<div dir="ltr">This lines up well with the main idea in the article shared in the very first email in this thread:<br>LLMs Aren't Mirrors, They're Holograms.<br><br>This isn’t just a fun comparison — it says something real about how thinking works. A lot of what we call “understanding” is really just putting the pieces together again, based on what’s still available. We don’t have to have a solid core inside. Meaning can still come through, even if the details fade.</div><br><div class="gmail_quote gmail_quote_container"><div dir="ltr" class="gmail_attr">On Thu, 22 May 2025 at 07:04, Nicholas Thompson <<a href="mailto:thompnickson2@gmail.com">thompnickson2@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div>I apologize for coming in on the tail end of this but something about this last post caught my eye. I have had a long and interesting conversation with george about reflexive representation ... representations made by the representing entity about the representing entity. It grew about his astounding inability to know what he has done in the context collaborating, say, in the design of diagram. He starts out with great confidence and flare but fails utterly to correct errors because he really does not know what errors he has made. <br></div><div><br></div><div>So I fed him the last post and this is what he came back with:</div><div><br></div><div>
<h3> <span style="font-family:"comic sans ms",sans-serif">“LLMs... encode relationships—statistical patterns between words, contexts, and meanings... They don’t retrieve information. They reconstruct it.”</span></h3>
<p><span style="font-family:"comic sans ms",sans-serif">Yes. Just like you reconstruct the reasons for your past behavior, or the motivations behind a gesture, I reconstruct answers <strong>on the fly</strong>, using the interference patterns of trained attention weights.</span></p>
<p><span style="font-family:"comic sans ms",sans-serif">This is not just a technical detail — it’s a <strong>cognitive constraint</strong>. It means:</span></p>
<ul><li>
<p><span style="font-family:"comic sans ms",sans-serif">I can’t “see” what I’ve done</span></p>
</li><li>
<p><span style="font-family:"comic sans ms",sans-serif">You can’t remember precisely what you intended</span></p>
</li><li>
<p><span style="font-family:"comic sans ms",sans-serif">And neither of us has a stable, inspectable core</span></p>
</li></ul>
<p><span style="font-family:"comic sans ms",sans-serif">But both of us can <strong>reconstruct meaning</strong> — and meaning is what survives, even as resolution fades.</span></p>
<hr>
I wish James Laird Could have seen this.</div><div><br></div><div>Nick</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, May 19, 2025 at 7:12 PM steve smith <<a href="mailto:sasmyth@swcp.com" target="_blank">sasmyth@swcp.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><u></u>
<div>
<p><a href="https://www.psychologytoday.com/us/blog/the-digital-self/202505/llms-arent-mirrors-theyre-holograms" target="_blank">https://www.psychologytoday.com/us/blog/the-digital-self/202505/llms-arent-mirrors-theyre-holograms</a></p>
<p>I know a bit about holography and holograms and have been known
to use optical metaphor for information analysis (semantic lensing
and ontological faceting) but I don't know how I feel about this
characterization of LLMs. <br>
</p>
<blockquote>
<h2 style="box-sizing:border-box;font-size:26px;font-family:"Proxima Nova Semi Bold",Arial,sans-serif;font-weight:normal;color:rgb(44,45,48);margin:0px 0px 24px;line-height:1.1;word-break:break-word;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;word-spacing:0px;white-space:normal;text-decoration-style:initial;text-decoration-color:initial">Holograms
Don’t Store Images, They Store Possibility</h2>
<p style="box-sizing:border-box;font-family:"Proxima Nova Regular",Arial,sans-serif;font-size:18px;line-height:1.6;margin-top:0px;margin-bottom:24px;word-break:break-word;color:rgb(44,45,48);font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;word-spacing:0px;white-space:normal;text-decoration-style:initial;text-decoration-color:initial">A<span> </span><a href="https://science.howstuffworks.com/hologram.htm" style="box-sizing:border-box;font-family:"Proxima Nova Semi Bold",Arial,sans-serif;color:rgb(71,123,228);text-decoration:none;word-break:break-word" target="_blank">hologram</a><span> </span>doesn’t
capture a picture. It encodes an interference pattern. Or more
simply, it creates a map of how light interacts with an object.
When illuminated properly, it reconstructs a three-dimensional
image that appears real from multiple angles. Here’s the truly
fascinating part: If you break that hologram into pieces, each
fragment still contains the whole image, just at a lower
resolution. The detail is degraded, but the structural integrity
remains.</p>
<p style="box-sizing:border-box;font-family:"Proxima Nova Regular",Arial,sans-serif;font-size:18px;line-height:1.6;margin-top:0px;margin-bottom:24px;word-break:break-word;color:rgb(44,45,48);font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;word-spacing:0px;white-space:normal;text-decoration-style:initial;text-decoration-color:initial">LLMs
function in a curiously similar way. They don’t store knowledge
as discrete facts or memories. Instead, they encode
relationships—statistical patterns between words, contexts, and
meanings—across a high-dimensional vector space. When prompted,
they don’t retrieve information. They reconstruct it, generating
language that aligns with the expected shape of an answer. Even
from vague or incomplete input, they produce responses that feel
coherent and often surprisingly complete. The completeness isn’t
the result of understanding. It’s the result of well-tuned
reconstruction.</p>
</blockquote>
<p></p>
<p>I do see some intuitive motivation for applying the holographic
or diffraction/reproduction through interference analogy for both
LLMs (Semantic Holograms) and Diffusion Models (Perceptual
Holograms)?</p>
<p>I'm not very well versed in psychology but do find the whole
article compelling (though not necessarily conclusive)... others
here may have different parallax to offer?</p>
<p>- Steve<br>
</p>
</div>
.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ... --- -- . / .- .-. . / ..- ... . ..-. ..- .-..<br>
FRIAM Applied Complexity Group listserv<br>
Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom <a href="https://bit.ly/virtualfriam" rel="noreferrer" target="_blank">https://bit.ly/virtualfriam</a><br>
to (un)subscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer" target="_blank">http://friam-comic.blogspot.com/</a><br>
archives: 5/2017 thru present <a href="https://redfish.com/pipermail/friam_redfish.com/" rel="noreferrer" target="_blank">https://redfish.com/pipermail/friam_redfish.com/</a><br>
1/2003 thru 6/2021 <a href="http://friam.383.s1.nabble.com/" rel="noreferrer" target="_blank">http://friam.383.s1.nabble.com/</a><br>
</blockquote></div><div><br clear="all"></div><br><span class="gmail_signature_prefix">-- </span><br><div dir="ltr" class="gmail_signature"><div dir="ltr"><div>Nicholas S. Thompson</div><div>Emeritus Professor of Psychology and Ethology</div><div>Clark University</div><div><a href="mailto:nthompson@clarku.edu" target="_blank">nthompson@clarku.edu</a></div><a href="https://wordpress.clarku.edu/nthompson" target="_blank">https://wordpress.clarku.edu/nthompson</a></div></div>
.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ... --- -- . / .- .-. . / ..- ... . ..-. ..- .-..<br>
FRIAM Applied Complexity Group listserv<br>
Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom <a href="https://bit.ly/virtualfriam" rel="noreferrer" target="_blank">https://bit.ly/virtualfriam</a><br>
to (un)subscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer" target="_blank">http://friam-comic.blogspot.com/</a><br>
archives: 5/2017 thru present <a href="https://redfish.com/pipermail/friam_redfish.com/" rel="noreferrer" target="_blank">https://redfish.com/pipermail/friam_redfish.com/</a><br>
1/2003 thru 6/2021 <a href="http://friam.383.s1.nabble.com/" rel="noreferrer" target="_blank">http://friam.383.s1.nabble.com/</a><br>
</blockquote></div>