<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">see also "The Roles of Symbols in
Neural-Based AI: They Are Not What You Think!" by Daniel L.
Silver, Tom M. Mitchell<br>
<a class="moz-txt-link-freetext" href="https://ebooks.iospress.nl/volumearticle/63710">https://ebooks.iospress.nl/volumearticle/63710</a><br>
<br>
<div class="abstract"> <b>Abstract</b><br>
<section>
<p>We propose that symbols are first and foremost external
communication tools used between intelligent agents that
allow knowledge to be transferred in a more efficient and
effective manner than having to experience the world
directly. But, they are also used internally within an agent
through a form of self-communication to help formulate,
describe and justify subsymbolic patterns of neural activity
that truly implement thinking. Symbols, and our languages
that make use of them, not only allow us to explain our
thinking to others and ourselves, but also provide
beneficial constraints (inductive bias) on learning about
the world. In this paper we present relevant insights from
neuroscience and cognitive science, about how the human
brain represents symbols and the concepts they refer to, and
how today’s artificial neural networks can do the same. We
then present a novel neuro-symbolic hypothesis and a
plausible architecture for intelligent agents that combines
subsymbolic representations for symbols and concepts for
learning and reasoning. Our hypothesis and associated
architecture imply that symbols will remain critical to the
future of intelligent systems NOT because they are the
fundamental building blocks of thought, but because they are
characterizations of subsymbolic processes that constitute
thought.</p>
</section>
</div>
<br>
<br>
Am 07.06.24 um 10:06 schrieb Weng, Juyang:<br>
</div>
<blockquote type="cite"
cite="mid:MN2PR12MB45495710FFA1037C67FD596AD0FB2@MN2PR12MB4549.namprd12.prod.outlook.com">
<div class="elementToProof">
Dear Asim,</div>
<div class="elementToProof">
You wrote, That single cell firing in a cat’s brain having
“meaning” is not due to “Asim” or “a Government.” These cells
with “meaning” develop NATURALLY.</div>
<div class="elementToProof">
Your statement "single cell firing having meaning" is not
mathematically meaningful. The brain is a vector of $10^{14}$
dimension. Each neuron corresponds to a dimension. Each neuron
does not have a one-to-one correspondence with a symbol (like
"Asim". Review the definition of one-to-one correspondence.
If you do not mean one-to-one correspondence, your statement is
not mathematically meaningful.</div>
<div class="elementToProof">
Best regards,</div>
<div class="elementToProof">
-John Weng</div>
<hr tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><b>From:</b> Asim Roy
<a class="moz-txt-link-rfc2396E" href="mailto:ASIM.ROY@asu.edu"><ASIM.ROY@asu.edu></a><br>
<b>Sent:</b> Thursday, June 6, 2024 11:21 PM<br>
<b>To:</b> Weng, Juyang <a class="moz-txt-link-rfc2396E" href="mailto:weng@msu.edu"><weng@msu.edu></a>; Stephen José
Hanson <a class="moz-txt-link-rfc2396E" href="mailto:jose@rubic.rutgers.edu"><jose@rubic.rutgers.edu></a>; Gary Marcus
<a class="moz-txt-link-rfc2396E" href="mailto:gary.marcus@nyu.edu"><gary.marcus@nyu.edu></a><br>
<b>Cc:</b> <a class="moz-txt-link-abbreviated" href="mailto:connectionists@mailman.srv.cs.cmu.edu">connectionists@mailman.srv.cs.cmu.edu</a>
<a class="moz-txt-link-rfc2396E" href="mailto:connectionists@mailman.srv.cs.cmu.edu"><connectionists@mailman.srv.cs.cmu.edu></a><br>
<b>Subject:</b> RE: Connectionists: short Op-ed to address AI
problems
<div> </div>
</div>
<div lang="EN-US">
<div class="x_WordSection1">
<p class="x_MsoNormal"><span>Dear John,</span></p>
<p class="x_MsoNormal"><span> </span></p>
<p class="x_MsoNormal"><span>There is no “Asim” or
“Government” in any brain, human or otherwise. That single
cell firing in a cat’s brain having “meaning” is not due
to “Asim” or “a Government.” These cells with “meaning”
develop NATURALLY. And that’s what you are missing in your
Development Network theory. You have not been able to
capture in your systems that side of development. Perhaps
time to go back to the drawing board. Symbols follow
directly from “single cells having meaning.”</span></p>
<p class="x_MsoNormal"><span> </span></p>
<p class="x_xmsonormal"><span>All the best,</span></p>
<p class="x_xmsonormal"><span>Asim Roy</span></p>
<p class="x_xmsonormal"><span>Professor, Information Systems</span></p>
<p class="x_xmsonormal"><span>Arizona State University</span></p>
<p class="x_xmsonormal"><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__lifeboat.com_ex_bios.asim.roy&d=DwMFaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=waSKY67JF57IZXg30ysFB_R7OG9zoQwFwxyps6FbTa1Zh5mttxRot_t4N7mn68Pj&s=oDRJmXX22O8NcfqyLjyu4Ajmt8pcHWquTxYjeWahfuw&e="
target="_blank" moz-do-not-send="true">Lifeboat
Foundation Bios: Professor Asim Roy</a></span></p>
<p class="x_xmsonormal"><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__isearch.asu.edu_profile_9973&d=DwMFaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=waSKY67JF57IZXg30ysFB_R7OG9zoQwFwxyps6FbTa1Zh5mttxRot_t4N7mn68Pj&s=jCesWT7oGgX76_y7PFh4cCIQ-Ife-esGblJyrBiDlro&e="
target="_blank" moz-do-not-send="true">Asim Roy |
iSearch (asu.edu)</a></span></p>
<p class="x_MsoNormal"><span> </span></p>
<p class="x_MsoNormal"><span> </span></p>
<p class="x_MsoNormal"><span> </span></p>
<p class="x_MsoNormal"><span> </span></p>
<div>
<div>
<p class="x_MsoNormal"><b><span>From:</span></b><span>
Weng, Juyang <a class="moz-txt-link-rfc2396E" href="mailto:weng@msu.edu"><weng@msu.edu></a>
<br>
<b>Sent:</b> Thursday, June 6, 2024 8:09 PM<br>
<b>To:</b> Asim Roy <a class="moz-txt-link-rfc2396E" href="mailto:ASIM.ROY@asu.edu"><ASIM.ROY@asu.edu></a>; Stephen
José Hanson <a class="moz-txt-link-rfc2396E" href="mailto:jose@rubic.rutgers.edu"><jose@rubic.rutgers.edu></a>; Gary
Marcus <a class="moz-txt-link-rfc2396E" href="mailto:gary.marcus@nyu.edu"><gary.marcus@nyu.edu></a><br>
<b>Cc:</b> <a class="moz-txt-link-abbreviated" href="mailto:connectionists@mailman.srv.cs.cmu.edu">connectionists@mailman.srv.cs.cmu.edu</a><br>
<b>Subject:</b> Re: Connectionists: short Op-ed to
address AI problems</span></p>
</div>
</div>
<p class="x_MsoNormal"> </p>
<div>
<p class="x_MsoNormal"><span>Dear Asim,</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> You wrote, "Let’s do one
issue at a time. Let’s try symbols first." This
approach misleads you to the wrong track.</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> Case 1: neuron level symbols
(your position).</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> Case 2: area level symbols.</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> Case 3: task level symbols.</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> They are all dead ends
because Asim is the government of the "brain" model.</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> For all those Asim knows,
it is too expensive to create all symbols for the
"brain" model. </span></p>
</div>
<div>
<p class="x_MsoNormal"><span> For all those Asim does not
know, the model does not know either.</span></p>
</div>
<div>
<p class="x_MsoNormal"><span> Deadends! If you continue
this "one issue at a time route," you waste too much
time in your life. This is because the first issue is
wrong to consider. </span></p>
</div>
<div>
<p class="x_MsoNormal"><span> Best regards,</span></p>
</div>
<div>
<p class="x_MsoNormal"><span>-John Weng </span></p>
</div>
<div class="x_MsoNormal">
<hr width="98%">
</div>
<div id="x_divRplyFwdMsg">
<p class="x_MsoNormal"><b><span>From:</span></b><span> Asim
Roy <<a href="mailto:ASIM.ROY@asu.edu"
moz-do-not-send="true" class="moz-txt-link-freetext">ASIM.ROY@asu.edu</a>><br>
<b>Sent:</b> Thursday, June 6, 2024 10:06 PM<br>
<b>To:</b> Weng, Juyang <<a
href="mailto:weng@msu.edu" moz-do-not-send="true"
class="moz-txt-link-freetext">weng@msu.edu</a>>;
Stephen José Hanson <<a
href="mailto:jose@rubic.rutgers.edu"
moz-do-not-send="true" class="moz-txt-link-freetext">jose@rubic.rutgers.edu</a>>;
Gary Marcus <<a href="mailto:gary.marcus@nyu.edu"
moz-do-not-send="true" class="moz-txt-link-freetext">gary.marcus@nyu.edu</a>><br>
<b>Cc:</b> <a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true" class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a>
<<a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true" class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a>><br>
<b>Subject:</b> RE: Connectionists: short Op-ed to
address AI problems</span> </p>
<div>
<p class="x_MsoNormal"> </p>
</div>
</div>
<div>
<div>
<p class="x_xmsonormal"><span>Dear John,</span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span>Let’s do one issue at a
time. Let’s try symbols first. There is plenty of
evidence in neurophysiology that one can associate
“meaning” to the activation of certain individual
cells. As far as I know, all of the brain-related
Nobel prizes were about finding “meaning” in the
activations of certain single neurons. Here I quote
from Wikipedia (<a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Single-unit_recording__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNkNR1ZZ9A$"
moz-do-not-send="true">Single-unit recording -
Wikipedia</a>):</span></p>
<p class="x_xmsonormal"><span> </span></p>
<ul type="disc">
<li class="x_xmsonormal"><span>1928: One of the earliest
accounts of being able to record from the nervous
system was by <a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Edgar_Adrian__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNnQtx1LXQ$"
title="Edgar Adrian" moz-do-not-send="true"><span>Edgar
Adrian</span></a> in his 1928 publication "The
Basis of Sensation". In this, he describes his
recordings of electrical discharges in
<u>single nerve fibers</u> using a <a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Lippmann_electrometer__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNn6lQGgzA$"
title="Lippmann electrometer"
moz-do-not-send="true"><span>Lippmann electrometer</span></a>.
He won the <span>Nobel Prize in 1932</span> for his
work revealing the function of neurons.<sup><a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Single-unit_recording*cite_note-11__;Iw!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNk6I8LNhA$"
moz-do-not-send="true"><span>[11]</span></a></sup></span></li>
<li class="x_xmsonormal"><span>1957: <a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/John_Carew_Eccles__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNkFV1ULMA$"
title="John Carew Eccles" moz-do-not-send="true"><span>John
Eccles</span></a> used intracellular <u>single-unit
recording</u> to study synaptic mechanisms in
motoneurons (for which he won the
<span>Nobel Prize in 1963</span>).</span></li>
<li class="x_xmsonormal"><span>1959: Studies by <a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/David_H._Hubel__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNkJaa_aew$"
title="David H. Hubel" moz-do-not-send="true"><span>David
H. Hubel</span></a> and <a
href="https://urldefense.com/v3/__https:/en.wikipedia.org/wiki/Torsten_Wiesel__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNndrzwVDg$"
title="Torsten Wiesel" moz-do-not-send="true"><span>Torsten
Wiesel</span></a>. They used <u>single neuron
recordings</u> to map the visual cortex in
unanesthesized, unrestrained cats using tungsten
electrodes. This work won them the
<span>Nobel Prize in 1981</span> for information
processing in the visual system.</span></li>
</ul>
<p class="x_xmsonormal"><span> </span></p>
<ul type="disc">
<li class="x_xmsonormal"><span>And the work of Mosers
and O’Keefe on grid and place cells that won them
the Nobel:
<span><a
href="https://urldefense.com/v3/__https:/www.nobelprize.org/prizes/medicine/2014/press-release/__;!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNm67pI7iQ$"
moz-do-not-send="true">The 2014 Nobel Prize in
Physiology or Medicine - Press release</a>. </span>Here’s
a quote about the work on place cells:</span></li>
</ul>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span>“</span><i><span>Most
neuroscientists once doubted that brain activity
could be linked with behaviour, but in the late
1960s, <strong><span>O</span></strong>’Keefe began
to record signals from individual neurons in the
brains of rats moving freely in a box. He put
electrodes in the hippocampus and was surprised to
find that <u>individual cells fired</u> when the
rats moved to particular spots</span></i><span>.”</span><span>
<a
href="https://urldefense.com/v3/__https:/www.nature.com/articles/514153a*:*:text=Most*20neuroscientists*20once*20doubted*20that*20brain*20activity*20could,fired*20when*20the*20rats*20moved*20to*20particular*20spots.__;I34lJSUlJSUlJSUlJSUlJQ!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNnqimHsFw$"
moz-do-not-send="true">
Nobel prize for decoding brain’s sense of place |
Nature</a></span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span>And then the findings about
concept cells (Jennifer Aniston cells), which are
single cell recordings. Here’s from
<a
href="https://urldefense.com/v3/__https:/www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2020.00059/full*B6__;Iw!!HXCxUKc!2pFG0g1tPh-88cfwjJImIxJxtBhaOQ1wWf15ZEUkChi5vUb8q_qEXUDZt7bsQ9QjqzSglNlbw-6x9Q$"
moz-do-not-send="true">
<span>Reddy and Thorpe (2014)</span></a><span>: “</span><span>concept
cells have “<strong><i><u><span>meaning</span></u></i></strong> of
a given stimulus in a manner that is <strong><i><span>invariant</span></i></strong> to
different representations of that stimulus.”</span></span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span>We all try to generalize
from data, right. If you examine these findings, the
most important feature is that they all found
“meaning” in single cell activations. So the most
fundamental question for you is: <span>
Do you accept these findings and the general
conclusion that single cell activations can have
meaning</span>? Again, beware that, beyond winning
Nobel prizes, much work in neuroscience and other
fields follows from these findings.</span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span>All the best,</span></p>
<p class="x_xmsonormal"><span>Asim Roy</span></p>
<p class="x_xmsonormal"><span>Professor, Information
Systems</span></p>
<p class="x_xmsonormal"><span>Arizona State University</span></p>
<p class="x_xmsonormal"><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__lifeboat.com_ex_bios.asim.roy&d=DwMFaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=waSKY67JF57IZXg30ysFB_R7OG9zoQwFwxyps6FbTa1Zh5mttxRot_t4N7mn68Pj&s=oDRJmXX22O8NcfqyLjyu4Ajmt8pcHWquTxYjeWahfuw&e="
target="_blank" moz-do-not-send="true">Lifeboat
Foundation Bios: Professor Asim Roy</a></span></p>
<p class="x_xmsonormal"><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__isearch.asu.edu_profile_9973&d=DwMFaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=waSKY67JF57IZXg30ysFB_R7OG9zoQwFwxyps6FbTa1Zh5mttxRot_t4N7mn68Pj&s=jCesWT7oGgX76_y7PFh4cCIQ-Ife-esGblJyrBiDlro&e="
target="_blank" moz-do-not-send="true">Asim Roy |
iSearch (asu.edu)</a></span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span> </span></p>
<p class="x_xmsonormal"><span> </span></p>
<div>
<div>
<p class="x_xmsonormal"><b><span>From:</span></b><span>
Weng, Juyang <<a href="mailto:weng@msu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">weng@msu.edu</a>>
<br>
<b>Sent:</b> Thursday, June 6, 2024 1:09 AM<br>
<b>To:</b> Asim Roy <<a
href="mailto:ASIM.ROY@asu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">ASIM.ROY@asu.edu</a>>;
Stephen José Hanson <<a
href="mailto:jose@rubic.rutgers.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">jose@rubic.rutgers.edu</a>>;
Gary Marcus <<a
href="mailto:gary.marcus@nyu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">gary.marcus@nyu.edu</a>>;
Weng, Juyang <<a href="mailto:weng@msu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">weng@msu.edu</a>><br>
<b>Cc:</b> <a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a><br>
<b>Subject:</b> Re: Connectionists: short Op-ed to
address AI problems</span></p>
</div>
</div>
<p class="x_xmsonormal"> </p>
<div>
<p class="x_xmsonormal"><span>Dear Asim,</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> You wrote, "We are
doing neurosymbolic with image processing – the
symbolic stuff on top of a DL model. It also brings
in the explanation side." </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Not only DL is
misconduct, but symbols are another devil. </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> In my IJCNN 2022
paper, </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> <a
href="https://urldefense.com/v3/__http:/www.cse.msu.edu/*weng/research/20M-IJCNN2022rvsd-cite.pdf__;fg!!IKRxdwAv5BmarQ!YZcFaLmNraAEJLpxRQGKzKZTVt_nn3J9i52_xG7zhEgKn6ZASf_q59sOFVdSPylt7_NueMymM_EI7GNl$"
moz-do-not-send="true">http://www.cse.msu.edu/~weng/research/20M-IJCNN2022rvsd-cite.pdf</a></span></p>
</div>
<div>
<p class="x_xmsonormal"><span> I proved "symbol-free"
as one of the 20 million-dollar problems for us to
understand human brains.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> The definition of
symbols requires a government, but government-free
is one of the 20 million-dollar problems for us to
understand human brains.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Let us consider three
cases:</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Case 1: If a human
designs symbols within a network (e.g., LSTM) and
assigns the symbols to some individual neurons
(e.g., task-specific gates) of the network, this
human is a government within the network since he is
task-aware. </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Case 2: If a human
designs symbols within a network and assigns roles
to blocks in a functional block diagram, e.g.,
[Starzyk10], this human is a government within the
network. </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Case 3: In the symbolic
AI school, a human programmer designs symbolic
representations for a task that is assigned to a
computer program or network. This human is a
government within the symbolic AI system since he is
task-aware. </span></p>
</div>
<div>
<p class="x_xmsonormal"><span> All the 3 cases do not
solve the government-free problem.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> I have attached an image
that further explains the symbol problem in the same
paper.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Let me know if you still
do not agree that the brain must be free from
symbols after you read the entire paper.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> By the way, I am
surprised that as a mathematician, you still do not
understand the Post-Selection misconduct in DL that
I raised to you earlier. Please use your own words
to explain Post-Selection and why you can handle
explanation using Post-Selection misconduct.</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> Best regards,</span></p>
</div>
<div>
<p class="x_xmsonormal"><span>-John Weng</span></p>
</div>
<div>
<p class="x_xmsonormal"><span> </span></p>
</div>
<div class="x_MsoNormal">
<hr width="98%">
</div>
<div id="x_x_divRplyFwdMsg">
<p class="x_xmsonormal"><b><span>From:</span></b><span> Connectionists
<<a
href="mailto:connectionists-bounces@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists-bounces@mailman.srv.cs.cmu.edu</a>>
on behalf of Asim Roy <<a
href="mailto:ASIM.ROY@asu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">ASIM.ROY@asu.edu</a>><br>
<b>Sent:</b> Wednesday, June 5, 2024 6:49 PM<br>
<b>To:</b> Stephen José Hanson <<a
href="mailto:jose@rubic.rutgers.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">jose@rubic.rutgers.edu</a>>;
Gary Marcus <<a href="mailto:gary.marcus@nyu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">gary.marcus@nyu.edu</a>><br>
<b>Cc:</b> <a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a>
<<a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a>><br>
<b>Subject:</b> Re: Connectionists: short Op-ed to
address AI problems</span> </p>
<div>
<p class="x_xmsonormal"> </p>
</div>
</div>
<div>
<p><span>Dear Stephen,</span></p>
<p><span> </span></p>
<p><span>We are doing neurosymbolic with image
processing – the symbolic stuff on top of a DL
model. It also brings in the explanation side. The
results are astounding. We get better performance
than a pure DL model. And exploring applications
with defense agencies. They are impressed with the
results we have so far. So, neurosymbolic is
definitely the way forward.</span></p>
<p><span> </span></p>
<p><span>Best,</span></p>
<p><span>Asim Roy</span></p>
<p><span>Professor, Information Systems</span></p>
<p><span>Arizona State University</span></p>
<p><u><span><a
href="https://urldefense.com/v3/__https:/search.asu.edu/profile/9973__;!!HXCxUKc!1ZzSj6Uim5wWu5W-JiBNqp_Cig3tUkK5DgMhDEBYnERP1f-pOAReghJiHzEk3hEHKL31roB_8qivsA$"
moz-do-not-send="true">Asim Roy | ASU Search</a></span></u></p>
<p><u><span><a
href="https://urldefense.com/v3/__https:/lifeboat.com/ex/bios.asim.roy__;!!HXCxUKc!1ZzSj6Uim5wWu5W-JiBNqp_Cig3tUkK5DgMhDEBYnERP1f-pOAReghJiHzEk3hEHKL31roAnhJU86A$"
moz-do-not-send="true">Lifeboat Foundation Bios:
Professor Asim Roy</a></span></u></p>
<p><span> </span></p>
<div>
<p><b><span>From:</span></b><span> Connectionists <<a
href="mailto:connectionists-bounces@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists-bounces@mailman.srv.cs.cmu.edu</a>>
<b>On Behalf Of </b>Stephen José Hanson<br>
<b>Sent:</b> Wednesday, June 5, 2024 6:06 AM<br>
<b>To:</b> Gary Marcus <<a
href="mailto:gary.marcus@nyu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">gary.marcus@nyu.edu</a>><br>
<b>Cc:</b> <a
href="mailto:connectionists@mailman.srv.cs.cmu.edu"
moz-do-not-send="true"
class="moz-txt-link-freetext">connectionists@mailman.srv.cs.cmu.edu</a><br>
<b>Subject:</b> Re: Connectionists: short Op-ed to
address AI problems</span></p>
</div>
<p><span> </span></p>
<p><span>Dear Flabbergasted:</span></p>
<p><span>Thankyou, I endeavor to provide short but
useful commentary that could be considered a "work
of art". Graci!</span></p>
<p><span>Now either my memory is failing since 2017(not
impossible), or you are smoothing over a time series
of claims that are actually like a seesaw.</span></p>
<p><span>I think if we just rewind some of the
connectionist comments; it would be clear, in fact,
for example, you had a long series of comments with
Geoff that seemed to indicate you were being
misreprented as well. Your complaints have always
be around the fact that DL-AI has false alarms (and
to be fair other problems) And sometimes pretty
odd-ones. LLMs human and non-human errors are even
more interesting. The fact that they seem to grow
circuits in the attention-heads is gobsmacking! I
thought then and think now you are complaining about
peas under a very thick mattress (oh-oh, metaphors
now- I may have opened pandora's box.)</span></p>
<p><span>But I will go look at the budding NeuroSymbolic
paper you mentioned, but I have my doubts that the
statistical bias is equivalent with the
architectually simplistic LLMs. Nonetheless, I have
not read it.</span></p>
<p><span>I will also make a coarse timeline of your
comments since 2017, but anyone out there that would
like to help, greatly appreciated!</span></p>
<p><span>Best,</span></p>
<p><span>Stephen</span></p>
<p><span>On 6/5/24 8:41 AM, Gary Marcus wrote:</span></p>
<blockquote>
<p><span>Wow, Stephen, you have outdone yourself. This
note is a startling mixture of rude,
condescending, inaccurate, and uninformed. A work
of art! </span></p>
<p><span> </span></p>
<p><span>To correct four misunderstandings:</span></p>
<p><span>1. Yes, my essay was written before LLMs were
popular (though around the time Transformers were
proposed as it happens). It was however
<i>precisely</i> “ a moonshot idea, that doesn't
involve leaving the blackbox in the hands of
corporate types who value profits over knowledge.”
Please read what I wrote. It’s one page, linked
below, and you obviously couldn’t be bothered,.
(Parenthetically, I was one of the first people to
warn that OpenAI was likely to be problematic,
and have done so repeatedly at my Substack.)</span></p>
<p><span>2. My argument throughout (back to 2012, in
the New Yorker, 2018 in my Deep Learning: A
Critical Appraisal, etc) has been that deep
learning has some role but cannot solve all
things, and that it would be not reliable on its
own. In 2019 onwards I emphasized many of the
social problems that arise from relying on such
unreliable architectures. I have never wavered
from any of that. (Again, please read my work
before so grossly distorting it.) Unreliable
systems that are blind to truth and values can
cause harm (bias), be exploited (to create
disinformation), etc. There is absolutely no
contradiction there, as I have explained numerous
times in my writings.</span></p>
<p><span>3. It’s truly rude to dismiss an entire field
as “flotsam and jetsam”, and you obviously aren’t
following the neurosymbolic literature, e.g., you
must have missed DeepMind’s neurosymbolic
AlphaGeometry paper, in Nature, with its state of
the art results, beating pure neural nets.</span></p>
<p><span>4. Again, nothing has changed about my view;
your last remark is gratuitous and based on a
misunderstanding.</span></p>
<p><span> </span></p>
<p><span>Truly flabbergasted,</span></p>
<p><span>Gary</span></p>
<p><span> </span></p>
<blockquote>
<p><span>On Jun 5, 2024, at 05:18, Stephen José
Hanson
</span><u><span><a
href="mailto:jose@rubic.rutgers.edu"
moz-do-not-send="true"><jose@rubic.rutgers.edu></a></span></u><span> wrote:</span></p>
</blockquote>
<blockquote>
<p><span></span></p>
<p><span>Gary, this was before the LLM discovery.
Pierre is proposing a moonshot idea, that
doesn't involve leaving the blackbox in the
hands of corporate types who value profits over
knowledge. OPENAI seems to be flailing and
having serious safety and security issues. It
certainly could be recipe for diaster.</span></p>
<p><span>Frankly your views have been all over the
place. DL doesn't work, DL could work but
should be merged with the useless flotsam and
jetsam from GOFAI over the last 50 years, and
now they are too dangerous because they work but
they are unreliable, like most humans.</span></p>
<p><span>Its hard to know what views of yours to
take seriously as they seem change so rapidly. </span></p>
<p><span>Cheers</span></p>
<p><span>Stephen</span></p>
<p><span>On 6/4/24 9:53 AM, Gary Marcus wrote:</span></p>
<blockquote>
<p><span>I would just point out that I first made
this suggestion [CERN for AI] in the New York
Times in 2017, and several others have since.
There is some effort ongoing to try to make it
happen, if you search you will see.</span></p>
<p><span> </span></p>
<table class="x_MsoNormalTable" width="300">
<tbody>
<tr>
<td>
<p><span><30gray-facebookJumbo.jpg></span></p>
</td>
</tr>
<tr>
<td>
<table class="x_MsoNormalTable"
width="300">
<tbody>
<tr>
<td>
<div>
<p><u><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__www.nytimes.com_2017_07_29_opinion_sunday_artificial-2Dintelligence-2Dis-2Dstuck-2Dheres-2Dhow-2Dto-2Dmove-2Dit-2Dforward.html-3Funlocked-5Farticle-5Fcode-3D1.xE0.mcIz.lT-5FK7BZdonGJ-26smid-3Dnytcore-2Dios-2Dshare-26referringSource-3DarticleShare-26u2g-3Di-26sgrp-3Dc-2Dcb&d=DwMGaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=fwBsbQ5xjEJFDg3c0iXuOBcr84mxEGxR0cEG4-hstVM8dJNyq3HVvpCACElUGWT2&s=TGZDkK1TsB_rNyjmal5jG1694upjB2JDhtj3UOe4Cws&e="
moz-do-not-send="true"><span>Opinion
| Artificial
Intelligence Is Stuck.
Here’s How to Move It
Forward. (Gift
Article)</span></a></span></u></p>
<p><u><span><a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__www.nytimes.com_2017_07_29_opinion_sunday_artificial-2Dintelligence-2Dis-2Dstuck-2Dheres-2Dhow-2Dto-2Dmove-2Dit-2Dforward.html-3Funlocked-5Farticle-5Fcode-3D1.xE0.mcIz.lT-5FK7BZdonGJ-26smid-3Dnytcore-2Dios-2Dshare-26referringSource-3DarticleShare-26u2g-3Di-26sgrp-3Dc-2Dcb&d=DwMGaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=wQR1NePCSj6dOGDD0r6B5Kn1fcNaTMg7tARe7TdEDqQ&m=fwBsbQ5xjEJFDg3c0iXuOBcr84mxEGxR0cEG4-hstVM8dJNyq3HVvpCACElUGWT2&s=TGZDkK1TsB_rNyjmal5jG1694upjB2JDhtj3UOe4Cws&e="
moz-do-not-send="true"><span>nytimes.com</span></a></span></u></p>
</div>
</td>
</tr>
</tbody>
</table>
</td>
</tr>
</tbody>
</table>
<p><span> </span></p>
<p><span> </span></p>
<blockquote>
<p><span>On Jun 3, 2024, at 22:58, Baldi,Pierre
</span><u><span><a
href="mailto:pfbaldi@ics.uci.edu"
moz-do-not-send="true"><pfbaldi@ics.uci.edu></a></span></u><span> wrote:</span></p>
</blockquote>
<blockquote>
<p><span></span><span><br>
I would appreciate feedback from this
group,especially dissenting feedback, on
the attached Op-ed. You can send it to my
personal email which you can find on my
university web site if you prefer. The basic
idea is simple:<br>
<br>
IF for scientific, security, or other
societal reasons we want academics to
develop and study the most advanced forms of
AI, I can see only one solution: create a
national or international effort around the
largest data/computing center on Earth with
a CERN-like structure comprising permanent
staff, and 1000s of affiliated academic
laboratories. There are many obstacles, but
none is completely insurmountable if we
wanted to.<br>
<br>
Thank you.<br>
<br>
Pierre<br>
<br>
<br>
</span></p>
<p><span><AI-CERN-Baldi2024FF.pdf></span></p>
</blockquote>
</blockquote>
<div>
<pre><span>-- </span></pre>
</div>
<div>
<pre><span>Stephen José Hanson</span></pre>
</div>
<div>
<pre><span>Professor of Psychology</span></pre>
</div>
<div>
<pre><span>Director of RUBIC</span></pre>
</div>
<div>
<pre><span>Member of Exc Comm RUCCS</span></pre>
</div>
</blockquote>
</blockquote>
<div>
<pre><span>-- </span></pre>
</div>
<div>
<pre><span>Stephen José Hanson</span></pre>
</div>
<div>
<pre><span>Professor of Psychology</span></pre>
</div>
<div>
<pre><span>Director of RUBIC</span></pre>
</div>
<div>
<pre><span>Member of Exc Comm RUCCS</span></pre>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>