You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@@ -146,6 +146,8 @@ Here the expected wait time is uniformly decreasing in <img src="https://latex.c
146
146
147
147
148
148
149
+
<p>==SUPERCEDED==</p>
150
+
<p>==MOVED TO .TEX==</p>
149
151
<p>This paper develops a simple model of human and AI ability to answer questions. Each question <img src="https://latex.codecogs.com/png.latex?%5Cbm%7Bq%7D"> is a high-dimensional vector, with a true scalar answer <img src="https://latex.codecogs.com/png.latex?a">. An agent’s estimate of the answer is an interpolation based on previously-seen questions and answers <img src="https://latex.codecogs.com/png.latex?(%5Cbm%7Bq%7D%5Ei,a%5Ei)_%7Bi=1,%5Cldots,n%7D">. This framework extends an <a href="https://tecunningham.github.io/posts/2023-09-05-model-of-ai-imitation.html">earlier model</a> developed for a different purpose.</p>
150
152
<p>The model yields several implications:</p>
151
153
<ol type="1">
@@ -219,7 +221,7 @@ Here the expected wait time is uniformly decreasing in <img src="https://latex.c
219
221
</tr>
220
222
</tbody>
221
223
</table>
222
-
<p>We can make some conjectures about adoption by task: :</p>
224
+
<p>We can make some conjectures about adoption by task:</p>
223
225
<table class="caption-top table">
224
226
<colgroup>
225
227
<col style="width: 42%">
@@ -327,14 +329,14 @@ Here the expected wait time is uniformly decreasing in <img src="https://latex.c
<li><ahref="#agrawal-et-al.-2018-exploring-the-impact-of-artificial-intelligence-prediction-versus-judgment" id="toc-agrawal-et-al.-2018-exploring-the-impact-of-artificial-intelligence-prediction-versus-judgment" class="nav-link" data-scroll-target="#agrawal-et-al.-2018-exploring-the-impact-of-artificial-intelligence-prediction-versus-judgment">Agrawal et al. (2018) “Exploring the Impact of Artificial Intelligence: Prediction versus Judgment”</a></li>
250
-
<li><ahref="#kleinberg-et-al.-2017-human-decisions-and-machine-predictions" id="toc-kleinberg-et-al.-2017-human-decisions-and-machine-predictions" class="nav-link" data-scroll-target="#kleinberg-et-al.-2017-human-decisions-and-machine-predictions">Kleinberg et al. (2017) “Human Decisions and Machine Predictions”</a></li>
<p>This paper develops a simple model of human and AI ability to answer questions. Each question <spanclass="math inline">\(\bm{q}\)</span> is a high-dimensional vector, with a true scalar answer <spanclass="math inline">\(a\)</span>. An agent’s estimate of the answer is an interpolation based on previously-seen questions and answers <spanclass="math inline">\((\bm{q}^i,a^i)_{i=1,\ldots,n}\)</span>. This framework extends an <ahref="https://tecunningham.github.io/posts/2023-09-05-model-of-ai-imitation.html">earlier model</a> developed for a different purpose.</p>
268
252
<p>The model yields several implications:</p>
269
253
<oltype="1">
@@ -337,7 +321,7 @@ <h2 id="toc-title">On this page</h2>
337
321
</tr>
338
322
</tbody>
339
323
</table>
340
-
<p>We can make some conjectures about adoption by task: :</p>
324
+
<p>We can make some conjectures about adoption by task:</p>
</code><buttontitle="Copy to Clipboard" class="code-copy-button"><iclass="bi"></i></button></pre><divclass="quarto-appendix-secondary-label">For attribution, please cite this work as:</div><divid="ref-cunningham2025" class="csl-entry quarto-appendix-citeas" role="listitem">
492
-
Cunningham, Tom. 2025. <span>“A Model of ChatGPT.”</span> June 14, 2025.
[==SUPERCEDED==].”</span> June 14, 2025. <ahref="https://tecunningham.github.io/posts/2025-06-14-model-of-chatgpt.html">tecunningham.github.io/posts/2025-06-14-model-of-chatgpt.html</a>.
0 commit comments