Talk:Hugo de Garis

From Wikipedia, the free encyclopedia

This article is within the scope of WikiProject Biography. For more information, visit the project page.
Biography Portal
??? This article has not yet received a rating on the project's quality scale. [FAQ]

Please rate the article and, if you wish, leave comments here regarding your assessment or the strengths and weaknesses of the article.

Contents

[edit] POV

This article looks really POV by AI enthousiasts ill-informed about the field. de Garis's academic career is mostly considered a failure and de Garis is not taken seriously any more by most AI researcher. CAM's brain machine is a joke as the idea of putting lot's of computational power will solve anything by itself is very naive. This article really need a serious cleanup !

--nct 09:44, 13 January 2007 (UTC)

I agree with the above user that cleanup is in order. I don't know de Garis, but this article is basically a smear job. Example: failing in some technology in 2001 is not exactly rare! 206.188.56.24

To say that de Garis did fail with some technology in 2001 is to miss the point. He's been working on his "artificial brains" for 15 years now. It's the thing that he's most famous for in the AI community - every one of his papers in the nineties was on this topic. The fact that he's made no progress doesn't mean that what he's doing is impossible (after all, it is a very hard problem), but his lack of success, and his failure to meet his own targets (2001 - cat brain, now - monkey brain) is relevant and should be mentioned. 172.188.190.67 13:34, 23 August 2007 (UTC)

[edit] de Garis Evolvable hardware

de Garis didn't really "invent" the field of evolvable hardware!

Robot controllers have been evolved using evolvable hardware

the russian revolution was fought between the whites (the supporters of the czar) and the reds, who were communists of all feathers

[edit] Removed

I've removed some extensive ramblings on de Garis's ideas from the discussion page, since they have no bearing on the biographical article.

[edit] Terrans

The link to Terran does not seem appropriate here since the disambiguation page does not contain an article on this use of the term. Is there a Wikipedia policy on linking to articles focused on different meanings of the word/phrase linked?

[edit] NPOV

The article as it is now violates NPOV. It seems that just about every paragraph, if not every sentence, in the early part of the article is followed by some statement starting with "To date, he has failed to" or "No proof exists that" or "It is unclear whether". If there is a reliable source which exhibits this pervading skepticism regarding de Garis, then fine, let's cite that reliable source. If this skepticism is just the opinion of some Wikipedia editor then it doesn't belong here. -- 192.250.34.161 20:01, 10 August 2007 (UTC)

You have to be more specific in what you're complaining about. de Garis claimed that by 2001 he would create an artificial cat brain. The comment you removed, and the other you complained about:
"To date, he has failed to demonstrate that this approach could work."
"but, up to now, has failed to evolve anything that could be considered as a brain, or even a serious robot control system."
are both factually correct. If you have any citation that shows de Garis has actually created an artificial brain, then please add it. The only work mentioned in his published papers is that he managed to reproduce simple 2-input logic gates. Creating an artificial brain would be major international news, so again, if you have any evidence that de Garis has done what he said he would do, please add it. I don't believe that, when stating a hypothesis, also stating that there is no published evidence supporting that hypothesis is POV, but to take into your account your opinion I'll reword it to " To date, he has failed to publish any research demonstrating that this approach could work.". If you have any citations that he has published such research, please add them. 172.188.190.67 13:27, 23 August 2007 (UTC)

[edit] Recent edits by 64.62.138.32 (now Profhugodegaris)

If you're going to add stuff like:

This technique, involving the artifical evolution of neural structures to perform such functions as pattern recognition, motion control, decision making, memory, etc.

then you're going to have to cite some evidence that these things have actually been done. AFAIK de Garis has never presented a working robot control system. 129.215.37.42 17:00, 4 September 2007 (UTC)

I have reverted the most recent edit by the same IP, because:

to computer science graduate students so that they can understand TQC (Topological Quantum Computing), a robust form of quantum computation.

First off, this says that all you need to understand TQC is CS, math and physics. Not true. Also this is uncited - de Garis teaches 3 subjects, if you want to say he teaches Topological Quantum Computing, then at least cite a page from Wuhan university, or better yet, some expert in TQC acknowledging de Garis' work.

He claimed that this approach would enable the creation of what he terms "artificial brains" which would quickly surpass human levels of intelligence.

I don't understand the deletion of this passage. de Garis has stated this many times..

From 1994 onwards, de Garis published a series of papers in which he discussed the use of artifical evolution to evolve neural networks which were implemented in FPGA technology. de Garis used this technique to evolve a wide variety of neural net circuits that act as e.g. pattern recognizers, motion controllers, decision makers, memory modules, etc. Connecting such modules enables him to build "artificial brains".

As stated above, none of these things have ever been done by de Garis' systems. If you are going to add this, you need to cite some expert (not de Garis) saying that these things have been done.

de Garis is still working on his evolvable hardware artificial brains

If this is true, you need to cite another expert (who is obviously not de Garis), otherwise it's just a claim of de Garis.

He has published no papers in this area.

This is in the section "recent research". It is a factual account of the current state of his "research" into topological quantum computing. If its *just* an interest, then it isn't even research and should be removed.

established a "brain builder group"

If this actually happened, you need to cite it. de Garis said he was going to do it, but as far as I can see nobody else has ever talked about this "group", and nothing was ever released by this group. 172.214.130.253 11:53, 5 September 2007 (UTC)

The previous poster is making the same changes, now using the account Profhugodegaris. If it is de Garis, it would explain some of the more POV edits. Chris Bainbridge 15:53, 5 September 2007 (UTC)
Reverted again. He replaced a cited quote with something not cited, removed a factual statement:
Ultimately the project failed to produce a functional robot control system, and ATR terminated it in 2001.
He added some unverifiable unpublished stuff:
He uses this board to evolve neural net modules 50 times faster than on an ordinary PC. He has evolved neural net modules that recognize patterns, control motions, make decisions, store memory, etc. Since the Celoxica board costs only about $1500, a robot less than $1000, his new brain building research project will cost less than $2500, to build an artificial brain with several 10,000s of neural net modules. At this price, nearly any interested research group in the world can afford to do similarly.
He removed the cited reference to Chris Malcolm's paper criticising his position. He also bizarrely and randomly removed the end of the cited quote:
between the white and the red Russians."[1]
Profhugodegaris has done this before. I wonder what he has against the complete quote and citation. Chris Bainbridge 13:46, 19 September 2007 (UTC)

[edit] cosmists

ai !ai ? It only just a from of Transhumanism, is hugo de garis donot know about this? Gx9900gundam (talk) 05:50, 23 November 2007 (UTC)