[CDSNL] False "Great Leap Forward" in AI
Juyang Weng
juyang.weng at gmail.com
Sat Jul 6 23:34:09 EDT 2024
Dear Sim,
You wrote, "I talked to a well-wisher for this community and he
suggested we cool things a bit". You are among very few scientists who
dared to openly discuss opposing views. It is very important for the AI
community to hear opposing views. The AI community must not be quiet about
this widespread misconduct any longer. Too much money is being wasted
every year, many of our students have been misguided, and the worldwide
public has been cheated by the AI community for too long! From the IEEE
CDS Newsletters, Vol. 1, No. 1 and No. 2, we can see that only those
researchers who are honest have openly discussed and admitted the
misconduct, but not the three Turing Award 2018 receipients and the authors
of some corporations that produded widely-cited but allegedly misconduct AI
papers, although I invited them to respond. If they responded like you,
the public could see the truth better and faster! Please continue our
debates.
You wrote, "My suggestion would be to have him present his viewpoints
on some website. Of course, he has a number of postings on this issues. I
would suggest he open up one of the postings for public comments and
discussion." Yes, the IEEE CDS Newsletters has been doing that. You and
all others are welcome to respond. Any email lenghs are fine. The
deadline is July 10, for the following issue No. 3, but all Dialogues from
this year hsve a unique stage (e.g., [AI Criss]) so that late responses can
also be published. I hope that this debate about the "Great Leep Foward
in AI" will continue. I am thinking about creating a video talk or even an
exhibition on AI controversies if there are enough supports.
You wrote, "I still recall Horace Barlow’s ... note to me on the
grandmother cell theory: ... though I fear that what I have written will
not be universally accepted, at least at first!”. If you understand DN3,
the first model for conscious learning that starts from a single cell, you
will see how the grandmother cell theory is naive.
Since you need to be a subscribing member of AMDNL listserv to post, I
have posted your views on AMDNL by quoting your exact comments. Thank you
for posting but you can publish formally on IEEE CDS Newsletters if you
submit your views to the [AI Crisis] Dialogue.
Best regards,
-John
On Sat, Jul 6, 2024 at 8:08 PM Asim Roy <ASIM.ROY at asu.edu> wrote:
> Dear All,
>
>
>
> I talked to a well-wisher for this community and he suggested we cool
> things a bit. I know John for many years. We have disagreements on many
> issues and love to argue. You see some of that on the Connectionists list.
> Freedom of thinking and freedom of expression is important to the field.
> So, my apologies to John for using some harsh words.
>
>
>
> John wanted to meet and talk about this in Yokohama. My suggestion would
> be to have him present his viewpoints on some website. Of course, he has a
> number of postings on this issues. I would suggest he open up one of the
> postings for public comments and discussion.
>
>
>
> I still recall Horace Barlow’s (Horace Barlow - Wikipedia
> <https://en.wikipedia.org/wiki/Horace_Barlow>, Darwin’s great grandson),
> note to me on the grandmother cell theory: “and of course the more widely
> discussed it is the better pleased I shall be, though I fear that what I
> have written *will not be universally accepted, at least at first*!” So,
> as scientists, we should not be afraid to present our viewpoints.
>
>
>
> Best,
>
> Asim
>
>
>
>
>
> *From:* Juyang Weng <juyang.weng at gmail.com>
> *Sent:* Friday, July 5, 2024 6:23 PM
> *To:* Asim Roy <ASIM.ROY at asu.edu>
> *Cc:* Russell T. Harrison <r.t.harrison at ieee.org>; Akira Horose <
> ahirose at ee.t.u-tokyo.ac.jp>; Hisao Ishibuchi <hisao at sustech.edu.cn>;
> Simon See <ssee at nvidia.com>; Kenji Doya <doya at oist.jp>; Robert Kozma <
> rkozma55 at gmail.com>; Simon See <Simon.CW.See at gmail.com>; Yaochu Jin <
> Yaochu.Jin at surrey.ac.uk>; Xin Yao <xiny at sustech.edu.cn>;
> amdnl at lists.cse.msu.edu; Danilo Mandic <d.mandic at imperial.ac.uk>; Irwin
> King <irwinking at gmail.com>; Jose Principe <principe at cnel.ufl.edu>; Marley
> Vellasco <marley at ele.puc-rio.br>
> *Subject:* Re: False "Great Leap Forward" in AI
>
>
>
> Dear Asim,
>
> Thank you for your response, so that people on this email list can
> get important benefits. The subject is very new. I can raise these
> misconducts because we have a holistic solution to the 20
> million-dollar problems.
>
> You wrote, "he does use an optimization method to weed out bad
> solutions." This is false. DN does not weed out bad solutions, since it
> has only one solution.
>
> You wrote, "In optimization, we only report the best solution." This
> is misconduct, if you hide bad-looking data, like hiding all other students
> in your class.
>
> You wrote, "There is no requirement to report any non-optimal
> solutions." This is not true for scientific papers and business reports.
>
> You wrote, "If someone is doing part of the optimization manually,
> post-hoc, there is nothing wrong with that either." This is false because
> the so-called post-hoc solution did not have a test!
>
> You wrote, "In fact, there is plenty of evidence in biology that it
> can create new circuits and reuse old circuits and cells/neurons. Thus,
> throwing out bad solutions happens in biology too." This is irrelevant,
> as your mother is not inside your skull, but a human programmer is doing
> that inside the "skull."
>
> You wrote, "at a higher level, there’s natural selection and survival
> of the fittest. So, building many solutions (networks) and picking the best
> fits well with biology." As I wrote before, this is false, since biology
> has built Aldof Hitler and many German soldiers who acted during the Second
> World War. We report them, not hiding them.
>
> You wrote, "John calls this process `cheating' and a `misdeed".”
> Yes, I still do.
>
> You wrote, "he claims his algorithm gets the globally optimal
> solution, doesn’t get stuck in any local minima." This is true, since we
> do not have a single objective function as you assumed. Such a single
> objective function is a restricted environment or government. Instead,
> the maximum likelihood computation in DN is conducted in a distributed way
> by all neurons, each of them having its own maximum likelihood mechanism
> (optimal Hebbain mechanism). Read a book, Juyang Weng, Natural and
> Artificial Intelligence, available at Amonzon.
>
> You wrote, "If that is true, he should get far better results than the
> folks who are “cheating” through post-selection." Off course, we did as
> early as 2016. See "Luckiest from Post vs Single DN" in the attached
> file 2024-06-30-IJCNN-Tutorial-1page.pdf. Furthermore, the luckiest
> from the cheating is only a fitting error on the validation set (not test),
> the single DN is a test error because DN does not fit the validation set.
> The latter should not be compared with the former, but we compared with
> them anyway.
>
> You wrote, "My hunch is, his algorithm falls short and can’t compete
> with the other ones." Your hunch is wrong. See above as you can see
> how wrong you are. DN is a lot better than even the false performance.
>
> You wrote, "And that’s the reason for this outrage against others."
> I am honest. All others should be honest too. Do not cheat like many
> Chinese in the Great Leap Forward.
>
> You wrote, "I would again urge IEEE to take action against John Weng
> for harassing plenary speakers at this conference and accusing them of
> “misdeeds.” I am simply trying to exercise my freedom of speech driven
> by my care for our community.
>
> Do you all see a "Great Leap Forward in AI" like the "Great Leap
> Forward" in 1958 in China?
>
> Best regards,
>
> -John
>
>
>
> On Fri, Jul 5, 2024 at 9:01 AM Asim Roy <ASIM.ROY at asu.edu> wrote:
>
> Dear All,
>
>
>
> Without getting into the details of his DN algorithm, he does use an
> optimization method to weed out bad solutions. In optimization, we only
> report the best solution. There is no requirement to report any non-optimal
> solutions. If someone is doing part of the optimization manually, post-hoc,
> there is nothing wrong with that either. In fact, there is plenty of
> evidence in biology that it can create new circuits and reuse old circuits
> and cells/neurons. Thus, throwing out bad solutions happens in biology too.
> And, of course, at a higher level, there’s natural selection and survival
> of the fittest. So, building many solutions (networks) and picking the best
> fits well with biology. However, John calls this process “cheating” and a
> “misdeed.” He also claims having a strong background in biology. So he
> should be aware of these processes.
>
>
>
> In addition, he claims his algorithm gets the globally optimal solution,
> doesn’t get stuck in any local minima. If that is true, he should get far
> better results than the folks who are “cheating” through post-selection. He
> should be able to demonstrate his superior solutions through the public
> competitions such as with Imagenet data. My hunch is, his algorithm falls
> short and can’t compete with the other ones. And that’s the reason for this
> outrage against others.
>
>
>
> I would again urge IEEE to take action against John Weng for harassing
> plenary speakers at this conference and accusing them of “misdeeds.”
>
>
>
> Best,
>
> Asim
>
>
>
> *From:* Juyang Weng <juyang.weng at gmail.com>
> *Sent:* Thursday, July 4, 2024 8:14 AM
> *To:* Asim Roy <ASIM.ROY at asu.edu>
> *Cc:* Russell T. Harrison <r.t.harrison at ieee.org>; Akira Horose <
> ahirose at ee.t.u-tokyo.ac.jp>; Hisao Ishibuchi <hisao at sustech.edu.cn>;
> Simon See <ssee at nvidia.com>; Kenji Doya <doya at oist.jp>; Robert Kozma <
> rkozma55 at gmail.com>; Simon See <Simon.CW.See at gmail.com>; Yaochu Jin <
> Yaochu.Jin at surrey.ac.uk>; Xin Yao <xiny at sustech.edu.cn>;
> amdnl at lists.cse.msu.edu; Danilo Mandic <d.mandic at imperial.ac.uk>; Irwin
> King <irwinking at gmail.com>
> *Subject:* Re: False "Great Leap Forward" in AI
>
>
>
> Dear Asim and All,
>
> I am happy that Asim responded so that he gave us all an opportunity to
> interactively participate in an academic discussion. We can defeat the
> false "Great Leap Forward".
>
> During the banquet of July 3, 2024, I was trying to explain to Asim why
> our Developmental Network (DN) only trains a single network, not multiple
> networks as all other methods do (e.g., neural networks with
> error-backprop, genetic algorithms, and fuzzy sets). (Let me know if there
> are other methods where one network is optimal and therefore is free from
> the local minima problem.)
>
> This single-network property is important because normally every
> developmental network (genome) must succeed in single-network development,
> from inception to birth, to death.
>
> Post-selection: A human programer trains multiple (n>1) predictors
> based on a fit set F, and then picks up the luckiest predictor based on a
> validation set (which is in the possession of the program). He suffers from
> the following two misconducts:
> Misconduct 1: Cheating in the absence of a test (because the test set
> T is absent).
>
> Misconduct 2: Hiding bad-looking data (other less lucky predictors).
>
> A. I told Asim that DN tests its performance from birth to death,
> across the entire life!
>
> B. I told Asim that DN does not hide any data because it trains a
> single brain and reports all its lifetime errors!
>
> Asim did not read our DN papers that I sent to him, or did not read
> them carefully, especially the proof of the maximum likelihood of DN-1.
> See Weng IJIS 2015,
> https://www.scirp.org/journal/paperinformation?paperid=53728
> <https://urldefense.com/v3/__https:/www.scirp.org/journal/paperinformation?paperid=53728__;!!IKRxdwAv5BmarQ!aCvWF-PEaRtFT0lr5G-TVd1WSX7BloN_D524nbIUhctg9BC609q63-E91LYTCtXzoEQMZbkc5gnl53le6ch2TgX4$>
> .
>
> At the banquet, I told Asim that the representation of DN is
> "distributed" like the brain and it collectively computes the maximum
> likelihood representation by very neuron using a limited resource and a
> limited amount of life experience. I told him that every brain is
> optimal, including his brain, my brain, and Aldolf Hitler's brain.
> However, every brain has a different experience. However, Asim apparently
> did not understand me and did not continue to ask what I meant by
> "distributed" maximum likelihood representation. Namely, every neuron
> incrementally computes the maximum likelihood representation of its own
> competition zone.
>
> Asim gave an expression about the maximum likelihood implying that
> every nonlinear objective function has many local minima! That seems to
> be a lack of understanding of my proof in IJIS 2015.
>
> (1) I told Asim that every (positive) neuron computes its competitors
> automatically (assisted by its dedicated negative neuron), so that every
> (positive) neuron has a different set of (positive) neuronal competitors.
> Because every neuron has a different competition zone, the maximum
> likelihood representation is distributed.
>
> (2) Through the distributed computing by all (limited number of)
> neurons that work together inside the DN, the DN computes the distributed
> maximum likelihood representations. Namely, every (positive) neuron
> computes its maximum likelihood representation incrementally for its unique
> competition zone. This is proven in IJIS 2015, based on the
> dual-optimality of Lobe Component Analysis. Through the proof, you can
> see how LCA converts a highly nonlinear problem for each neuron into a
> linear problem for each neuron, by defining observation as a
> response-weighted input (i.e., dually-optimal Hebbian learning). Yes, with
> this beautifully converted linear problem (inspired by the brain), neuronal
> computation becomes computing an incremental mean through time in every
> neuron. Therefore, a highly nonlinear problem of computing lobe components
> becomes a linear one. We know that there is no local minima problem in
> computing the mean of a time sequence.
>
> (3) As I presented in several of my IJCNN tutorials, neurons in DN
> start from random weights, but different random weights lead to the same
> network, because the initial weights only change the neuronal resources,
> but not the resulting network.
>
> In summary, the equation that Asim listed is for each neuron, but each
> neuron has a different instance of the expression. There is no search,
> not that Asim implied (without saying)! This corresponds to a holistic
> solution the 20-million dollar problems (i.e., the local minuma problem
> solved by the maximum-likelihood optimality). See
> https://ieeexplore.ieee.org/document/9892445
> <https://urldefense.com/v3/__https:/ieeexplore.ieee.org/document/9892445__;!!IKRxdwAv5BmarQ!aCvWF-PEaRtFT0lr5G-TVd1WSX7BloN_D524nbIUhctg9BC609q63-E91LYTCtXzoEQMZbkc5gnl53le6cX7cdu2$>
>
> However, all other learning algorithms have not solved this local
> minima problem. Therefore, they have to resort to trials and errors
> through training many predictors.
>
> Do you have any more questions?
> Best regards,
>
> -John
>
>
>
> On Thu, Jul 4, 2024 at 4:20 PM Asim Roy <ASIM.ROY at asu.edu> wrote:
>
> Dear All,
>
>
>
> There’s quite a bit of dishonesty here. John Weng can be accused of the
> same “misconduct” that he is accusing others of. He didn’t quite disclose
> what we discussed at the banquet last night. He is hiding all that.
>
>
>
> His basic argument is that we pick the best solution and report results on
> that basis. In a sense, when you formulate a machine learning problem as an
> optimization problem, that’s essentially what you are trying to do – get
> the best solution and weed out the bad ones. And HE DOES THE SAME IN HIS
> DEVELOPMENT NETWORK. When I asked him how his DN algorithm learns, he said
> it uses the maximum likelihood method, which is an old statistical method (Maximum
> likelihood estimation - Wikipedia
> <https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Maximum_likelihood_estimation__;!!IKRxdwAv5BmarQ!aCvWF-PEaRtFT0lr5G-TVd1WSX7BloN_D524nbIUhctg9BC609q63-E91LYTCtXzoEQMZbkc5gnl53le6RquWOgs$>).
> I quote from Wikipedia:
>
>
>
> The goal of maximum likelihood estimation is to find the values of the
> model parameters that *maximize the likelihood function over the
> parameter space*,[6]
> <https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Maximum_likelihood_estimation*cite_note-:0-6__;Iw!!IKRxdwAv5BmarQ!aCvWF-PEaRtFT0lr5G-TVd1WSX7BloN_D524nbIUhctg9BC609q63-E91LYTCtXzoEQMZbkc5gnl53le6VGxpEh8$> that
> is
>
> 𝜃^=argmax𝜃∈Θ𝐿𝑛(𝜃;𝑦) .[image: {\displaystyle {\hat {\theta
> }}={\underset {\theta \in \Theta }{\operatorname {arg\;max} }}\,{\mathcal
> {L}}_{n}(\theta \,;\mathbf {y} )~.}]
>
>
>
> So, by default, HE ALSO HIDES ALL THE BAD SOLUTIONS AND DOESN’T REPORT
> THEM. He never talks about all of this. He never mentions that I had talked
> about this in particular.
>
>
>
> I would suggest that based on his dishonest accusations against others
> and, in particular, against one of the plenary speakers here at the
> conference, that IEEE take some action against him. This nonsense has been
> going on for a longtime and it’s time for some action.
>
>
>
> By the way, I am not a member of IEEE. I am expressing my opinion only
> because he has falsely accused me also and I had enough of it. I have added
> Danilo Mandic and Irwin King to the list.
>
>
>
> Thanks,
>
> Asim Roy
>
> Professor, Information Systems
>
> Arizona State University
>
> Asim Roy | ASU Search <https://search.asu.edu/profile/9973>
>
> Lifeboat Foundation Bios: Professor Asim Roy
> <https://urldefense.com/v3/__https:/lifeboat.com/ex/bios.asim.roy__;!!IKRxdwAv5BmarQ!aCvWF-PEaRtFT0lr5G-TVd1WSX7BloN_D524nbIUhctg9BC609q63-E91LYTCtXzoEQMZbkc5gnl53le6QZXPE1y$>
>
>
>
> *From:* Juyang Weng <juyang.weng at gmail.com>
> *Sent:* Wednesday, July 3, 2024 5:54 PM
> *To:* Russell T. Harrison <r.t.harrison at ieee.org>
> *Cc:* Akira Horose <ahirose at ee.t.u-tokyo.ac.jp>; Hisao Ishibuchi <
> hisao at sustech.edu.cn>; Simon See <ssee at nvidia.com>; Kenji Doya <
> doya at oist.jp>; Robert Kozma <rkozma55 at gmail.com>; Simon See <
> Simon.CW.See at gmail.com>; Yaochu Jin <Yaochu.Jin at surrey.ac.uk>; Xin Yao <
> xiny at sustech.edu.cn>; Asim Roy <ASIM.ROY at asu.edu>; amdnl at lists.cse.msu.edu
> *Subject:* False "Great Leap Forward" in AI
>
>
>
> Dear Asim,
>
> It is my great pleasure to finally have somebody who argued with me
> about this important subject. I have attached the summary of this
> important issue in pdf.
>
> I alleged widespread false data in AI from the following two
> misconducts:
> Misconduct 1: Cheating in the absence of a test.
>
> Misconduct 2: Hiding bad-looking data.
>
> The following is a series of events during WCCI 2024 in Yokohama Japan.
>
> These examples showed that some active researchers in the WCCI community
> were probably not aware of the severity and urgency of the issue.
> July 1, in public eyes, Robert Cozma banned the chance for Simon See at
> NVidea to respond to my question pointing to a False "Great Leap Forward"
> in AI.
> July 1, Kenji Doya suggested something like "let misconduct go ahead
> without a correction" because the publications are not cited. But he still
> did not know that I alleged that AlphaFold as well as many almost all
> published Google's deep learning products suffer from the same
> Post-Selection misconduct.
> July 1, Asim Roy said to me "We need to talk" but he did not stay
> around to talk. I had a long debate during the Banquet last night. He
> seems to imply that post-selections of few networks and hiding the
> performance information of the entire population is "survival of the
> fittest". He did not seem to agree that all 3 billion human populations
> need to be taken into account in human evolution, at least a large number
> of samples like in human sensus.
> July 3, Yaochu Jin did not let me ask questions after a keynote talk.
> Later he seemed to admit that many people in AI only report the data they
> like.
>
> July 3, Kalanmoy Deb said that he just wanted to find a solution using
> genetic algorithms but did not know that his so-called solution did not
> have a test at all.
>
> July 1, I saw all books on the display on the Springer Table appear to
> suffer from Post-Selection misconduct.
>
> Do we have a false data flooded "Great Leap Forward" in AI? Why?
>
> I welcome all those interested to discuss this important issue.
> Best regards,
> -John Weng
> --
>
> Juyang (John) Weng
>
>
>
>
> --
>
> Juyang (John) Weng
>
>
>
>
> --
>
> Juyang (John) Weng
>
--
Juyang (John) Weng
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cse.msu.edu/pipermail/amdnl/attachments/20240707/ac627945/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 183 bytes
Desc: not available
URL: <http://lists.cse.msu.edu/pipermail/amdnl/attachments/20240707/ac627945/attachment-0001.png>
More information about the Amdnl
mailing list