ADD / XOR / ROL: A non-anthropomorphized view of LLMs

content/interets/liens-interessants/2025/07/07/3add12ff/data/screenshot.yaml

@@ -0,0 +1,4 @@
+title: "Capture d'écran de https://addxorrol.blogspot.com/2025/07/a-non-anthropomorphized-view-of-llms.html"
+description: "Capture effectuée le 7 juillet 2025 à 09:28"
+attribution: "Richard Dern"
+file: "images/screenshot.png"

content/interets/liens-interessants/2025/07/07/3add12ff/index.md

@@ -0,0 +1,25 @@
+---
+links:
+  - name: Page d'origine
+    url: https://addxorrol.blogspot.com/2025/07/a-non-anthropomorphized-view-of-llms.html
+    lang: en
+  - name: Archive
+    url: http://web.archive.org/web/20250707001913/https://addxorrol.blogspot.com/2025/07/a-non-anthropomorphized-view-of-llms.html
+    archive: true
+title: "ADD / XOR / ROL: A non-anthropomorphized view of LLMs"
+date: 2025-07-07
+keywords: []
+cover:
+  "images/screenshot.png"
+  # Example:
+  # links:
+  #   - name: "LEGO (France)"
+  #     lang: "fr"
+  #     url: "https://www.lego.com/fr-fr/product/t-rex-rampage-75936"
+  #     official: true
+status: [{ "date": "2025-07-07T07:28:35.158Z", "http_code": 200 }]
+---
+
+> In many discussions where questions of "alignment" or "AI safety" crop up, I am baffled by seriously intelligent people imbuing almost magical human-like powers to something that - in my mind - is just MatMul with interspersed nonlinearities.
+
+Je suis entièrement d'accord avec l'auteur, qui pense que nous anthropomorphisons l'intelligence artificielle, ce qui nuit fortement à toute discussion saine sur le sujet.