commit

Author: tecunningham

_freeze/posts/2025-09-19-transformative-AI-notes/execute-results/html.json

@@ -254,54 +254,54 @@ <h3 class="no-anchor listing-title">
 </a>
 </div>
 </div>
-<div class="quarto-post image-right" data-index="1" data-listing-date-sort="1759388400000" data-listing-file-modified-sort="1759438659676" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="43" data-listing-word-count-sort="8406">
+<div class="quarto-post image-right" data-index="1" data-listing-date-sort="1759474800000" data-listing-file-modified-sort="1735428686821" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="10" data-listing-word-count-sort="1848">
 <div class="thumbnail">
-<p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a></p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">
-</a><p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a></p>
+<p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a></p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">
+</a><p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a></p>
 </div>
 <div class="body">
 <h3 class="no-anchor listing-title">
-<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">Economics and AI</a>
+<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">Too Much Good News is Bad News</a>
 </h3>
 <div class="listing-subtitle">
-<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a>
+<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a>
 </div>
 <div class="listing-description">
-<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a>
+<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a>
 </div>
 </div>
 <div class="metadata">
-<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">
+<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">
 <div class="listing-date">
-Oct 2, 2025
+Oct 3, 2025
+</div>
+<div class="listing-author">
+Tom Cunningham
 </div>
 </a>
 </div>
 </div>
-<div class="quarto-post image-right" data-index="2" data-listing-date-sort="1759388400000" data-listing-file-modified-sort="1735428686821" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="10" data-listing-word-count-sort="1848">
+<div class="quarto-post image-right" data-index="2" data-listing-date-sort="1759388400000" data-listing-file-modified-sort="1759505769965" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="43" data-listing-word-count-sort="8405">
 <div class="thumbnail">
-<p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a></p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">
-</a><p><a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a></p>
+<p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a></p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">
+</a><p><a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a></p>
 </div>
 <div class="body">
 <h3 class="no-anchor listing-title">
-<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">Too Much Good News is Bad News</a>
+<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">Economics and AI</a>
 </h3>
 <div class="listing-subtitle">
-<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a>
+<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a>
 </div>
 <div class="listing-description">
-<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external"></a>
+<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external"></a>
 </div>
 </div>
 <div class="metadata">
-<a href="./posts/2024-12-26-heavy-tailed-noise.html" class="no-external">
+<a href="./posts/2025-09-19-transformative-AI-notes.html" class="no-external">
 <div class="listing-date">
 Oct 2, 2025
 </div>
-<div class="listing-author">
-Tom Cunningham
-</div>
 </a>
 </div>
 </div>

docs/index.html

@@ -3,8 +3,8 @@
     "listing": "/index.html",
     "items": [
       "/posts/2024-10-27-from-citlali.html",
-      "/posts/2025-09-19-transformative-AI-notes.html",
       "/posts/2024-12-26-heavy-tailed-noise.html",
+      "/posts/2025-09-19-transformative-AI-notes.html",
       "/posts/2020-10-02-on-deriving-things.html",
       "/posts/2024-05-10-premature-optimization.html",
       "/posts/2023-01-23-peer-effects-norms-culture-sin-taxes.html",

docs/index.xml

@@ -452,7 +452,7 @@ <h1>We don’t have a standard model of AI</h1>
 </dd>
 <dt>A pocket model: LLMs share knowledge.</dt>
 <dd>
-<p>Here is a simple mental model that I often use: <em>LLMs share knowledge</em>. The model is unsatisfactory in many respects but has the virtues of being very simple and very general. Consider an LLM as just a database of answers to questions, containing the set of answers that already exist in the public domain (i.e., in the LLM’s training set).<a href="#fn5" class="footnote-ref" id="fnref5" role="doc-noteref"><sup>5</sup></a> LLMs therefore lower the cost of access to existing knowledge, and people will consult an LLM when they encounter a problem for which (i) they do know the answer, but (ii) they expect that someone else does have know the answer (and the answer was included in the training set).</p>
+<p>Here is a simple mental model that I often use: <em>LLMs share knowledge</em>. The model is unsatisfactory in many respects but has the virtues of being very simple and very general. Consider an LLM as just a database of answers to questions, containing the set of answers that already exist in the public domain (i.e., in the LLM’s training set).<a href="#fn5" class="footnote-ref" id="fnref5" role="doc-noteref"><sup>5</sup></a> LLMs therefore lower the cost of access to existing knowledge, and people will consult an LLM when they encounter a problem for which (i) they do know the answer, but (ii) they expect that someone else does know the answer (and the answer was included in the training set).</p>
 <p>This is a very crude model of an LLM but I think it gives a reasonable characterization of their adoption and effect so far. Around 1/3 of adults in rich countries are regularly using chatbots, and I think it’s fair to say the majority of the use is solving problems outside the domain of the user’s own expertise, but inside someone else’s expertise (see our <a href="https://www.nber.org/papers/w34255">ChatGPT paper</a>). This knowledge-sharing model predicts that LLMs will flatten comparative advantage, so we should see more home production (people solve their own problems), less trade, and lower returns to experience.</p>
 <p>The model has a number of imperfections as a general model of AI: (1) LLMs are often used to do tasks that don’t require knowledge outside the user’s domain, e.g.&nbsp;solving a problem that requires time and patience but not knowledge such as certain types of computer programming, writing, or creating images; (2) the model treats LLMs as strictly bound by the limits of human knowledge, this was a good approximation for early LLMs but it’s clear that AI is progressively expanding the boundary of human knowledge in a variety of ways.</p>
 <p>This model is related to the Garicano-Ide-Talamas models in which an AI shares existing knowledge.</p>
@@ -740,7 +740,16 @@ <h1>AI scientists will be unlike human scientists</h1>
 <div id="ref-varian2011economic" class="csl-entry" role="listitem">
 Varian, Hal. 2011. <span>“Economic Value of Google.”</span> <a href="https://dl.icdst.org/pdfs/files1/f87de5ba3c43760ebcbc2a1d90950dbc.pdf">https://dl.icdst.org/pdfs/files1/f87de5ba3c43760ebcbc2a1d90950dbc.pdf</a>.
 </div>
-</div></section></div></main> <!-- /main -->
+</div></section><section class="quarto-appendix-contents" id="quarto-citation"><h2 class="anchored quarto-appendix-heading">Citation</h2><div><div class="quarto-appendix-secondary-label">BibTeX citation:</div><pre class="sourceCode code-with-copy quarto-appendix-bibtex"><code class="sourceCode bibtex">@online{2025,
+  author = {},
+  title = {Economics and {AI}},
+  date = {2025-10-02},
+  url = {tecunningham.github.io/posts/2025-09-19-transformative-AI-notes.html},
+  langid = {en}
+}
+</code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre><div class="quarto-appendix-secondary-label">For attribution, please cite this work as:</div><div id="ref-2025" class="csl-entry quarto-appendix-citeas" role="listitem">
+<span>“Economics and AI.”</span> 2025. October 2, 2025. <a href="https://tecunningham.github.io/posts/2025-09-19-transformative-AI-notes.html">tecunningham.github.io/posts/2025-09-19-transformative-AI-notes.html</a>.
+</div></div></section></div></main> <!-- /main -->
 <script id="quarto-html-after-body" type="application/javascript">
 window.document.addEventListener("DOMContentLoaded", function (event) {
   const toggleBodyColorMode = (bsSheetEl) => {
@@ -1162,7 +1171,7 @@ <h1>AI scientists will be unlike human scientists</h1>
 });
 </script>
 </div> <!-- /content -->
-<script>var lightboxQuarto = GLightbox({"closeEffect":"zoom","loop":false,"selector":".lightbox","descPosition":"bottom","openEffect":"zoom"});
+<script>var lightboxQuarto = GLightbox({"closeEffect":"zoom","selector":".lightbox","loop":false,"openEffect":"zoom","descPosition":"bottom"});
 (function() {
   let previousOnload = window.onload;
   window.onload = () => {

docs/listings.json

@@ -110,6 +110,6 @@
   </url>
   <url>
     <loc>tecunningham.github.io/posts/2025-09-19-transformative-AI-notes.html</loc>
-    <lastmod>2025-10-02T20:57:39.676Z</lastmod>
+    <lastmod>2025-10-03T15:36:09.965Z</lastmod>
   </url>
 </urlset>

docs/posts/2025-09-19-transformative-AI-notes.html

@@ -1,5 +1,6 @@
 ---
 title: Economics and AI
+citation: true
 bibliography: ai.bib
 reference-location: section
 citation-location: document
@@ -266,7 +267,7 @@ Structural models of AI.
 
 A pocket model: LLMs share knowledge.
 
-: Here is a simple mental model that I often use: *LLMs share knowledge*. The model is unsatisfactory in many respects but has the virtues of being very simple and very general. Consider an LLM as just a database of answers to questions, containing the set of answers that already exist in the public domain (i.e., in the LLM's training set).[^blog] LLMs therefore lower the cost of access to existing knowledge, and people will consult an LLM when they encounter a problem for which (i) they do know the answer, but (ii) they expect that someone else does have know the answer (and the answer was included in the training set).
+: Here is a simple mental model that I often use: *LLMs share knowledge*. The model is unsatisfactory in many respects but has the virtues of being very simple and very general. Consider an LLM as just a database of answers to questions, containing the set of answers that already exist in the public domain (i.e., in the LLM's training set).[^blog] LLMs therefore lower the cost of access to existing knowledge, and people will consult an LLM when they encounter a problem for which (i) they do know the answer, but (ii) they expect that someone else does know the answer (and the answer was included in the training set).
 
       This is a very crude model of an LLM but I think it gives a reasonable characterization of their adoption and effect so far. Around 1/3 of adults in rich countries are regularly using chatbots, and I think it's fair to say the majority of the use is solving problems outside the domain of the user's own expertise, but inside someone else's expertise (see our [ChatGPT paper](https://www.nber.org/papers/w34255)). This knowledge-sharing model predicts that LLMs will flatten comparative advantage, so we should see more home production (people solve their own problems), less trade, and lower returns to experience.