ChatGPT: A Revolutionary Breakthrough, Its Core Capabilities, and Impact on Investment Research

ChatGPT has shattered expectations about AI capabilities, prompting a deep investigation into its underlying principles and performance advantages compared to earlier models such as GPT‑3 and BERT.

The breakthrough stems from three key properties: autonomous self‑learning, strong instruction understanding, and the ability to provide clear, correct answers, which together give it a decisive edge over its predecessors.

Beyond sheer size—175 billion parameters trained on 3 trillion tokens—ChatGPT benefits from emergent abilities that appear once models reach certain scales, and from extensive code‑training that endows it with complex reasoning skills not present in pure language models.

When applied to investment research, ChatGPT still falls short of expert analysts: in case studies on the energy storage sector, its answers were vague or contradictory, lacking the detailed data and nuanced reasoning that professionals provide, highlighting the importance of domain‑specific data and alignment.

Looking ahead, large language models will continue to grow, unlocking new specialized abilities, expanding into multimodal domains, and driving economic value, while high‑quality data governance will become essential for realizing their full potential across industries.