:26 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9749</guid>
		<description>&lt;i&gt;Qubit may not yet have a precise meaning, but “commercially viable” does.&lt;/i&gt;

I would say that the meaning of "qubit" as about as precise as "commercially viable".  What you really mean is that "qubit" has a much more arcane meaning.  The rigorous meaning of "commercially viable" is of course important to any good businessman, so that they do not have the excuse that it's okay to pantomime to the ignorant.</description>
		<content:encoded><![CDATA[<p><i>Qubit may not yet have a precise meaning, but “commercially viable” does.</i></p>
<p>I would say that the meaning of &#8220;qubit&#8221; as about as precise as &#8220;commercially viable&#8221;.  What you really mean is that &#8220;qubit&#8221; has a much more arcane meaning.  The rigorous meaning of &#8220;commercially viable&#8221; is of course important to any good businessman, so that they do not have the excuse that it&#8217;s okay to pantomime to the ignorant.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Steve Demuth</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9748</link>
		<author>Steve Demuth</author>
		<pubDate>Sat, 17 Feb 2007 21:17:34 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9748</guid>
		<description>The phrase "the first commercially viable quantum computer" tells me everything I need to know about the basic intellectual integrity of D-Wave as a company.  Qubit may not yet have a precise meaning, but "commercially viable" does: it means that you've built something, put it on the market, and managed to sell enough in an ongoing business to show an ongoing rate of return on investment that is greater than what one could get investing in low-risk, fixed interest securities.  Anything short of this may be interesting, may be a breakthrough, may even be marginally profitable, but it isn't "commercially viable."  

In other words, you can't claim a commercially viable product until after the fact, because it's the buyers who vote on whether you're viable, not your marketing department. 

Furthermore, D-Wave essentially claims to have solved the class of NP-Complete problems, even though they clearly know that they have not in any meaningful sense done so.  (A lawyer might argue that they don't really make that claim, but that's how it reads.  I know this because a couple of my colleagues "impeached" a short tutorial I gave on Quantum Computing earlier this week by bringing me the press release with the remark "Didn't you say that QC cannot solve problems in the class NP in P time?  These guys think otherwise.")  Rose's response in this column that their claim is comparable to ILOG's (disclaimer, and claim to competence: I work for ILOG) about their optimization products is baloney.  Here's the headline claim for ILOG's CPlex product: "[ILOG's] mathematical optimization technology enables better decision-making for efficient resource utilization."  That's about as racy as it gets.

There's good reason for this.  We're a publicly held company, and we can be held to account by investors for claims we make that we can't back up.   If D-Wave were a public company, and made the claims in an investment prospectus that they make on their website, they'd all be at risk of jail time.</description>
		<content:encoded><![CDATA[<p>The phrase &#8220;the first commercially viable quantum computer&#8221; tells me everything I need to know about the basic intellectual integrity of D-Wave as a company.  Qubit may not yet have a precise meaning, but &#8220;commercially viable&#8221; does: it means that you&#8217;ve built something, put it on the market, and managed to sell enough in an ongoing business to show an ongoing rate of return on investment that is greater than what one could get investing in low-risk, fixed interest securities.  Anything short of this may be interesting, may be a breakthrough, may even be marginally profitable, but it isn&#8217;t &#8220;commercially viable.&#8221;  </p>
<p>In other words, you can&#8217;t claim a commercially viable product until after the fact, because it&#8217;s the buyers who vote on whether you&#8217;re viable, not your marketing department. </p>
<p>Furthermore, D-Wave essentially claims to have solved the class of NP-Complete problems, even though they clearly know that they have not in any meaningful sense done so.  (A lawyer might argue that they don&#8217;t really make that claim, but that&#8217;s how it reads.  I know this because a couple of my colleagues &#8220;impeached&#8221; a short tutorial I gave on Quantum Computing earlier this week by bringing me the press release with the remark &#8220;Didn&#8217;t you say that QC cannot solve problems in the class NP in P time?  These guys think otherwise.&#8221;)  Rose&#8217;s response in this column that their claim is comparable to ILOG&#8217;s (disclaimer, and claim to competence: I work for ILOG) about their optimization products is baloney.  Here&#8217;s the headline claim for ILOG&#8217;s CPlex product: &#8220;[ILOG&#8217;s] mathematical optimization technology enables better decision-making for efficient resource utilization.&#8221;  That&#8217;s about as racy as it gets.</p>
<p>There&#8217;s good reason for this.  We&#8217;re a publicly held company, and we can be held to account by investors for claims we make that we can&#8217;t back up.   If D-Wave were a public company, and made the claims in an investment prospectus that they make on their website, they&#8217;d all be at risk of jail time.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Greg Kuperberg</title>
		<link>http://scottaaronson.com/blog/?p=201#comment-9747</link>
		<author>Greg Kuperberg</author>
		<pubDate>Sat, 17 Feb 2007 20:25:23 +0000</pubDate>
		<guid>http://scottaaronson.com/blog/?p=201#comment-9747</guid>
		<description>Well, bully for Wikipedia.  Its article on D-Wave is duly skeptical.  It also links to an astute AP article in which D-Wave itself admits that the world's first commercially viable quantum computer could just be stone soup.  Note that the journalist, Jordan Robertson, said nothing about the NP-in-one-shot nonsense.  In my experience, science journalists are at least smarter when you tell them the truth.</description>
		<content:encoded><![CDATA[<p>Well, bully for Wikipedia.  Its article on D-Wave is duly skeptical.  It also links to an astute AP article in which D-Wave itself admits that the world&#8217;s first commercially viable quantum computer could just be stone soup.  Note that the journalist, Jordan Robertson, said nothing about the NP-in-one-shot nonsense.  In my experience, science journalists are at least smarter when you tell them the truth.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
