Wikipedia:Articles for deletion/Confabulation (neural networks)


 * The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review).  No further edits should be made to this page.

The result was   no consensus. Alot of commentary over a two week period with no strong consensus emerging. Nja 247 08:56, 24 July 2009 (UTC)

Confabulation (neural networks)

 * ( [ delete] ) – (View AfD) (View log)

No independent sources to affirm notability per WP:GNG, and I can't seem to find any; all seems to be primary sources  Chzz  ►  01:24, 10 July 2009 (UTC)


 * — LongJohnPlatinum (talk • contribs) has made few or no other edits outside this topic.
 * — Be-A-Bhodi (talk • contribs) has made few or no other edits outside this topic.
 * — Aravind.g1001 (talk • contribs) has made few or no other edits outside this topic.
 * — Rbsmth42 (talk • contribs) has made few or no other edits outside this topic.
 * — Port Tiger (talk • contribs) has made few or no other edits outside this topic.


 * I see secondary sources. Doesn't Scientific American, various reputable book authors, and Pullitzer Prize winning newspapers count? —Preceding unsigned comment added by Periksson28 (talk • contribs) 04:25, 10 July 2009 (UTC)
 * Sources that are independent of the subject need to provide significant coverage. Simply having secondary sources isn't enough. It's the significance given to this subject in those sources that is being questioned, I assume. Sancho 07:57, 10 July 2009 (UTC)


 * Comment : The term comes up on citeseer, which is a good pace to look for academic computer topics, http://citeseer.ist.psu.edu/cis?q=confabulation&cs=1 , but what got me was ref to thalamic-corital loop ( and lack of caffeine). There was another up for deletion on EM theories of consciousness as it seemed to be a collection of fringe theories but that seems to be the "state of the art" in that area. I would suggest a merge or a link or list or catagory on "theories of consciousness" if that is what this is about. EM theories are more interesting and offer some hope for people who don't want to be thought of as neural networks (brain-independent mind of course brings in people who learned science from watching Ghost Busters but so does credible cold fusion work). Nerdseeksblonde (talk) 09:56, 10 July 2009 (UTC)
 * Note: This debate has been included in the list of Science-related deletion discussions.  -- TexasAndroid (talk) 11:57, 10 July 2009 (UTC)

I see all Wikipedia criteria being met here. Before, this was a paltry, one-sided article. Then others balanced it, with credible peer-reviewed references (i.e., the prestigious journal, Neural Networks, two very reputable science reporters with Ph.D.s in biology, Dr. Tina Hesman and Dr. Robert Holmes, plausible and successful authors in both robotics and legal implications of technology, Levy and Plotkin, not to mention respected NASA engineers and scientists. No, plenty of significance has been given here by secondary sources and none of these sources are ghostbusters franchisees. The fact of the matter is that millions around the world have been using products and services produced by confabulating neural networks, while the military throws millions into a new generation of battlefield robotics guided by these same principles. I sincerely recommend that this article be retained. —Preceding unsigned comment added by Periksson28 (talk • contribs) 17:10, 11 July 2009 (UTC)
 * Thaler usurping the majority of the article for himself is in no way "balancing" it. I don't think you're approaching this subject from a neutral standpoint.  Your comments make you out to be a fanboy, or perhaps Thaler himself (which, given what has happened on the article already, wouldn't be terribly surprising).  Whatever the case, lets leave the decision about the notability of the article and the sources, be they from Thaler himself or merely from someone pretending to be him, to people with more comment history on Wikipedia, shall we? LongJohnPlatinum (talk) 04:19, 13 July 2009 (UTC)

If Wikipedia featured an article claiming John Glenn to be the first man to orbit the earth, then anyone correcting that assertion would be considered usurping that piece, according to Mr. Long John. Naturally anyone responding to such a gross error, would be fans of Yuri Gagarin. In the case of this article, there have been several contributors.

Just check out the validity of the references. In many cases, they are the same as those referenced by the RHN fan who originally contributed and fibbed via omission. Whoever is adding to this stub (and I emphasize stub) is doing so at the invitation of Wikipedia, not necessarily because they are fans of one camp or another. Further, such claims are solidly backed by non-trivial sources, in many cases those used by RHN himself. —Preceding unsigned comment added by Periksson28 (talk • contribs) 14:35, 13 July 2009 (UTC) I recently read the book by Plotkin, The Genie in the Machine: How Computer-Automated Inventing is Revolutionizing Law and Business published by Stanford University Press. The book provides real-world examples of the products confabulation theory has been used to create -- everything from toothbrushes to creative robots. Also the paper by Patrick published in IEEE summarizes the efforts of NASA’s marshall space flight center harnessing such technologies to autonomous space vehicle docking to future missions. In another paper by Mayer again published in IEEE, uses this technique to design an autonomous mobile robot to adapt to changing environments.
 * Straw man arguments and attacks on my character are not germane to the issue of the notability of the article and its sources. As asked earlier, are those sources you keep mentioning covering the subject significantly?  That is a question best left to those with more Wikipedia experience than ourselves, and thus more credibility. LongJohnPlatinum (talk) 18:53, 13 July 2009 (UTC)

All these sources sound like secondary to me and also credible. I think this article is balanced and should be retained. (User talk:Rbsmth42 —Preceding undated comment added 17:30, 13 July 2009 (UTC).
 * The issue is not whether the article is "balanced" but if it, and its references, meet the notability requirements. LongJohnPlatinum (talk) 18:53, 13 July 2009 (UTC)

Good. I’m glad we’ve narrowed the debate to one of notability.

Please observe that consistent with Wikipedia Notability Guidelines, (1) Reliable sources (i.e., Stanford University Press, Springer, Scientific American, Elsevier, IEEE, NASA, etc.) have provided coverage of the subject. In each case, the context has not been that of confabulation in neural nets, but legal repercussions of autonomous discovery, brain death, materials discovery, and robotic control. --- check

(2) No empirical research is needed to extract the content. In each case, the research has been thoroughly laid out, especially when presented by science reporters aiming to explain the concept to the public (i.e., Hesman & Holmes). --- check

(3) Multiple secondary sources have been included, including scientists (Yam, Hesman, Holmes, engineers (Patrick), and lawyers (Plotkin). Each of these groups were required to perform a rigorous due diligence to determine authenticity of the subject science and technology. --- check

(4) All of the above sources solicited Thaler (not the other way around). I can’t imagine any of these authors writing articles that support an entrepreneurial business. --- check [User talk: aravind.g1001] —Preceding unsigned comment added by Aravind.g1001 (talk • contribs) 23:04, 13 July 2009 (UTC)
 * How can you know who solicited who definitively without being either one of the sources or Thaler himself? Further, how could you know what sort of due diligence each group was required to do without being Thaler or one of the sources? Given that you, Periksson28, and Rbsmith42 are all new users posting on the same obscure topic in neural networking, with the same lack of appreciation for wiki editing and commenting standards, and the same opinion/confusion over what the discussion on this page is about, I am wondering if you are not all sock puppets.  LongJohnPlatinum (talk) 03:54, 14 July 2009 (UTC)

Our loyalties are not germane to the discussion. We represent the truth and the current article meets all Wikipedia standards. It seems that only one person, Mr. Long John, is intent upon the destruction of this article because others had the gumption to point out that the subject of confabulation in neural nets has already been addressed in peer-reviewed papers, credible press, and US and international patents. The topic is not obscure, because there are many commercial, government, and military customers of this technological capability who have effectively taken part in the due diligence process.

When pinned down, Mr. Long John dogmatically denies the relevance of his loyalties. The truth is that we don't have any idea whose payroll he is on. —Preceding unsigned comment added by Periksson28 (talk • contribs) 15:17, 14 July 2009 (UTC)
 * Your continued assertions that the article and its sources meet the notability requirements do not constitute a demonstration of your claims, and neither do they constitute a demonstration of your credibility on the topic of notability to make such a claim. Your continued conspiracy theory assertions about my "loyalties" or about "who's payroll I'm on" are not germane to the discussion of notability.  LongJohnPlatinum (talk) 16:02, 14 July 2009 (UTC)

Then please provide substantive proof, and not dogmatic assertions, that article and references are not notable. Evidently, the article met Wikipedia notability requirements when presnted by an obvious RHN loyalist who conveniently omitted a very significant literature trail. Subsequent additions were solidly backed by verifiable references. —Preceding unsigned comment added by Periksson28 (talk • contribs) 16:36, 14 July 2009 (UTC)
 * I have made no assertions about the notability or lack thereof of this article and its references, and I shall refrain from making any because I am a new user/editor to Wikipedia and do not have the experience necessary to make an accurate judge of notability. I shall repeat myself yet again - let's leave the decision about the notability of the article and the sources to people with more comment history on Wikipedia, shall we?  Additionally, your assertions about the article's previous status are untrue - if it does not meet the notability requirements now, then it most certainly did not meet them previously, as it was little more than a stub.  More than likely, since this is such an obscure topic, it was simply overlooked by the admins for a long time.  Your continued conspiracy claims about the "loyalties" of previous contributors are not germane to the discussion. LongJohnPlatinum (talk) 17:34, 14 July 2009 (UTC)

No, you're the inexperienced one who pulled the train's emergency brake. Let's hear your reasons for lack of notability.

BTW, creative/cogent confabulation is a front and center topic judging from the press it has received over the last 15 years. All you need to do is look at the references. —Preceding unsigned comment added by Periksson28 (talk • contribs) 17:46, 14 July 2009 (UTC)
 * I remind you that it was not I that put the article up for deletion, but the administrator Chzz. Neither I, nor yourself, have any sufficient editing experience (as evidenced by our respective contribution histories) so as to give an authoritative opinion on the topic of notability as used by the Wikipedia standards.  LongJohnPlatinum (talk) 18:15, 14 July 2009 (UTC)

Let's see. First you raised the issues with the article. Subsequently Chzz nominated it for deletion. As a result, significant secondary sources have been added. The article has been toned down so as not to sound like a press release for anyone. Now you've introduced the next hurdle, notability. What is the next barrier? Please, again, what are your issues with notability? —Preceding unsigned comment added by Periksson28 (talk • contribs) 18:29, 14 July 2009 (UTC)
 * The whole reason that the article was nominated for deletion in the first place was notability - it wasn't an issue I raised. The article in its current state (with references and from a decidedly more NPOV) might meet the requirements, but again, that's something that people with more experience should decide. LongJohnPlatinum (talk) 18:40, 14 July 2009 (UTC)


 * — Be-A-Bhodi (talk • contribs) has made few or no other edits outside this topic. Strong Keep: It is clear that the sources provided are notable and of a respectable, authoritative and peer-reviewed nature. The article should be allowed to remain. Anyone who takes the time to read the topic and review the sources can clearly see this is not a vanity site; rather, the article meets all the necessary notability requirements by providing multiple, substantial credible sources that clearly establish precedence for the information being provided, and, provide readers a set of scientifically credible resources from which to launch (or continue) their own investigation of this topic. --Be-A-Bhodi (talk) 18:32, 15 July 2009 (UTC)

 Please add new comments below this notice. Thanks, ( X! ·  talk )  · @180  · 03:19, 17 July 2009 (UTC)
 * With regards to the notability of the sources, I would still prefer that someone with more editing experience (and thus, more credibility) chime in as to whether or not the sources do indeed meet the requirements. On a different note, am I the only one who finds it curious how many new users this topic is attracting in such a short period of time? LongJohnPlatinum (talk) 19:22, 15 July 2009 (UTC)
 * Comment Sources used in an article as references are not required to be notable, they only have to be reliable. It really isn't all that surprising that the AfD has attracted more attention to the article because AfDs tend to bring editors out of the woodwork. --Tothwolf (talk) 07:16, 16 July 2009 (UTC)
 * Keep Given the discussion and work that has gone into this article since the AfD nomination, I see no reason for deletion. The subject of the article is clearly notable, and I believe the original concerns that the article was mainly sourced to one particular body of work have largely been addressed. Given the discussion here and on the article's talk page I suspect the editors who have been working on the article will continue to improve it as well. --Tothwolf (talk) 07:16, 16 July 2009 (UTC)
 * Relisted to generate a more thorough discussion so consensus may be reached.


 * Keep. The article is still rather unclear, but it seems to be a reasonably noable idea.  I'm not sure, though, that this use of the word differs so strongly from the standard meaning of confabulation that it might not be merged there.  - Smerdis of Tlön (talk) 14:21, 17 July 2009 (UTC)
 * Please take a look at my review of the listed references. I found no proof of the subject's notability on the Web. — Rankiri (talk) 16:24, 17 July 2009 (UTC)


 * Delete: fails WP:N, falls under WP:POVFORK, possibly WP:PROMOTION. Confabulation is a valid psychological concept but I see no evidence that the term is discussed in this particular manner by anyone but Stephen Thaler. The article's editing history—as well as this AfD discussion—raise major WP:COI, WP:SPA or WP:CANVASS concerns. Exclusive Google searches like, or  show no relevant results. None of the listed references are valid or indicative of notability: the topic has to receive significant coverage in reliable secondary sources that are independent of the subject in order to satisfy the inclusion criteria for standalone articles.


 * Both Computers Compose Personalized Music and "The Machine That Invents" are not about confabulation but about Stephen Thaler, his views and creations. "Computers Compose Personalized Music" is fully available online — the term confabulation is not even mentioned.


 * Robots unlimited, by David N. L. Levy: no results found in this book for Confabulation


 * The Genie in the Machine, by Robert Plotkin: no results found in this book for Confabulation


 * "Demonstration of Self-Training Autonomous Neural Networks in Space Vehicle Docking Simulations" is co-written with Stephen Thaler. Even then, from what I can see, the term is enclosed in quotation marks and its mention in the article is trivial: “confabulations” of the information to arrive at new possibilities for solutions.


 * "A Modular Neurocontroller for Creative Mobile Autonomous Robots Learning by Temporal Difference, Systems, Man, and Cybernetics" does not use the term confabulation at all . — Rankiri (talk) 16:24, 17 July 2009 (UTC)


 * Delete: agree with Raniki. No significant coverage in reliable secondary sources that are independent of the subject. Sancho 20:34, 17 July 2009 (UTC)

— Port Tiger (talk • contribs) has made few or no other edits outside this topic. *Keep. I have to admit that my editing experience is limited, and I do know both cognitive scientists this article references (does that disqualify me?). i also have a firm grasp of the term "confabulation" in contexts of both high level and computational psychology. In the latter sense the terms confabulation-false memory-degraded memory-novel pattern-constraint violation-strange attractors-etc are all used interchangeably in the artificial neural network sense. both sources are trying to point out that degraded memories in the brain are at the heart of cognition and the literature seems to be steering toward "confabulation". After my perusal of the reference, I do agree with Rankiri that the second article by Hesman about music generation should be stricken. The other citations make sense, ultimately mentioning one or more of these equivalent terms. There are some important references missing here. Port Tiger (talk) 20:49, 17 July 2009 (UTC) — Port Tiger (talk&#32;• contribs) has made few or no other edits outside this topic.
 * Mentions are not "significant coverage". There needs to be significant coverage in secondary sources that are independent of the subject (in this case, independent of the subject should be taken to mean not written by the originator(s) of the idea). Sancho 21:44, 17 July 2009 (UTC)


 * Sorry for the poor word choice. For two consecutive days, the entire front page of a Pullitzer newspaper significantly covered this notion of "faulty memory" generation (i.e., confabulation) to generate name brand products and control battlefield robots. Thats significant, considering it was written by a very reputable Ph.D. biologist and science reporter, Tina Hesman. Cliffover, who I've just added to the references, is a prodigious writer and Ph.D. physicist, who hits the mark, discussing the noise-induced generation of "both memories and unusual juxtapoistions of those memories unprecedented in the network's experience." (i.e., confabulation). Upon a thorough read of the remaining references you will find that the identification is made between confabulation and false memory generation. After all, Wikipedia would want the terms "land beaver" and "groundhog" to be unified and searchable under one name. In all these cases, the resulting literature was significant and not written by the originators of the idea. Maybe a move would work, but as a newcomer I can't imagine any of these sources wanting to be called "insignificant" even in Wiki-parlance.Port Tiger (talk) 22:14, 17 July 2009 (UTC)
 * Once again, none of the mentioned sources satisfy WP:GNG. "Significant coverage" means that sources address the subject directly in detail, and no original research is needed to extract the content. If the term is not even mentioned, it's not really addressed directly in detail, is it? — Rankiri (talk) 03:27, 18 July 2009 (UTC)
 * Agree again with Rankiri. The problem is that you have to say "ie. confabulation" in your justification, because the articles don't even mention the term. It would be like me inventing a term bogification to mean the death of a celebrity, then turning to all the coverage of celebrity deaths to say that they're giving my term bogification significant coverage. Sancho 05:15, 18 July 2009 (UTC)
 * Disagree with Rankiri and Sancho. "Bogification" is not an appropriate analogy. The debate is not about something fairly common like a celebrity death, but a newly observed phenomenon in neural networks that has earned quite a bit of fame in both the press and books published by outside authors. Sure, there have been a lot of red herrings tossed out by recent writers on the subject as they change terminology to throw off the public. This article serves to undo that mischief. Furthermore, we are not talking about just a term, we are talking about a concept that goes by several different names as delineated in the article's body. Therefore, a word search alone doesn't prove much. As noted above, "ground hog" and "land beaver" should land a Wikipedia user at the same article, and citations to the ground hog artcle should not be abandoned because Ctrl F was unsuccessful in finding the term "gound hog" in that reference. Periksson28 (talk) 16:26, 18 July 2009 (UTC)
 * Disagree with Rankiri. The Hesman article alone addresses the topic significantly and in detail under the equivalent concept of degraded memories, as she clearly states in the beginning of the article. Nevertheless, LongJohnPlatinum continues to repeatedly destroy that citation. I am speaking as someone having a lifetime of experience in the field of neural networks. My research into the topic shows equivalency of the terms in question (and that was not a key word search).Periksson28 (talk) 19:51, 18 July 2009 (UTC)
 * From WP:NEO: A new term does not belong in Wikipedia unless there are reliable sources specifically about the term — not just sources which mention it briefly or use it in passing. As for the groundhog analogy, please see arguments to avoid in deletion discussions — Rankiri (talk) 23:00, 18 July 2009 (UTC)


 * See Neural Networks archive. Volume 18, Issue 2 (March 2005). Q.E.D. Periksson28 (talk) 23:44, 18 July 2009 (UTC)


 * Dr. Hesman does not just mention it briefly. She posits the phenomenon as the basis of a new brand of artificial intelligence, makes contact with how this mechanism works in the brain, and shows how it can be appplied to do many things.Periksson28 (talk) 23:51, 18 July 2009 (UTC)


 * I found nothing germane to the discussion in arguments to avoid in deletion discussions. Periksson28 (talk) 23:59, 18 July 2009 (UTC)

— 71.246.47.3 (talk) has made few or no other edits outside this topic. *---Keep---i agree with port tiger! —Preceding unsigned comment added by 71.246.47.3 (talk) 01:00, 18 July 2009 (UTC) — 71.246.47.3 (talk) has made few or no other edits outside this topic.
 * The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.