AI Reading List

Daily curated list of news, articles, and other interesting posts related to AI.

2025-11-21

• FAWK: LLMs can write a language interpreter (martin.janiczek.cz)

  Interesting look at how large language models like Sonnet 4.5 can generate fully working interpreters for new languages with minimal human guidance. The author reflects on both the possibilities and pitfalls of using LLMs for rapid prototyping, noting impressive results but cautioning about losing familiarity with the generated code.

• The New AI Consciousness Paper (www.astralcodexten.com)

  Overview of recent efforts to identify signs of consciousness in AI, focusing on computational theories like recurrent processing, global workspace, and higher order thought. The discussion highlights both the challenges and promise of assessing AI consciousness, especially in light of new research by leading experts.

• Boom, bubble, bust, boom. Why should AI be different? (crazystupidtech.com)

  Interesting take on how the AI boom closely mirrors the internet bubble of the late 1990s, with even faster growth and significantly higher spending. While AI promises transformative innovation, the risks of concentrated investment and unclear long-term business value suggest the bubble could burst with widespread consequences.