Stock Markets July 23, 2026 05:24 AM

DeepSeek Founder Says Firm Will Focus on AGI Over Near-Term Profit; Top Models Likely to Stay Open-Source

Founder emphasizes computing resources as the primary bottleneck and signals plans for large AI clusters while remaining undecided on custom chips

By Marcus Reed
Share
Twitter Reddit Facebook LinkedIn

DeepSeek's founder told investors the company will prioritise long-term artificial general intelligence (AGI) development ahead of maximising short-term profits, and indicated the firm's most advanced models will likely remain open-source. The remarks, made during a recent four-hour investor meeting, highlighted computing power as the chief constraint and disclosed that the company recently raised capital at an approximate $52 billion valuation after a previous policy of rejecting external funding.

DeepSeek Founder Says Firm Will Focus on AGI Over Near-Term Profit; Top Models Likely to Stay Open-Source
Summarize with
ChatGPT Perplexity Claude Grok Gemini

Key Points

  • DeepSeek's founder told investors the company will prioritise long-term AGI development over maximising near-term profits.
  • The company is likely to keep its most advanced models open-source, with the founder asserting open-source development and commercial monetisation can coexist.
  • DeepSeek completed its first fundraising round at an approximate $52 billion valuation after previously rejecting external capital; it faces growing domestic competition.

BEIJING, July 23 - The founder of Chinese AI startup DeepSeek told investors that the company intends to prioritise the pursuit of long-term artificial general intelligence rather than focusing on maximising near-term profits. According to a transcript of a recent four-hour meeting with investors, the founder argued that developing AGI remains the firm's central objective.

At the meeting, the founder said DeepSeek is likely to keep its most advanced models open-source, asserting that open-source development does not preclude commercial monetisation. The comment framed openness and revenue generation as complementary rather than mutually exclusive approaches.

DeepSeek rose to global attention early last year after its low-cost V3 and R1 models drew notice for challenging assumptions about China’s ability to compete with US AI leaders despite restrictions on advanced-chip exports. The company has since shifted strategy: it recently completed its first fundraising round at a roughly $52 billion valuation, marking a reversal of a years-long position against taking external capital. That change comes amid heightened competition from domestic players, including established tech firms and well-funded startups.

During the investor session, the founder described access to computing power as the company’s main constraint, saying the principal gap with the US is rooted in resource availability rather than in technical expertise. He noted plans to construct large AI computing clusters in the future to address that constraint, while remaining undecided on whether to pursue development of proprietary AI chips.

On the topic of chips, the founder was quoted saying: "I hope to be able to buy chips at a reasonable price so that I don’t have to make chips myself." This underscores a preference for procuring components on the market if supply and pricing permit, rather than entering semiconductor manufacturing prematurely.

The founder’s comments appear against a backdrop of export controls. The United States has already restricted exports to China of some of the most advanced AI chips. However, authorities have so far refrained from adding DeepSeek, Chinese memory chipmaker CXMT and more than 100 other companies flagged as national security risks to a trade blacklist, in part to avoid escalating tensions.


Contextual note: The founder’s remarks were delivered in a lengthy investor meeting and are drawn from a meeting transcript. The company’s recent fundraising round and valuation were cited at the meeting.

Risks

  • Access to computing resources is identified by the founder as the main constraint on DeepSeek's progress, which could slow model training and deployment; this primarily affects the cloud and data center sectors.
  • Uncertainty remains over chip supply and pricing; the founder is undecided on building proprietary AI chips, leaving hardware supply and semiconductor markets as potential risk factors.
  • Geopolitical and export-control dynamics introduce uncertainty: while some advanced AI chips are already subject to export limits, the absence of a trade blacklist listing for DeepSeek to date leaves open the possibility of future policy shifts affecting operations and partnerships.

More from Stock Markets

AMD Says Second-Generation Helios AI Servers Are in Full Production, Shipping by Quarter-End Jul 23, 2026 Tech Titans Slide as Magnificent 7 Lose $123.96 Billion in a Single Session Jul 23, 2026 Morgan Stanley Views Power-Rich Bitcoin Miners as Fastest Route to AI Data Center Capacity Jul 23, 2026 Albertsons Lowers Sales and Profit Targets, Boosts Price Investments as Shares Dive Jul 23, 2026 Blackstone outlines measures to ease environmental and community concerns around AI data centers Jul 23, 2026