autosave: 2026-03-21 10:45:46

Author: tecunningham

_freeze/posts/2025-09-13-recursive-self-improvement-explosion-optimization/execute-results/html.json

@@ -9,7 +9,7 @@
 <link>tecunningham.github.io/</link>
 <atom:link href="tecunningham.github.io/index.xml" rel="self" type="application/rss+xml"/>
 <description>{{&lt; meta description-meta &gt;}}</description>
-<generator>quarto-1.8.25</generator>
+<generator>quarto-1.5.57</generator>
 <lastBuildDate>Sat, 07 Mar 2026 08:00:00 GMT</lastBuildDate>
 <item>
   <title>When You Overtake More Runners than You’re Overtaken by</title>

docs/index.html

@@ -3,7 +3,6 @@
     "listing": "/index.html",
     "items": [
       "/posts/2024-10-27-from-citlali.html",
-      "/posts/2026-03-19-heebie-jeebies.html",
       "/posts/2023-08-13-runner-percentile.html",
       "/posts/2026-01-29-knowledge-creating-llms.html",
       "/posts/2025-12-30-llm-verification.html",

docs/index.xml

@@ -2,14 +2,10 @@
 <urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
   <url>
     <loc>tecunningham.github.io/index.html</loc>
-    <lastmod>2025-10-29T15:46:32.992Z</lastmod>
+    <lastmod>2023-07-26T14:50:16.883Z</lastmod>
   </url>
   <url>
     <loc>tecunningham.github.io/posts/2023-01-31-social-media-suspensions-data.html</loc>
     <lastmod>2025-03-10T06:05:27.346Z</lastmod>
   </url>
-  <url>
-    <loc>tecunningham.github.io/posts/2026-03-19-heebie-jeebies.html</loc>
-    <lastmod>2026-03-20T03:28:02.301Z</lastmod>
-  </url>
 </urlset>

docs/listings.json

@@ -19,6 +19,30 @@ execute:
 ==OFFCUTS FILE==
 
 
+Extrapolating from these models.
+: It's dangerous to talk about "the elasticity" between A and B: it is meaningful to talk about the elasticity *at a given point* but it's a strong functional-form assumption that it'll have the same elasticity everywhere. 
+
+Mathematical assumptions on intelligence -> R&D:
+:  
+    1. _Full automation._ Once you hit a threshold of intelligence then it's worth entirely replacing humans with agents. E.g. replacing tollbooth operator with automatic tollbooth.
+    2. _Partial automation._ There's some subset of tasks, and you replace them one-by-one.
+    3. _Speedup._ -> Humans and agents are complements, an agent increases efficiency on all tasks.
+
+Concrete assumptions.
+:  
+    1. _Autonomous algorithm-search machine._ You spend a lot of tokens, it autonomously learns to navigate the space. Like AlphaZero, trained from scratch.
+    2. _Speeds up routine tasks._ The AI R&D guy can do certain things quicker, e.g. parse logs, write visualization, fix bugs.
+    3. _Shares knowledge._ Can learn tricks. It doesn't affect people at the frontier very much.
+
+Q: what if the capability frontier is the data frontier?
+: 
+    - Nice prediction: will not speed up experts, only behind the frontier.
+
+Natural processes often have explosive *regions*.
+: 
+    You 
+
+
 #           Summary
 
 

docs/posts/2025-09-13-recursive-self-improvement-explosion-optimization.html

@@ -1054,7 +1054,7 @@ @article{davidson2021could
   volume   = {25},
   year     = {2021},
   abstract = {Abstract unavailable.},
-  url      = {https://www.semanticscholar.org/search?q=Could%20advanced%20AI%20drive%20explosive%20economic%20growth}
+  url      = {https://www.openphilanthropy.org/research/could-advanced-ai-drive-explosive-economic-growth}
 }
 @article{deming2017growing,
   title     = {The growing importance of social skills in the labor market},
@@ -1187,6 +1187,13 @@ @article{erdil2025gate
   abstract             = {Assessing the economic impacts of artificial intelligence requires integrating insights from both computer science and economics. We present the Growth and AI Transition Endogenous model (GATE), a dynamic integrated assessment model that simulates the economic effects of AI automation. GATE combines three key ingredients that have not been brought together in previous work: (1) a compute-based model of AI development, (2) an AI automation framework, and (3) a semi-endogenous growth model featuring endogenous investment and adjustment costs. The model allows users to simulate the economic effects of the transition to advanced AI across a range of potential scenarios. GATE captures the interactions between economic variables, including investment, automation, innovation, and growth, as well as AI-related inputs such as compute and algorithms. This paper explains the model's structure and functionality, emphasizing AI development for economists and economic modeling for the AI community. The model is implemented in an interactive sandbox, enabling users to explore the impact of AI under different parameter choices and policy interventions. The modeling sandbox is available at: www.epoch.ai/GATE.},
   abstract_source      = {http://export.arxiv.org/api/query?id_list=2503.04941}
 }
+@misc{erdil2025automatingrd,
+  title  = {Most AI value will come from broad automation, not from R\&D},
+  author = {Erdil, Ege and Barnett, Matthew},
+  year   = {2025},
+  note   = {Gradient Updates / Epoch AI},
+  url    = {https://epoch.ai/gradient-updates/most-ai-value-will-come-from-broad-automation-not-from-r-d}
+}
 @misc{evgeniou2023navigating,
   url      = {https://knowledge.insead.edu/operations/navigating-trust-and-safety-world-generative-ai},
   author   = {Theodoros Evgeniou, Jeff Dunn and Alice Hunsberger},
@@ -3068,8 +3075,7 @@ @article{jones2022semiendogenous
   number  = {1},
   pages   = {125--152},
   url     = {https://www.annualreviews.org/content/journals/10.1146/annurev-economics-080521-012458},
-  url     = {https://www.annualreviews.org/content/journals/10.1146/annurev-economics-080521-012458},
-  doi     = {10.1146/annurev-economics-080321-033317}
+  doi     = {10.1146/annurev-economics-080521-012458}
 }
 
 @misc{ho2024algorithmicprogresslm,
@@ -3083,6 +3089,16 @@ @misc{ho2024algorithmicprogresslm
   url           = {https://arxiv.org/abs/2403.05812}
 }
 
+@misc{ho2025explosionexperiments,
+  title  = {The software intelligence explosion debate needs experiments},
+  author = {Ho, Anson and Whitfill, Parker},
+  year   = {2025},
+  month  = nov,
+  day    = {15},
+  note   = {Gradient Updates / Epoch AI},
+  url    = {https://epoch.ai/gradient-updates/the-software-intelligence-explosion-debate-needs-experiments}
+}
+
 @misc{ho2026leastunderstooddriver,
   title  = {The least understood driver of AI progress},
   author = {Ho, Anson},