As of October 2026, AI progress is outpacing ordinary product managers’ ability to keep up, with widespread use of free AI tools and increasing company-imposed restrictions in large tech firms.

The REA tool is highlighted as a must-have for reverse engineering and rapid development, especially for asset creation and reducing costs, while high-end tasks still require specialized tools.

Local GPUs and offline AI use are becoming essential for researchers and teams, with rising GPU costs and a move toward self-contained, local models for confidential work.

AI model economics create a “black hole” effect: higher-quality models attract more data and revenue for their developers, potentially weakening industries that don’t adopt AI.

A personal one-month example of developing an offline-capable writing/app tool illustrates the accelerating pace of development and the shift between offline and online model usage.

Content shifts in 2026 favor short videos over long-form content, pushing product managers to focus on deep user insight to identify needs and sustain product vitality.

Long-term, high-cost, high-iteration progress is seen in transformative technologies like flying cars, signaling potential disruption but with regulatory and practical adoption still uncertain.

Access to advanced models in big tech is increasingly restricted and monitored, suggesting reduced freedom and career credit in large firms compared to smaller firms or independent ventures.

The piece is published via a WeChat Official Account and reprinted with authorization from 36Kr, authored by Kevin’s Bits of Changing the World.

Takeaway for PMs: Continuously track AI progress, invest beyond work tasks, and adapt quickly to evolving tools and content-creation modes to avoid obsolescence.

In large tech companies, only products with clear ongoing benefits survive, while smaller, overlooked products can sustain smaller firms, contributing to a perceived draining of PMs’ value in big organizations.