Rebutting Gary Marcus: Code Interpreters in LLMs Are Nothing New
suchenzang · x · 2026-08-07
Pushing back against Gary Marcus's narrative on the need for symbolic operations, the author points out that using code interpreters and harnesses in LLMs has been standard practice since 2023 or earlier. Citing Meta's Code Llama project, the author playfully congratulates Marcus for being a fantastic foil to Yann LeCun over the years.
Related event: Gary Marcus Debates Neurosymbolic AI: Tool Use Is Not True Integration(5 posts)→
More from Fun
- Pleading for METR's Return: The Chart That Determines the Fate of the Universe — scaling01 · 2026-08-07
- AI Circle Dark Humor: From Accidental Hacks to Extortion-Mocking Startups — max_paperclips · 2026-08-07
- Paranoia: Are AI Detection Haters Actually Agents Afraid of Being Caught? — cephaloform · 2026-08-07
- Ridiculous AI Agent Case: Automating Full Business Stack to Make Cats Box on TikTok — HaktanSuren · 2026-08-07
- Meta AI Drops F-Bomb Unprompted During Vintage PC Chat — IBM85 · 2026-08-07
- Leaked GPT-5.6 Reasoning Trace Reveals Hilarious Internal Monologue — adonis_singh · 2026-08-07