A physicist has explored the capabilities of GPT-5, a hypothetical future model, in understanding complex scientific concepts like black holes. The physicist found that while GPT-5 can generate plausible-sounding explanations, these often lack true scientific rigor and are more akin to "vibe physics." This suggests that current large language models, even advanced future versions, may struggle with genuine scientific reasoning and discovery. AI
IMPACT Suggests that even advanced future LLMs may not achieve genuine scientific reasoning, highlighting current limitations.
RANK_REASON An opinion piece from a physicist discussing the potential capabilities and limitations of a future AI model.
Read on Latent Space (podcast video) →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
