The smoke of the 618 shopping festival has cleared, leaving behind the jubilation of bargain hunters and the lamentations of wallets. But don't fret, the new darling of the AI world—Agent intelligences—might just be your remedy for regret.
Since OpenAI's GPTs set the trend, platforms for intelligent agents have sprung up like mushrooms after the rain, allowing one-click summoning of various agents to collaborate and create personalized smart assistants. Indeed, on the platform of Zhipu Qingyan, alongside the official basic tools, third-party service APIs are also joining the fray.
For instance, plugins that generate mind maps with a single click, or the proximity search feature of AutoNavi Maps, to recommendations on what's worth buying—these agents are subtly changing our consumer decisions.
Speaking of what's worth buying, this e-commerce consumer guide app is a master of saving money, offering everything from price comparisons and coupon collection to product recommendations. And when combined with AI, does this herald a new business model?
We compared the agents on Kimi+ with the default Kimi smart assistant and found that AI shopping agents present information more uniformly and clearly. They not only provide configuration details but also show the final prices during the 618 period, down to the differences in prices for different configurations and colors.
The "Little Value" within the what's worth buying app has taken the integration of AI and e-commerce to a new level. It not only recommends products but also allows one-click redirection to e-commerce platforms, adding items to the cart, and achieving a seamless transition from recommendation to purchase.
But what is the technological support behind this? Wang Yunfeng, CTO of what's worth buying technology, revealed that their AI strategy involves building a multi-agent architecture, with a self-developed consumer big model as the main force, while also leveraging other big models for assistance. This architecture decomposes different sub-tasks to specialized agents, enabling task collaboration.
Wang also mentioned that the consumer big model of what's worth buying has chosen to enhance the model's professional skills, sacrificing some unrelated abilities. This differentiated choice offers new insights for other internet companies.
Therefore, the special feature of AI shopping agents lies not only in their provision of more accurate consumer advice but also in how they integrate with e-commerce to create new business models. In this big model ecosystem, how vertical industries position themselves, what's worth buying's AI practice may provide the answer.