<!DOCTYPE html>
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body>
<p><a
href="https://www.psychologytoday.com/us/blog/the-digital-self/202505/llms-arent-mirrors-theyre-holograms"
class="moz-txt-link-freetext">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; -webkit-font-smoothing: antialiased; 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; overflow-wrap: break-word; word-break: break-word; hyphens: auto; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; white-space: normal; text-decoration-thickness: initial; text-decoration-style: initial; text-decoration-color: initial;">Holograms
Don’t Store Images, They Store Possibility</h2>
<p
style="box-sizing: border-box; -webkit-font-smoothing: antialiased; font-family: "Proxima Nova Regular", Arial, sans-serif; font-size: 18px; line-height: 1.6; margin-top: 0px; margin-bottom: 24px; overflow-wrap: break-word; word-break: break-word; hyphens: auto; color: rgb(44, 45, 48); font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; white-space: normal; text-decoration-thickness: initial; text-decoration-style: initial; text-decoration-color: initial;">A<span> </span><a
href="https://science.howstuffworks.com/hologram.htm"
style="box-sizing: border-box; -webkit-font-smoothing: antialiased; font-family: "Proxima Nova Semi Bold", Arial, sans-serif; color: rgb(71, 123, 228); text-decoration: none; word-break: break-word; overflow-wrap: break-word; hyphens: auto;">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; -webkit-font-smoothing: antialiased; font-family: "Proxima Nova Regular", Arial, sans-serif; font-size: 18px; line-height: 1.6; margin-top: 0px; margin-bottom: 24px; overflow-wrap: break-word; word-break: break-word; hyphens: auto; color: rgb(44, 45, 48); font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; white-space: normal; text-decoration-thickness: initial; 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>
</body>
</html>