字节跳动首次针对特定业务部门派发独立期权。据英国《金融时报》报道，为了防范腾讯等竞争对手挖角，字节跳动近期向旗下 AI 实验室 Seed 部门员工发放了以豆包命名的低价特股期权，使员工可以直接分享 AI 业务的增长红利，而不会受到集团不同业务线股权稀释的影响。期权方案是字节跳动成立以来首次发行与单一业务单元绑定的特殊期权。 随着国内大模型领域的人才争夺战升温，基础设施和数据标注等核心工程岗位成为抢夺重点。字节跳动自 2023 年建立 Seed 实验室以来，凭借大规模的算力基础设施投资成为 Nvidia 在中国的最大客户，吸引了大量顶尖 AI 研究员。然而，腾讯自 2025 年底承认大模型开发滞后以来，近期由前 OpenAI 研究员姚顺雨带队，开展了密集的定向挖角。近期离职的 Seed 视觉平台资深成员肖雪峰与基础设施专家张驰均已加入腾讯。 为了应对挖角，字节跳动在本月向全球 Seed 员工提供了每股 13 美元的豆包特股认购选择权，相比 2025 年底的每股 10 美元上涨了近 30%，上涨反映出豆包大模型商业化与模型性能的提升。尽管期权机制能够提供丰厚的潜在收益，但组织内部也存在争议。部分内部人士透露，Seed 实验室团队规模已达 2000 人，内部组织架构面临官僚化和职责碎片化挑战。相比之下，腾讯的 AI 团队规模更小，且向研究人员承诺了更广泛的自主权、管理职责和更具吸引力的薪酬方案。除腾讯外，DeepSeek 同样面临激烈的人才竞争压力，目前正向投资者募集首轮资金以保留核心团队，而字节跳动在 2025 年也曾成功挖角多位 DeepSeek 的基础设施核心成员。

Human communication rely on expression, timing, attention, and presence to build trust and understanding. A real-time human avatar add a critical visual layer to voice agents, making interactions more natural, more engaging: youtube.com/watch?v=xOnMYhca…

Just open-sourced VibeTensor — the first deep learning system fully generated by an AI agent, with 0 lines of human-written code: github.com/NVlabs/vibetensor It’s a working DL system with RCU style dispatcher, a cache allocator and reverse-mode autograd. The agent also invented a Fabric Tensor system — something that doesn’t exist in any current framework. The Vibe Kernel includes 13 kinds and 47k LOC of generated Triton and CuteDSL kernels with strong performance. VibeTensor was generated by our 4th-generation agent. It shows a “Frankenstein Effect”: the system is correct, but some critical paths are designed in inefficient ways. As a result, performance isn’t comparable to PyTorch. I haven’t written a single line of code since summer 2025. I started this effort after @karpathy 's podcast — I didn’t agree with his arguments, so Terry Chen and I began using it as a stress test for our agents. The “Frankenstein Effect” ended up exposing some of our agent’s limitations — but the direction is clear.