Darren LI

To address these concerns, several of the leading model companies are working on improved steering —a way to place better controls on LLM outputs—to focus model outputs and help models better understand and execute on complex user demands.

sim2real确实不是大问题了,但是zero-shot/few-shot sim2real永远是大问题,因为zero/few-shot learning本身就是大问题

我希望未来Bolt投出的几十个项目中,最后能够做出AI-native的super global app,对应1-2个十亿美元以上的公司。这其实就是Bolt成功的体现了

具体到Bolt,因为单笔金额灵活、数量会比较多,我告诉LP,可以把Bolt看成一个一揽子计划(basket),最终的总体投资额(包括后续加码)会在2000万到5000万美元。这些项目中最终能几个有在contextual data、personal data、collaborative data这三个方面实现积累并且与时俱进的团队,最终的收益也不会差。而且, 未来应用公司创造的总价值会远超过大模型。

While same-day delivery and free returns are eroding much of the rest: we’ve now reached the point that it’s not clear that there is anything that cannot be bought online. But actually, the real barrier now is often not touching it, but knowing about it. These are things where a lot of the role of a physical shop is curation, recommendation and... See more