Tencent chief AI scientist Yao Shunyu says race is a 'long-term game'
Synopsis
Tencent's chief AI scientist Yao Shunyu, a former OpenAI researcher, publicly dismissed lag concerns and declared the AI race is only in its second half — pointing to trillion-dollar opportunities in coding agents and embodied intelligence that no company has yet captured.
Key Takeaways
Yao Shunyu , chief AI scientist at Tencent Holdings , said 'AI is a long-term game, with the second half of the race just starting' at a Shenzhen summit on 6 June 2026 .
Yao compared the current AI landscape to the personal-computer industry of the 1970s , arguing that foundational opportunities remain largely unclaimed.
He expressed doubt that OpenAI 's ChatGPT and Anthropic 's Claude Code would be the only AI super apps, predicting a 'steady stream of new opportunities.' Tencent identified coding agents, multimodal AI, and embodied intelligence as 'trillion-dollar markets yet to be filled.' Yao heads both the Hunyuan large language model and AI infrastructure divisions at Tencent .
He acknowledged that setbacks in early LLM development are normal and called for honesty about shortcomings as the key to competing in the AI race's second phase.
Yao Shunyu, Tencent Holdings' chief AI scientist and head of its Hunyuan large language model division, pushed back on Friday against perceptions that the company is falling behind in artificial intelligence, declaring the competition is only entering its second half with trillion-dollar markets still untapped. Speaking at Tencent's Cloud AI Industry Applications Summit in Shenzhen on 6 June 2026, the former OpenAI researcher framed the current AI landscape as analogous to the nascent personal-computer era of the 1970s.
Why it matters
Yao argued that the AI race is far from settled, pointing to coding agents, multimodal AI, and embodied intelligence as frontiers where no incumbent has yet claimed dominance. 'While coding agents and productivity will undoubtedly become more critical, there are trillion-dollar markets yet to be filled as we are witnessing the emergence of new things including multimodal AI and embodied intelligence,' he said. His remarks signal that Tencent is positioning itself for a prolonged contest rather than a sprint against rivals such as Alibaba Group Holding and ByteDance.
The competitive backdrop
Yao directly addressed the dominance of Western AI products, expressing scepticism that OpenAI's ChatGPT and Anthropic's Claude Code would remain the only super apps. 'That would be a very bleak world. Instead, a steady stream of new opportunities will inevitably emerge,' he said. The comment underscores Tencent's bet that platform fragmentation — rather than winner-takes-all consolidation — will define the next phase of AI adoption.
Acknowledging setbacks, betting on patience
Yao, who also oversees Tencent's AI infrastructure, conceded that detours and setbacks are a normal part of early LLM and AI product exploration. He stressed that success in the second half of the AI race hinges on intellectual honesty about shortcomings and patience when making strategic adjustments. The candid acknowledgement is notable given the intense public scrutiny Chinese tech giants face over their AI progress relative to US peers.
What's next
Tencent's emphasis on embodied intelligence — AI systems that interact with the physical world — aligns with a broader industry pivot toward robotics and agentic workflows that analysts expect to accelerate through 2026 and beyond. With Pony Ma Huateng's company already investing heavily in Hunyuan model iterations, the next milestones to watch are product releases in coding-agent tooling and any disclosed partnerships in the robotics supply chain. How quickly Tencent can convert its cloud customer base into AI application users will be the clearest indicator of whether its long-game thesis is paying off.
Point of View
Arriving just as US export controls on advanced chips create asymmetric pressure on compute-intensive model training — pivoting toward robotics and agentic workflows partially sidesteps that bottleneck. Investors should watch whether Tencent's patience thesis translates into measurable cloud AI revenue growth in the next two earnings cycles.
NationPress
21 Jul 2026
Frequently Asked Questions
What did Tencent's chief AI scientist say about the AI race?
Yao Shunyu , Tencent 's chief AI scientist, said 'AI is a long-term game, with the second half of the race just starting,' speaking at a company summit in Shenzhen on 6 June 2026 . He compared the current moment to the personal-computer industry in the 1970s , arguing that the most significant opportunities are still ahead.
Who is Yao Shunyu at Tencent?
Yao Shunyu is Tencent Holdings ' chief AI scientist and head of both its Hunyuan large language model and AI infrastructure divisions. He previously worked at OpenAI before joining Tencent .
What AI markets is Tencent targeting?
Tencent is targeting coding agents, multimodal AI, and embodied intelligence, which Yao described as 'trillion-dollar markets yet to be filled.' He said productivity and coding agents will become increasingly critical, while embodied intelligence represents an emerging frontier.
Does Tencent think ChatGPT and Claude will dominate AI?
Yao Shunyu said he does not expect OpenAI 's ChatGPT and Anthropic 's Claude Code to be the only AI super apps, calling such an outcome 'a very bleak world.' He predicted that 'a steady stream of new opportunities will inevitably emerge,' suggesting the market will remain fragmented.
How is Tencent competing against Alibaba and ByteDance in AI?
Tencent is differentiating itself through its Hunyuan model ecosystem and a long-term investment posture, according to Yao Shunyu . Rather than racing to match rivals Alibaba Group Holding and ByteDance on near-term benchmarks, the company is betting on application-layer opportunities in embodied intelligence and agentic AI that have yet to mature.
