Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Ask HN: What can LLMs not do?

Dec 27, 2024 - news.ycombinator.com
The author reflects on their past prediction about the advancement of AI, specifically regarding autonomous-only highway lanes by 2024, and notes that while progress seemed slow initially, AI development has accelerated, particularly with large language models (LLMs). They express surprise at how quickly AI capabilities have advanced, to the point where their previous distinctions between human and AI abilities have mostly been surpassed.

The author seeks new criteria to differentiate humans from AI, emphasizing the need for objective, testable measures rather than subjective or philosophical ones like consciousness or soul. They are not interested in limitations related to form or intentional design, such as memory or physical embodiment, and are looking for insights from a technically knowledgeable and skeptical community.

Key takeaways:

  • The author has been interested in AI for over a decade and made a bet about autonomous highway lanes by 2024.
  • They have observed faster-than-expected progress in AI, particularly with large language models (LLMs).
  • The author seeks new criteria to differentiate humans from AI, avoiding subjective measures and focusing on testable aspects.
  • They are not interested in limitations related to form factor or intentional constraints, such as stateful memory or robotic bodies.
View Full Article

Comments (0)

Be the first to comment!