ByteDance’s Founder Puts Frontier Parity at 2029-2030

Zhang Yiming told an internal AI all-hands that ByteDance can compete at the frontier but needs at least three to four more years to catch up.

At an all-hands for ByteDance’s AI team on 12 November 2026, reported by Bloomberg, Zhang Yiming said the company can compete with OpenAI and Anthropic at the frontier, but genuine parity would take until 2029 or 2030 at the earliest. He described the pre-training dataset behind the Seed model series as several orders of magnitude smaller than those used by leading overseas labs, and called that the most urgent problem to solve. The commercial read is straightforward. A founder publicly setting a multi-year gap is a capital-allocation instruction rather than a concession: it prioritises spending on data and compute over launch marketing, and it hands domestic peers a smaller, more defensible target to compete against.