Darren LI
@darrenli
Creator Economy, AI & retail-tech investor | J.D. & J.M. | PNG collector | Musical lover | Happy to chat, using Cal.com link below to book calls (https://cal.com/darrenli)
twitter.comDarren LI
1. Nearly every company in the Sequoia network is building language models into their products.
Lauren Reeder • The New Language Model Stack
【SaaS+AI】
- Vertical SaaS是最好的做产品的机会 :1)找到垂直行业,持续获得用户内部/反馈且对模型训练有用的数据,垂类数据更容易对模型通过reward改进,让产品效果提升更快;2)获客成本相对低,尤其是后续拓展产品
- 重视how to access to data :如果SaaS企业非常重视AI,就一定要建立数据飞轮,用户的数据相当于免费标注:找人标注会很贵,但如果用户数据本身就是标注,就可以大幅提高迭代效率(如特斯拉)带着B2B2C、挖掘C端需求的思维做saas产品,比如客服 saas 的价值是解决C端问题
- 做企业级客户要非常非常有耐心
Article
Specifically, we believe the Pixar of the next century won’t emerge through traditional film or animation, but rather through interactive video . This new storytelling format will blur the line between video games and television/film – fusing deep storytelling with viewer agency and “play,” opening up a vast new market.
Jonathan Lai • The Next Generation Pixar: How AI will Merge Film & Games | Andreessen Horowitz
Key unlock: users can better tailor the outputs of LLMs.
Sarah Wang • The Next Token of Progress: 4 Unlocks on the Generative AI Horizon
AgentBench: Evaluating LLMs as Agents
Evaluating Large Language Models (LLMs) as agents in interactive environments, highlighting the performance gap between API-based and open-source models, and introducing the AgentBench benchmark.
arxiv.orgGenerative AI and
AgentBench: Evaluating LLMs as Agents
Great framework for understanding the current stack of AI agents companies; most companies only aggregate low-hanging fruits of tools and memories components (e.g., vector database) without extra works of planning built upon off-shelf LLMs
样本效率做到了之前的deepmind的5-6倍。预计能在10h内实现双足的实机学习
在机器人领域应用深度强化学习，目前主流的一些思路是什么？ - 知乎
