DeepSeek’s Liang Wenfeng Says Coding Agents Come Before General Agents and Vertical AI
oran_ge · x · 2026-07-23
A long set of quotes from DeepSeek founder Liang Wenfeng’s recent investor meeting, covering product strategy, open source, competition, and the AGI roadmap.
Main points
- He argues product is only a byproduct on the road to AGI, not the main destination.
- He downplays 3D, video generation, and world models as having limited direct relation to intelligence ceilings.
- He says multimodality matters, but is not intelligence itself, and that hallucinations are mainly a product problem.
- He believes coding agents are the priority right now, ahead of general agents or vertical agents in finance and healthcare.
- The next major step after CoT and agents, in his view, is continual learning; only after that could AI accelerate frontier research and eventually reach embodied intelligence.
- On business, he says they aim for reasonable profits, low prices, no sales team and no customer support, and that open source is strategically compatible with that approach.
- He attributes China’s AI gap mainly to resources, expects the domestic model market to consolidate, and says large-model companies will not capture most of AI’s profits.
- He also stresses team stability, restraint, and vision-driven culture over KPI-style management.
The quotes read like a compact manifesto: AGI first, product second, cost discipline, and open source as strategy rather than charity.
Related event: DeepSeek Founder's Investor Call Reveals AGI Roadmap and Compute Plans(37 posts)→
More from AGI Musings
- Anthropic's S-1 Shows $6.6M Revenue Per Employee as AI Agents Log 6x Human Hours — jacalulu · 2026-10-26
- Ben Recht: The world needs a truly open language model, and it's easier to build than you think — beenwrekt · 2026-10-26
- Which jobs fall to AI first? The rule: any layer whose rules can be written as code — FinanceYF5 · 2026-10-26
- Anthropic S-1: expert-level AI 'within a few years', 90% of code written by Claude — kimmonismus · 2026-10-26
- The identity crisis in software engineering: multiagent systems demand a new problem set — rakyll · 2026-10-26
- Frontier LLMs negotiate contracts poorly and frequently renege for selfish gain, study finds — xuanalogue · 2026-10-26