Skip to content

Commit 3296c54

Browse files
committed
commit
1 parent 8e27f8c commit 3296c54

6 files changed

Lines changed: 68 additions & 222 deletions

File tree

_freeze/posts/2026-03-13-apple-picking-ai/execute-results/html.json

Lines changed: 2 additions & 2 deletions
Large diffs are not rendered by default.
Binary file not shown.
473 Bytes
Loading

docs/posts/2026-03-13-apple-picking-ai.html

Lines changed: 19 additions & 53 deletions
Original file line numberDiff line numberDiff line change
@@ -165,7 +165,7 @@
165165
<meta name="twitter:card" content="summary_large_image">
166166
</head>
167167

168-
<body class="nav-fixed fullcontent quarto-light">
168+
<body class="nav-fixed quarto-light">
169169

170170
<div id="quarto-search-results"></div>
171171
<header id="quarto-header" class="headroom fixed-top quarto-banner"><div id="quarto-draft-alert" class="alert alert-warning"><i class="bi bi-pencil-square"></i>Draft</div>
@@ -253,9 +253,11 @@ <h1 class="title">An Apple-Plucking Model of Agents</h1>
253253
</header><div id="quarto-content" class="quarto-container page-columns page-rows-contents page-layout-article page-navbar">
254254
<!-- sidebar -->
255255
<!-- margin-sidebar -->
256-
256+
<div id="quarto-margin-sidebar" class="sidebar margin-sidebar zindex-bottom">
257+
258+
</div>
257259
<!-- main -->
258-
<main class="content quarto-banner-title-block" id="quarto-document-content">
260+
<main class="content quarto-banner-title-block page-columns page-full" id="quarto-document-content">
259261

260262

261263

@@ -268,7 +270,10 @@ <h1 class="title">An Apple-Plucking Model of Agents</h1>
268270
dt {grid-column-start: 1; width: 6cm;}
269271
dd {grid-column-start: 2; margin-left: 2em;}
270272
</style>
271-
<dl>
273+
274+
<div class="no-row-height column-margin column-container"><div class="">
275+
<p>Thanks to Nate Rush, Manish Shetty, Basil Halperin for helpful comments.</p>
276+
</div></div><dl>
272277
<dt>An apple-picking model of AI work.</dt>
273278
<dd>
274279
<p>Here’s a simple model useful for thinking about AI’s contribution to solving problems.</p>
@@ -279,11 +284,11 @@ <h1 class="title">An Apple-Plucking Model of Agents</h1>
279284
<dt>Implications of the apple-picking model.</dt>
280285
<dd>
281286
<ol type="1">
282-
<li>Agents can autonomously advance the state-of-the-art on an optimization problem, yet still not be a perfect substitute for human labor (they can find the unpicked low apples).</li>
283-
<li>Agent contribution to a problem, as you scale the expenditure, will be higher than human contribution, but then asymptote to a lower maximum value.</li>
284-
<li>Agents will have relatively bigger value (relative to humans) for problems that are not yet heavily optimized.</li>
287+
<li>Agents can autonomously advance the state-of-the-art on an optimization problem, yet still not be a perfect substitute for human labor (they can find the low apples that haven’t been picked yet).</li>
288+
<li>Agent contribution to a problem, as you scale the expenditure, will be higher than human contribution, but then asymptote to a lower maximum value (they’ll pick a lot of apples, but will never be able to pick them all).</li>
289+
<li>Agents will have relatively bigger value, relative to humans, for problems that are not yet heavily optimized (robots are useful for trees that have never been picked).</li>
285290
<li>Agents will asymptote to higher points if they are given better human starting points (if a tree is partly picked).</li>
286-
<li>To gauge the ability of agents we want to test not just for their ability to improve performance, but their <em>reach</em>.</li>
291+
<li>To gauge the ability of agents we want to test not just for their ability to improve performance, but their <em>reach</em> (we want to benchmark robots not on how many apples they can pick, but the height of the highest apple they can reach).</li>
287292
</ol>
288293
</dd>
289294
<dt>Relation to other literature.</dt>
@@ -393,22 +398,13 @@ <h3 class="anchored" data-anchor-id="state-variables-and-dynamics">1. State vari
393398
<h3 class="anchored" data-anchor-id="crisp-conditions">2. Crisp conditions</h3>
394399
<p><strong>A) Activation.</strong> With <span class="math inline">\(\lambda_t = 0\)</span>, <span class="math inline">\(a_t = h_t \to 1\)</span>. So the agent can ever turn on <strong>iff <span class="math inline">\(\bar{a} &lt; 1\)</span></strong>. If <span class="math inline">\(\bar{a} \ge 1\)</span>, <span class="math inline">\(\lambda_t \equiv 0\)</span> forever. Activation-time approximation: <span class="math inline">\(h_t = 1 - p^t \ge \bar{a}\)</span> <span class="math inline">\(\Leftrightarrow\)</span> <span class="math inline">\(t \ge \ln(1-\bar{a})/\ln p\)</span>; <span class="math inline">\(p\)</span> mainly shifts <em>when</em> activation happens.</p>
395400
<p><strong>B) Crossing human level.</strong> As <span class="math inline">\(t \to \infty\)</span>, <span class="math inline">\(h_t \to 1\)</span>. If <span class="math inline">\(\lambda_t &lt; 1\)</span>, <span class="math inline">\(a_t \to 1\)</span>; if <span class="math inline">\(\lambda_t \ge 1\)</span>, <span class="math inline">\(a_t = \lambda_t\)</span>. So asymptotically <span class="math inline">\(\lambda_{t+1} \to f(1)\)</span> with <span class="math inline">\(f(1) = 0\)</span> if <span class="math inline">\(1 &lt; \bar{a}\)</span>, and <span class="math inline">\(f(1) = \alpha + \beta(1-\bar{a})\)</span> if <span class="math inline">\(1 \ge \bar{a}\)</span>. So <strong>takeoff past human level</strong> (eventually <span class="math inline">\(\lambda &gt; 1\)</span>) <strong>iff</strong> <span class="math display">\[\boxed{\alpha + \beta(1-\bar{a}) &gt; 1.}\]</span> Interpretation: “If the orchard were fully human-level (<span class="math inline">\(a=1\)</span>), would the next agent be at least human-level?” If not, the system stays below 1. This condition is essentially independent of <span class="math inline">\(p\)</span> (timing, not whether).</p>
396-
<p><strong>C) Above human level: runaway vs saturation.</strong> For <span class="math inline">\(\lambda_t \ge 1\)</span>, <span class="math inline">\(a_t = \lambda_t\)</span> and <span class="math display">\[\lambda_{t+1} = \alpha + \beta(\lambda_t - \bar{a}).\]</span> - <strong>Runaway / hard takeoff</strong> iff <span class="math inline">\(\boxed{\beta &gt; 1}\)</span> (roughly geometric growth in <span class="math inline">\(\lambda_t\)</span>). - <strong>Soft takeoff / saturation</strong> iff <span class="math inline">\(\boxed{\beta &lt; 1}\)</span>: convergence to <span class="math display">\[\lambda^* = \frac{\alpha - \beta\bar{a}}{1-\beta}\]</span> (provided the system crosses 1 first). - <strong>Knife-edge</strong> <span class="math inline">\(\beta = 1\)</span>: linear growth.</p>
401+
<p><strong>C) Above human level: runaway vs saturation.</strong> For <span class="math inline">\(\lambda_t \ge 1\)</span>, <span class="math inline">\(a_t = \lambda_t\)</span> and <span class="math display">\[\lambda_{t+1} = \alpha + \beta(\lambda_t - \bar{a}).\]</span></p>
402+
<ul>
403+
<li><p><strong>Runaway / hard takeoff</strong> iff <span class="math inline">\(\boxed{\beta &gt; 1}\)</span> (roughly geometric growth in <span class="math inline">\(\lambda_t\)</span>).</p></li>
404+
<li><p><strong>Soft takeoff / saturation</strong> iff <span class="math inline">\(\boxed{\beta &lt; 1}\)</span>: convergence to <span class="math display">\[\lambda^* = \frac{\alpha - \beta\bar{a}}{1-\beta}\]</span> (provided the system crosses 1 first).</p></li>
405+
<li><p><strong>Knife-edge</strong> <span class="math inline">\(\beta = 1\)</span>: linear growth.</p></li>
406+
</ul>
397407
<hr>
398-
</section>
399-
<section id="illustrations" class="level3">
400-
<h3 class="anchored" data-anchor-id="illustrations">3. Illustrations</h3>
401-
<p>The figures below implement this model: four canonical trajectories, phase diagram in <span class="math inline">\((\alpha,\beta)\)</span>, cobweb plots of the asymptotic map, and sensitivity to <span class="math inline">\(p\)</span> and <span class="math inline">\(\bar{a}\)</span>.</p>
402-
<div class="cell" data-layout-align="center">
403-
<div class="cell-output-display">
404-
<div class="quarto-figure quarto-figure-center">
405-
<figure class="figure">
406-
<p><img src="2026-03-13-apple-picking-ai_files/figure-html/apple-rsi-four-regimes-1.png" class="img-fluid figure-img" width="672"></p>
407-
<figcaption>Four regimes: no activation (&lt;U+0101&gt;&lt;U+2265&gt;1), activated but stuck (f(1)&lt;1), soft takeoff (f(1)&gt;1, &lt;U+03B2&gt;&lt;1), hard takeoff (f(1)&gt;1, &lt;U+03B2&gt;&gt;1)</figcaption>
408-
</figure>
409-
</div>
410-
</div>
411-
</div>
412408
<div class="cell">
413409
<div class="cell-output-display">
414410
<div>
@@ -418,36 +414,6 @@ <h3 class="anchored" data-anchor-id="illustrations">3. Illustrations</h3>
418414
</div>
419415
</div>
420416
</div>
421-
<div class="cell" data-layout-align="center">
422-
<div class="cell-output-display">
423-
<div class="quarto-figure quarto-figure-center">
424-
<figure class="figure">
425-
<p><img src="2026-03-13-apple-picking-ai_files/figure-html/apple-rsi-phase-1.png" class="img-fluid figure-img" width="528"></p>
426-
<figcaption>Phase diagram: f(1)=1 (cross human level) and &lt;U+03B2&gt;=1 (runaway vs saturation) at fixed &lt;U+0101&gt; = 0.3</figcaption>
427-
</figure>
428-
</div>
429-
</div>
430-
</div>
431-
<div class="cell" data-layout-align="center">
432-
<div class="cell-output-display">
433-
<div class="quarto-figure quarto-figure-center">
434-
<figure class="figure">
435-
<p><img src="2026-03-13-apple-picking-ai_files/figure-html/apple-rsi-cobweb-1.png" class="img-fluid figure-img" width="672"></p>
436-
<figcaption>Cobweb: asymptotic map &lt;U+03BB&gt;_{t+1} = f(max(&lt;U+03BB&gt;_t,1)); soft (&lt;U+03B2&gt;&lt;1) vs hard (&lt;U+03B2&gt;&gt;1)</figcaption>
437-
</figure>
438-
</div>
439-
</div>
440-
</div>
441-
<div class="cell" data-layout-align="center">
442-
<div class="cell-output-display">
443-
<div class="quarto-figure quarto-figure-center">
444-
<figure class="figure">
445-
<p><img src="2026-03-13-apple-picking-ai_files/figure-html/apple-rsi-sensitivity-1.png" class="img-fluid figure-img" width="672"></p>
446-
<figcaption>Sensitivity: effect of p (left) and of &lt;U+0101&gt; (right) on &lt;U+03BB&gt;_t trajectory</figcaption>
447-
</figure>
448-
</div>
449-
</div>
450-
</div>
451417

452418

453419

473 Bytes
Loading

0 commit comments

Comments
 (0)