corpus.blog
54,302 blogs · [ { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5312e2c6", "title": "Jabberwocky: translating nonsense to different languages", "url": "https://agarriga.substack.com/p/jabberwocky-translating-nonsense", "published_at": "2026-04-22T05:34:30+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa535e775a", "title": "Lojban is a terribly inefficient language", "url": "https://agarriga.substack.com/p/lojban-is-a-terribly-inefficient", "published_at": "2026-04-21T06:54:27+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa539179aa", "title": "General Chinese: romanization chaos in early 20th century China", "url": "https://agarriga.substack.com/p/general-chinese-romanization-chaos", "published_at": "2026-04-18T06:48:50+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa53ee0c56", "title": "The wizard’s word: an essay in verse", "url": "https://agarriga.substack.com/p/the-wizards-word-an-essay-in-verse", "published_at": "2026-04-17T06:46:23+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5460fac9", "title": "Confronting the tyrant", "url": "https://agarriga.substack.com/p/confronting-the-tyrant", "published_at": "2026-04-16T06:59:48+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa54bca987", "title": "Being Miguel Acevedo (pt. 1)", "url": "https://agarriga.substack.com/p/being-miguel-acevedo", "published_at": "2026-04-15T06:07:12+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa55a8e0c8", "title": "The difficulty of self-modeling explains why humans feel like they have free will.", "url": "https://agarriga.substack.com/p/the-difficulty-of-self-modeling-explains", "published_at": "2026-04-14T06:31:16+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa564defdb", "title": "Inside me there are at least two wolves", "url": "https://agarriga.substack.com/p/inside-me-there-are-at-least-two", "published_at": "2025-11-30T07:53:56+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa56609f33", "title": "Concrete reasons why alignment is on track to be solved", "url": "https://agarriga.substack.com/p/concrete-reasons-why-alignment-is", "published_at": "2025-11-27T19:20:36+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa573fe904", "title": "Spatially distributed consciousness is not an abstract thought experiment if AI is conscious", "url": "https://agarriga.substack.com/p/due-to-ai-spatially-distributed-consciousness", "published_at": "2025-11-26T04:55:11+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5780c1b3", "title": "Alignment will happen by default. What’s next?", "url": "https://agarriga.substack.com/p/alignment-will-happen-by-default", "published_at": "2025-11-25T07:11:39+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5852de01", "title": "Why investigate neural networks that plan? (Part 1)", "url": "https://agarriga.substack.com/p/why-investigate-neural-networks-that", "published_at": "2025-11-24T07:46:05+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa58ec6f5c", "title": "Two ways to advance science: mapping and search", "url": "https://agarriga.substack.com/p/two-ways-to-advance-science-mapping", "published_at": "2025-11-23T06:07:06+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5957a938", "title": "Does training LLMs with RL improve their ability to reason?", "url": "https://agarriga.substack.com/p/are-llms-trained-with-rl-actually", "published_at": "2025-11-22T07:52:43+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5a13e4af", "title": "Two rules for good research taste", "url": "https://agarriga.substack.com/p/two-rules-for-good-research-taste", "published_at": "2025-11-20T07:29:15+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5af3d288", "title": "Policy gradients (part 2): using 'probability-until-now'", "url": "https://agarriga.substack.com/p/policy-gradients-part-2-using-probability", "published_at": "2025-11-19T06:22:02+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5b4b81ec", "title": "Policy gradients (part 1): what is it?", "url": "https://agarriga.substack.com/p/policy-gradients-part-1", "published_at": "2025-11-18T07:02:56+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5b5f4ca4", "title": "Nvidia rules ML, not because its products are cost-effective", "url": "https://agarriga.substack.com/p/nvidia-rules-ml-because-its-products", "published_at": "2025-11-17T07:54:01+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5c5c30f9", "title": "Cultural norms shape contingent facts about AI releases", "url": "https://agarriga.substack.com/p/cultural-norms-shape-contingent-facts", "published_at": "2025-11-16T06:25:00+00:00" }, { "id": "01a0dcd2-3bc6-72c9-9188-c4aa5d236f40", "title": "Is memory-safe Linux within reach?", "url": "https://agarriga.substack.com/p/is-memory-safe-linux-within-reach", "published_at": "2025-11-15T06:40:17+00:00" } ] posts Claim your blog
Back to agarriga.substack.com
Blog · corpus.blog/blogs/agarriga.substack.com/posts

agarriga.substack.com

agarriga.substack.com

2026

Jabberwocky: translating nonsense to different languagesoriginal ↗
22 Apr 2026
Lojban is a terribly inefficient languageoriginal ↗
21 Apr 2026
General Chinese: romanization chaos in early 20th century Chinaoriginal ↗
18 Apr 2026
The wizard’s word: an essay in verseoriginal ↗
17 Apr 2026
Confronting the tyrantoriginal ↗
16 Apr 2026
Being Miguel Acevedo (pt. 1)original ↗
15 Apr 2026
The difficulty of self-modeling explains why humans feel like they have free will.original ↗
14 Apr 2026

2025

Inside me there are at least two wolvesoriginal ↗
30 Nov 2025
Concrete reasons why alignment is on track to be solvedoriginal ↗
27 Nov 2025
Spatially distributed consciousness is not an abstract thought experiment if AI is consciousoriginal ↗
26 Nov 2025
Alignment will happen by default. What’s next?original ↗
25 Nov 2025
Why investigate neural networks that plan? (Part 1)original ↗
24 Nov 2025
Two ways to advance science: mapping and searchoriginal ↗
23 Nov 2025
Does training LLMs with RL improve their ability to reason?original ↗
22 Nov 2025
Two rules for good research tasteoriginal ↗
20 Nov 2025
Policy gradients (part 2): using 'probability-until-now'original ↗
19 Nov 2025
Policy gradients (part 1): what is it?original ↗
18 Nov 2025
Nvidia rules ML, not because its products are cost-effectiveoriginal ↗
17 Nov 2025
Cultural norms shape contingent facts about AI releasesoriginal ↗
16 Nov 2025
Is memory-safe Linux within reach?original ↗
15 Nov 2025