An AI agent tasked with buying a pack of eight Pears Pure & Gentle bathing bars highlighted the complexities of agentic commerce, according to medianama.com. The experiment was part of preparations for an agentic commerce hackathon scheduled for August 1, where the goal is to build AI that can autonomously negotiate and complete purchases online.
The experiment involved instructing the AI agent, named Saki, to shop for a single product while optimizing for the lowest cost across sellers. The process required the agent to navigate product selection, evaluate sellers, handle login procedures, calculate final prices after offers, choose delivery addresses, and complete payment. The exercise revealed that even simple shopping tasks involve multiple layers of complexity for AI agents.
This experiment sheds light on the challenges in developing AI agents capable of fully autonomous online shopping and negotiation. While agentic commerce promises automation, tasks such as product standardization, price evaluation, and transaction handling remain intricate. The hackathon aims to push the boundaries of AI negotiation bots that can manage these complexities end-to-end, a step beyond current AI shopping assistants.
The hackathon where participants will build AI negotiation bots is set for August 1, providing a platform to advance research and development in agentic commerce, as reported by medianama.com.