The author tests CodeX, GPT5.6‑Sol and Fable5, exposing simple yet puzzling errors—missed tasks, inconsistent CSS changes, and contradictory answers—that may stem from catastrophic forgetting and highlight emerging reliability bottlenecks in large language models.
The article examines how multimodal AI models repeatedly miscount a six‑finger image, explores the underlying bias revealed in the paper “Vision Language Models are Biased,” and warns that such prior‑knowledge‑driven errors can have serious safety implications in real‑world applications.
The article analyzes major limitations of large language models—hallucination, outdated knowledge, and insufficient domain expertise—and explains how Retrieval‑Augmented Generation and various fine‑tuning strategies can address these issues while outlining practical cost considerations.
In a detailed interview, Tsinghua researcher Zhao Sicheng and JD Retail senior director Peng Changping discuss the latest progress in multimodal large models, their practical applications in advertising and e‑commerce, persistent challenges such as hallucinations and data alignment, and the skills engineers need to thrive in the emerging AI era.
