吴浩哲 / Wu Haozhe

Overview

吴浩哲 is presented in the source as a Wuhan University mathematics graduate, University of Hong Kong researcher, and author and maintainer of the open-source Vibe Trading project.

Current Profile

His account of AI-assisted investing is shaped by experience across traditional multifactor research, deep learning, large-model text processing, and agents. He argues that financial AI should not optimize only for producing more signals: it should make claims understandable, testable, time-valid, and auditable before any capital is committed.

He keeps human responsibility at the end of the chain. Models can read material, form hypotheses, write code, and propose judgments, while deterministic checks and constraint systems filter possible action; people still set risk boundaries, position size, exceptions, and rules.

Key Characteristics

• Builds at the intersection of quantitative finance, natural-language research, and agentic workflows.
• Treats Financial Data Alignment / 金融数据对齐 and Point-in-Time Backtesting / 时点回测 as more durable system value than unbounded model-generated intelligence.
• Favors explainable, auditable research chains over opaque signals or bare probability outputs.
• Uses AI extensively in personal research while retaining final review and responsibility for each position.
• Frames human participation as both a safety mechanism and a source of agency, commitment, and meaning.

Evidence

Technical and product orientation

• AI Trading —— 决策便宜，行动很贵｜对谈超 3w Star Vibe-Trading 作者浩哲 describes his path through multifactor, deep-learning, language-model, and agent approaches and his role maintaining Vibe Trading.

Research and control philosophy

• AI Trading —— 决策便宜，行动很贵｜对谈超 3w Star Vibe-Trading 作者浩哲 records his material–hypothesis–validation–bounded-execution–review workflow and his insistence on data timing, contrary evidence, costs, and explicit stopping when evidence is missing.

Human responsibility

• AI Trading —— 决策便宜，行动很贵｜对谈超 3w Star Vibe-Trading 作者浩哲 attributes risk limits, position size, exceptions, and ultimate responsibility to people even when AI performs most of the research.

Qualifications

• The profile is based on one podcast source and does not independently verify education, institutional role, project metrics, investment performance, or the empirical basis of every market claim.
• His reported reliance on AI for roughly 80%–90% of personal trades describes his workflow, not a recommendation or measured safety threshold.
• Statements about future market speed, data value, and multi-agent behavior are forward-looking or source-reported interpretations.

What Changed

• Added the first canonical profile of 吴浩哲 and his AI-trading research philosophy.

Relationships

• Vibe Trading - open-source project he authors and maintains.
• AI Trading - system approach he defines around verified, bounded movement from judgment to action.
• Decision–Action Cost Asymmetry / 决策—行动成本不对称 - central framing for why cheap decisions require expensive action gates.
• Financial Data Alignment / 金融数据对齐 - control layer he identifies as especially important for trustworthy financial research.
• 42章经 - podcast in which the bounded profile was presented.